Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An examination of underlying motor planning and execution processes that are reflected in variations… Cheng, Darian T. 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2017_november_cheng_darian.pdf [ 2.68MB ]
Metadata
JSON: 24-1.0355397.json
JSON-LD: 24-1.0355397-ld.json
RDF/XML (Pretty): 24-1.0355397-rdf.xml
RDF/JSON: 24-1.0355397-rdf.json
Turtle: 24-1.0355397-turtle.txt
N-Triples: 24-1.0355397-rdf-ntriples.txt
Original Record: 24-1.0355397-source.json
Full Text
24-1.0355397-fulltext.txt
Citation
24-1.0355397.ris

Full Text

AN EXAMINATION OF UNDERLYING MOTOR PLANNING AND EXECUTION PROCESSES THAT ARE REFLECTED IN VARIATIONS IN CORTICAL COMPONENTS DURING TARGET ENCODING AND EXECUTION OF GOAL DIRECTED REACHING MOVEMENTS  by  Darian T. Cheng A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate and Postdoctoral Studies  (Kinesiology) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) August 2017 © Darian T. Cheng, 2017 ii  Abstract In this thesis the neural correlates associated with reaches under varying conditions of visual feedback, delay and movement difficulty were examined. The events of interest, wherein these correlates are observed were: 1) the encoding of a target (P3); and 2) the execution of a reaching movement towards the target (motor MP & N4).  In study 1, the neural correlates associated with variations in task difficulty and target visible vs. target occluded were examined. The results of target encoding showed that P3 was greater when observed prior to reaches without than with vision of the target. Results of MP and N4 both varied based on task difficulty, as reaches to the far target yielded larger amplitudes than reaches to the close target. In study 2, the effects of visual feedback, delay condition, and presentation schedules (i.e., blocked vs. randomized) on target encoding and movement related cortical potentials was examined. For target encoding, component P3 failed to yield any significant differences across all vision, delay and presentation schedules. As for the movement related potentials, significant effects of delay were observed for component MP in the randomized protocol (i.e., larger MP in the long vs. the short delay), but not the blocked protocol. The analysis of N4 for the randomized protocol yielded a main effect of vision, as reaches with vision of the target attained greater negative amplitudes as compared with reaches without vision of the target. In the blocked protocol the analyses of N4 yielded a main effect of delay period as movements following long delay periods resulted in larger negative amplitudes. In study 3, the effects of real-time vision of the hand and target during a reaching movement on target encoding and movement execution were investigated. The findings revealed that the P3 component was modulated by the visibility of target. For the movement related cortical potentials, larger amplitudes for N4 were yielded for reaches without vision of iii  the hand vs. reaches with vision of the hand. The three studies taken together provide insight into the neural events associated with goal-directed manual aiming under various reaching conditions.    iv  Lay Summary In this thesis the electrical activity emitted by the brain and observed over the scalp was examined while individuals were performing reaching movements. The rationale for these studies was to observe differences in electrical activity relating to the different reaching conditions in order to gain a better idea of the underlying processes that allow individuals to plan and execute reaching movements with their arms and hands. In the first study, the difficulty of the reaching movement was manipulated by having participants reach to targets of varying sizes and distances. In the second study, the delay period between seeing the target and reaching to the target was manipulated. In the finally study, vision of the reaching hand/limb was manipulated. In all, the studies provided greater insight into the processes associated with planning and execution of reaching movements.   v  Preface This study was approved by the University of British Columbia research ethics board (H07-01734). The experimental designs were devised by Gord Binsted. The data was collected at the University of British Columbia Okanagan (Kelowna, BC) by Darian Cheng, Krista Fjeld and Robert Hermosillo.    vi  Table of Contents Abstract ........................................................................................................................................... ii Lay Summary ................................................................................................................................. iv Preface............................................................................................................................................. v Table of Contents ........................................................................................................................... vi List of Figures ................................................................................................................................. x 1 General Introduction ................................................................................................................ 1 1.1 Human Movement .......................................................................................................... 3 1.1.1 Neural Processes Underlying Planning and Control of Movement .......................... 3 1.1.2 Neural Underpinnings of Movement ........................................................................ 6 1.1.3 Using a Visual Memory Representation as a Proxy for Vision ................................ 8 1.2 Neural Processes and Anatomical Substrates for Movement ....................................... 11 1.2.1 Visual Encoding of the Environment ...................................................................... 12 1.2.2 Movement Planning ................................................................................................ 13 1.2.3 Movement Execution and Control .......................................................................... 14 1.3 Evoked Brain Potentials ................................................................................................ 16 1.3.1 ERP Components Evoked by Visual Stimuli ......................................................... 16 1.3.2 N2 and P3 Components .......................................................................................... 17 1.3.3 Variations in the P3 Component and Impending Movement Difficulty ................. 21 1.3.4 Movement-Related Cortical Potentials ................................................................... 23 1.3.5 Motor-Related Cortical Potentials and with Error Detection ................................. 27 1.3.6 fERN and Reaching Movements ............................................................................ 29 1.3.7 Motor-Related Cortical Potentials and Visual Feedback Use................................. 31 1.3.8 Hierarchical Model of Error Detection ................................................................... 33 1.3.9 Modulation of Cortical Potentials during Memory-Guided Reaches ..................... 34 1.4 Overarching Questions.................................................................................................. 39 1.4.1 Target Encoding ...................................................................................................... 39 1.4.2 Movement Execution .............................................................................................. 40 1.5 Experimental Approaches ............................................................................................. 42 1.5.1 Study 1: The Neural Correlates of Visually and Memory-Guided Reaches Under Varying Task Difficulty ......................................................................................................... 42 vii  1.5.2 Study 2: The Neural Correlates of Reaching Under Varying Delay and Visual feedback 43 1.5.3 Study 3: The Neural Correlates of Real-Time Vision of the Hand and Target During a Goal-Directed Reaching Movement ....................................................................... 44 2 Study 1: The Neural Correlates of Visually and Memory-Guided Reaches Under Varying Task Difficulty .............................................................................................................................. 46 2.1 Introduction ................................................................................................................... 46 2.1.1 Target Encoding ...................................................................................................... 51 2.1.2 Movement Evoked Potentials ................................................................................. 52 2.2 Methods......................................................................................................................... 54 2.2.1 Participants .............................................................................................................. 54 2.2.2 Task ......................................................................................................................... 54 2.2.3 Apparatus ................................................................................................................ 54 2.2.4 Procedure ................................................................................................................ 55 2.2.5 Data Reduction and Analysis .................................................................................. 56 2.3 Results ........................................................................................................................... 59 2.3.1 Behavioural Results ................................................................................................ 59 2.3.2 Electroencephalographic results ............................................................................. 63 2.3.2.1 Target Encoding ...................................................................................................... 63 2.3.2.2 Movement Evoked Potentials ................................................................................. 65 2.4 Discussion ..................................................................................................................... 70 2.4.1 Behavioural Results ................................................................................................ 70 2.4.2 Target Encoding ...................................................................................................... 71 2.4.3 Movement Evoked Potentials ................................................................................. 72 2.5 Conclusion .................................................................................................................... 73 Bridging Summary ........................................................................................................................ 75 3 Study 2: The Neural Correlates of Reaching Under Varying Delay and Visual feedback ... 77 3.1 Introduction ................................................................................................................... 77 3.1.1 Target Encoding (P3) .............................................................................................. 81 3.1.2 Movement Evoked Potentials (MP and N4) ........................................................... 82 3.2 Methods......................................................................................................................... 83 viii  3.2.1 Participants .............................................................................................................. 83 3.2.2 Task ......................................................................................................................... 83 3.2.3 Apparatus ................................................................................................................ 84 3.2.4 Procedure ................................................................................................................ 85 3.2.4.1 Experiment 1 (Random Procedure) ........................................................................ 85 3.2.4.2 Experiment 2 (Blocked procedure) ......................................................................... 87 3.2.5 Data Reduction and Analysis ........................................................................................ 87 3.3 Results ........................................................................................................................... 90 3.3.1 Experiment 1: Random Protocol ............................................................................. 90 3.3.1.1 Behavioural Results ................................................................................................ 90 3.3.1.2 Electroencephalographic Results ............................................................................ 93 3.3.2 Experiment 2: Blocked protocol ............................................................................. 96 3.3.2.1 Behavioural Results ................................................................................................ 96 3.3.2.2 Electroencephalographic Results ............................................................................ 99 3.4 Discussion ................................................................................................................... 102 3.4.1 Behavioural Findings ............................................................................................ 103 3.4.2 Target Encoding (P3) ............................................................................................ 103 3.4.3 Early Movement Potential (MP) ........................................................................... 104 3.4.4 Late Movement Potential (N4) ............................................................................. 106 3.5 Conclusions ................................................................................................................. 107 Bridging Summary ...................................................................................................................... 109 4 Study 3: The Neural Correlates of Real-Time Vision of the Hand and Target During a Goal-Directed Reaching Movement .................................................................................................... 111 4.1 Introduction ................................................................................................................. 111 4.1.1 Target Encoding .................................................................................................... 116 4.1.2 Movement Execution ............................................................................................ 117 4.2 Methods....................................................................................................................... 118 4.2.1 Participants ............................................................................................................ 118 4.2.2 Task ....................................................................................................................... 118 4.2.3 Apparatus .............................................................................................................. 118 4.2.4 Procedure .............................................................................................................. 119 ix  4.2.5 Data Reduction and Analysis ................................................................................ 121 4.3 Results ......................................................................................................................... 123 4.3.1 Behavioural Results .............................................................................................. 123 4.3.2 Electroencephalographic Results .......................................................................... 128 4.3.2.1 Target Encoding .................................................................................................... 128 4.3.2.2 Movement Execution ............................................................................................ 130 4.4 Discussion ................................................................................................................... 133 4.4.1 Replication of Previous Behavioural Findings ..................................................... 134 4.4.2 Target Encoding .................................................................................................... 135 4.4.3 Movement Execution ............................................................................................ 136 4.5 Conclusions ................................................................................................................. 137 4.6 Summary of Findings .................................................................................................. 138 4.6.1 Study 1: The Neural Correlates of Visually and Memory-Guided Reaches Under Varying Task Difficulty ....................................................................................................... 138 4.6.2 Study 2: The Neural Correlates of Reaching Under Varying Delay and Visual feedback 142 4.6.3 Study 3: The Neural Correlates of Real-Time Vision of the Hand and Target During a Goal-Directed Reaching Movement ..................................................................... 146 4.7 Study Integration ......................................................................................................... 150 4.7.1 Implications on Target Encoding .......................................................................... 150 4.7.2 Implications on Movement Related Cortical Potentials ....................................... 152 4.7.3 Implications on Movement Planning and Control ................................................ 156 Bibilography ............................................................................................................................... 159     x  List of Figures Figure 1.1. From Elliott et al. (2010) and reproduced with permission. This schematic depicts an updated view of the processes underlying human limb movements through the incorporation of internal models into the hybrid model of open and closed loop control first proposed by Woodworth (1899). ......................................................................................................................... 6 Figure 1.2. From Binsted et al. (2006) and reproduced with permission. Standard deviation of movement endpoints on the Y- axis and time on the X-axis. Individuals were aiming to targets to the left and right until vision was removed, as indicated by the dashed line. What is noticeable is the initial increase in error when vision is first removed, followed by a plateau for 5 seconds followed by another increase in error. .......................................................................................... 11 Figure 1.3. A schematic of the brain areas sub-serving the dorsal and ventral stream, adapted from Milner and Goodale (1995). The flow of information starts at the eye, goes through the lateral geniculate nucleus (LGN) and to the primary visual cortex (V1). From V1, the visual information is either streamed dorsally to parietal regions of the brain or ventrally towards temporal regions............................................................................................................................ 13 Figure 1.4. From Kourtis et al. (2012) and and reproduced with permission. The figure shows grand averaged VEP waveforms derived from pooled electrode sites Pz, P1, P2, POz, PO3 and PO4. Notice the amplitude of the P3b was greater when individuals were observing the cue that indicated greater task difficulty (i.e., thick black line) as compared to the cue indicating lesser difficulty (i.e., dashed line). .......................................................................................................... 22 Figure 1.5. From Kirsch, Hennighausen & Rosler, 2010; and reproduced with permission. The figure shows a guided limb movement to a physical stop on the left. This was followed by a xi  replication of the limb movement without the stop on the right, where individuals attempted to replicate the previous movement. Labelled on the graph to the left are the peaks associated with the limb movements. ..................................................................................................................... 24 Figure 1.6. From Kirsch & Hennighausen (2010) and reproduced with permission. The above figure depicts the displacement (p) velocity (v) and acceleration (a) profiles of reaching to different movement amplitudes. The bottom figure shows the motor related cortical potential related to the above conditions. Notice the amplitude of the second component (N4), scales with the amplitude of the target, with reaches of longer amplitudes yielded larger N4 amplitudes. .... 26 Figure 1.7. Motor related cortical potentials associated with the shooting task. Note the modulation of the second negative component occurring approximately 250-300ms following movement onset. Figure from Torrecillos et al. (2014), reproduced with permission. ................ 30 Figure 1.8. Taken from Krigolson & Holroyd (2007). This figure depicts the grand averaged waveform generated over electrode Pz, over the parietal region of the scalp during a reaching movement. Notably the large dashed control condition waveform depicts the parietal activity observed during reaching movement (i.e., guiding a cursor) without a target perturbation. The correctable and uncorrectable conditions are trials where the target perturbation occurs with and without the ability to correct the trajectory of the cursor, respectively. ....................................... 32 Figure 1.9. Taken with permission from Krigolson et al. (2012). Grand averaged ERP waveforms locked to target presentation for the proximal target. This graph illustrates that the memory guided waveform is markedly suppressed as compared to the visually guided waveform. The authors suggest that the suppression of the memory-guided trace may have to do with the involvement of motor planning processes when individuals are encoding the target, which is supports the motor hypothesis. ..................................................................................................... 36 xii  Figure 1.10. Taken with permission from Krigolson et al., (2012). ERP average waveform time locked to movement start. Notice the second peak observed at approximately 300 ms post movement initiation is lower in magnitude than in the visually guided condition. This coincided with lower deceleration values in the memory guided reach condition........................................ 38 Figure 2.1. The sequence of events occuring on a given trial. Participants first fixated on the fixation cross on the right and positioned the stylus over the home position on the left of the tablet. This was followed by the target preview of 250ms, when the target preview was marked as an event. Following the target preview, vision of the target was removed for 2 s. A ‘go’ tone was presented following the delay period, which informed participants to move to the target location. On FV trials, the target reappeared with the ‘go’ tone, while on NV trials, the target did not reappear. .................................................................................................................................. 57 Figure 2.2. Average movement time (ms) with standard error bars, as a function of vision condition, target size across the close and far targets ................................................................... 60 Figure 2.3. Acceleration (mm/s/s) profiles as a function of target amplitude (Close target and Far target). ..................................................................................................................................... 61 Figure 2.4. Variable error in movement amplitude (mm) in the primary x-axis with standard error bars as a function of vision condition and target size (FV = target vision, NV = no target vision) ........................................................................................................................................... 62 Figure 2.5. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement. .................................................... 63 Figure 2.6. Grand average ERP waveforms locked to the onset of the target preview as a function of vision condition (FV: target vision; NV: no target vision). The P3, as indicated by the xiii  triangle, shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition ............................................................................................................... 64 Figure 2.7. Average peak amplitudes of the P3 component observed at the pooled parietal electrodes. Analysis of the peak amplitudes revealed a main effect of vision, as ........................ 65 Figure 2.8. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition and target amplitude for electrode Fz. The N4 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition. ...................................................................................................................................... 66 Figure 2.9. Grand average amplitudes of MP as a function of vision condition and target amplitude with standard error bars ............................................................................................... 67 Figure 2.10. N4 grand average amplitudes as a function of target vision (FV: target vision; NV: no target vision) with standard error bars ............................................................................. 67 Figure 2.11. Grand average ERP waveforms, at electrode Cz, time-locked to the onset of the reaching movement as a function of vision, target size and target location. The peak of interest, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition. ......................................................................................................... 68 Figure 2.12. Averages for the negative component observed at approximately 170 ms post movement onset, as a function of vision, target size and movement amplitude with standard error bars. ............................................................................................................................................... 69 Figure 3.1. The visual display containing the home position on the left, fixation to the right as well as all possible target locations. From the top of the target array to the bottom, the targets are T3, T2, T1, B1, B2, B3) ................................................................................................................ 85 xiv  Figure 3.2. A depiction of the sequence of events on a typical trial. The trial started after the participant foveated on the fixation cross and positioned the stylus onto the home position. A cue was subsequently presented, which informed the participant of the delay period and the visual feedback condition on the impending movement. Following the cue, there was a preview of the target for 250 ms and a delay period (2 s or 5 s) during which the target was removed. Following the delay period, an imperative tone was presented, which informed the participant to move to the target location. On trials with vision, the target reappeared, whereas on trials without vision the target did not reappear. ............................................................................................................ 86 Figure 3.3. Variable error in movement amplitude (mm) with standard error bars as a function of vision condition (FV = full vision, NV = no-vision) ................................................................ 91 Figure 3.4. Constant error in movement amplitude (mm) with standard error bars as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay) ................................ 92 Figure 3.5. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement. .................................................... 93 Figure 3.6. Grand average ERP waveforms of electrode PO3, time locked to the onset of the target preview as a function of vision condition. The P3 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. ...................................................................................................................................... 94 Figure 3.7. Grand average ERP waveforms time locked to the onset of the reaching movement as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay). The N4 peak, xv  indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition.................................................................................. 95 Figure 3.8. Averaged N4 peak amplitude as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay) with standard error bars. ......................................................... 96 Figure 3.9. Constant error in movement amplitude (mm) with standard error bars as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay) ................................ 98 Figure 3.10. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement. (FV = full vision, NV = no-vision)....................................................................................................................................................... 99 Figure 3.11. Grand average ERP waveforms locked to the onset of the target preview as a function of vision condition. The P3 peak, indicated by the arrows shows the location of the peak where the 50 ms window was taken for averaging of each condition. ....................................... 100 Figure 3.12. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition. The N4 peak, indicated by the arrows shows the location of the peak where the 50 ms window was taken for averaging of each condition. ............................... 101 Figure 3.13. Grand average N4 peak amplitude as a function of vision condition with standard error bars. .................................................................................................................................... 102 Figure 4.1. The experimental procedure used in the study. Note that for target preview, the target remained on for Hand/Target and Target only conditions, whereas for the Hand Only, No xvi  hand/No target conditions the target was extinguished. Also two target amplitudes were used (Close: 40 and 44 cm) ................................................................................................................. 120 Figure 4.2. Time after peak velocity (ms) with standard error bars as a function of vision condition and target location. ...................................................................................................... 124 Figure 4.3. Velocity profiles as a function of each reaching condition .................................. 125 Figure 4.4. Variable error in movement amplitude (mm) with standard error bars as a function of vision condition. ..................................................................................................................... 126 Figure 4.5. Constant error in movement amplitude (mm) with standard error bars as a function of vision condition. ..................................................................................................................... 127 Figure 4.6. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement. .................................................. 128 Figure 4.7. Grand average ERP waveforms locked to the onset of the target preview for all vision conditions at electrode PO3. The P3 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. The dashed lines indicate the control condition where participants were not required to perform a subsequent movement to the target. ............................................................................................ 129 Figure 4.8. Bar graph depicting the grand average P3 amplitudes as a function of preview condition with standard error bars. ............................................................................................. 130 Figure 4.9. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition. The N4 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. The dotted line indicates the control condition where participants were not required to perform a subsequent xvii  movement to the target, while the vertical dashed line indicates the time where the average movement end occurred. ............................................................................................................. 131 Figure 4.10. Bar graph depicting grand averages of the N4 peak, as a function of reaching condition, with standard error bars. ............................................................................................ 132 Figure 4.11. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition. The peak, as indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. The dotted line indicates the control condition where participants were not required to perform a subsequent movement to the target, while the vertical dashed line indicates the time where movement end occurred....................................................................................................................................... 133 Figure 4.12. A schematic depicting the components of interest (P3, MP and N4) and the factors from each study that influenced the amplitude of these components. ............................ 155 Figure 4.13. Taken with permission from Elliott et al., (2010). This figure depicts the different neural processes that occur prior to and during the execution of a limb movement. ................. 157   1  1 General Introduction How does the brain plan and control limb movements? For healthy individuals, performing movements is relative easy despite being in an environment that is constantly changing. These changes occur from movement of objects in the environment or by changes in our position in the environment. As such, humans have a highly evolved sensorimotor system that allows for the rapid adaptation to change and the maintenance of highly precise actions.   One of the earliest systematic studies of human goal-directed movement was performed by Woodworth (1899). In these experiments, he had subjects performing reciprocal aiming movements between two target lines on smoked paper that was scrolled along at a constant speed. The reciprocal aiming movements were performed with different movement velocities and visual conditions (i.e. eyes open or eyes closed) to manipulate the task demands. These experiments yielded two core findings. First, when aiming with eyes open, corrections to the trajectory of the stylus were observed when the participants homed-in on the target. Second, with increasing speed the accuracy of the movements was just as poor with eyes open or closed. This led Woodworth (1899) to propose that aiming movements are composed of two phases: an initial impulse phase where a movement is planned ahead of time, followed by a current control phase, where vision is used to amend the position of the limb as it nears the target. Also, he observed that the processing of visual information takes a specific amount of time to occur (i.e., approx. 450 ms, a value which has been shown to be an overestimation cf. Whiting, Gill & Stephenson, 1970). Therefore, if the speed of the movements surpassed the speed with which visual information could be processed, aiming accuracy and precision was compromised. While the understanding of planning and control of limb movements has progressed since Woodworth (1899), the central tenets of this dual-phase model of manual aiming remain current and are 2  prominent in most extant theories of sensorimotor control. More recently however, the idea of predictive control and forward modeling (Wolpert & Ghahramani, 2000) has been added to explain how the system can account for early trajectory amendments which the previous pre-planned and control hybrid model could not explain (Desmurget & Grafton, 2003; Desmurget & Grafton, 2000).  The overarching theme of this thesis is to examine the brain activity that underlies the planning and execution of goal-directed reaching movements. To date there has been significant behavioural research examining the planning and control of upper limb movements: from studies with individuals with acquired brain injuries (Milner & Goodale, 1995), to studies with non-human primates (Ungerleider & Mishkin, 1982), and healthy subjects (Westwood, Heath & Roy, 2003). Although these studies have made large inroads into understanding the underlying processes of planning and control of movement, more recent imaging studies have revealed more details on the brain regions involved. Namely, functional magnetic resonance imaging (fMRI) has allowed us to observe the oxygen metabolism associated with the brain regions that are active during the process of planning and control of movement (Ball et al., 1999; Lotze et al., 1999). While this method is spatially precise, it lacks the temporal resolution necessary to examine the millisecond-to-millisecond changes in brain activity that occurs within rapid goal-directed movements like reaching.  Notably, it would be very difficult to examine the different brain regions that are active at specific time points prior-to and during the reaching movement. As a result, methods such as electroencephalography (EEG), which have excellent temporal resolution, present an opportunity to uncover the time-specific activity associated with these processes. Thus, the objective of this thesis is to use EEG to identify and examine the brain potentials associated with reaching movements and how they vary as a function of different 3  reaching and environmental parameters. By studying these brain potentials we will better understand how and when the neural processes facilitating reaching movements occur, either prior-to and/or during the movement. Furthermore, these findings can be related back to the neural structures that have been identified for movement in previous imaging and lesion studies. To provide the necessary background context to this research, this Introduction begins with a brief presentation of the behavioural findings pertaining to goal-directed reaching. This will provide insight into the processes underlying the planning and control of human movement. The Introduction continues with an account of the seminal neurobehavioural work relating perception to action as a model for understanding the anatomical underpinnings of goal-directed action. Lastly, an overview is provided of the event-related potentials that accompany the visual encoding of stimuli and cortical activity relevant to the planning and execution of movement.  1.1 Human Movement  1.1.1 Neural Processes Underlying Planning and Control of Movement A recent body of work, using the concept of internal models has provided a unifying account of how reaching movements are generated. The main concept of internal models proposes that the central nervous system makes sensory predictions of the outcome of the movement from the motor output by way of forward modeling. These sensory predictions are based on previous experience with performing the given task, allowing for early corrections to the limb to occur which are faster than corrections stemming from closed-loop feedback systems (Desmurget & Grafton, 2000; Kawato et al., 1999; Wolpert & Ghahramani, 2000). As the sensory prediction 4  aspect is the main idea brought forth by the concept of internal models, there are many ideas pertaining to movement planning and execution that have been incorporated from open and closed-loop motor control theories (Meyer et al, 1988; Plamondon, 1995; Plamondon & Alimi, 1997; Also see Figure 1.1). Firstly, before the initiation of a movement a motor plan is generated through inverse modeling. This idea of a motor plan is similar to the ideas put forth by proponents of open-loop theories (Henry & Rogers, 1960; Keele, 1968; Plamondon, 1995; Plamondon & Alimi, 1997; Wadman, Van Der Gon, Geuze & Mol, 1979), which propose that the nervous system translates a desired movement goal-an abstract objective (i.e., reaching a target), into physical action by specifying and subsequently executing motor commands containing different sequences of muscle activations. In choosing the appropriate plan however, priority is given to optimizing the smoothness of the movement. Hence limb movements often exhibit stereotyped characteristics, such as the speed and shape of the limb trajectory, which is observed across many individuals.  In addition to inverse modelling, the internal models concept introduces the process of forward modelling which simulates the sensory consequences of the motor commands by comparing the motor output with and the sensory inflow of the current limb position. This forward prediction allows for fast non-feedback based corrections to occur (Desmurget & Grafton, 2000; Wolpert, & Gharhrami, 2000). As indicated above, a motor plan is generated through inverse modelling to achieve a desired movement outcome. As the movement is being initiated, motor outflow from motor areas is sent to the acting limb as well as to the posterior parietal cortex and cerebellum by way of an efference copy. This is where the forward modelling predicts the endpoint location of the movement based on the current state of the limb (via sensory inflow) and the efference copy. This prediction is made based on previous experience 5  and knowledge with performing these actions, called priors. If the forward model predicts that the desired movement outcome will be not achieved, an error signal is generated. In response to this error signal, corrections to the motor command are made before or shortly after the initiation of the movement. As this revised motor command is being sent, the forward model then again makes another prediction and executes another correction if needed. As a result, forward modeling is a continuous process that occurs throughout the movement.  Furthermore, once enough time has elapsed for visual and proprioceptive feedback to be received from the closed-loop, feedback systems are used to fine-tune the limb trajectory as it homes-in to the desired endpoint location. In addition, this sensory feedback is also used to assess whether the predicted sensory feedback generated during the forward modeling process indeed reflected the actual sensory outcome of the movement. If a discrepancy between the two exists, the prior that pertains to this action will be updated. This is done by combining the previous sensory predictions of the prior and the sensory events encountered on the most recent attempt, through Bayesian integration (Wolpert, & Gharhrami, 2000). In sum, the addition of predictive control along with open and closed loop models provides a more unified understanding of goal-directed human movement.  6   Figure 1.1. From Elliott et al. (2010) and reproduced with permission. This schematic depicts an updated view of the processes underlying human limb movements through the incorporation of internal models into the hybrid model of open and closed loop control first proposed by Woodworth (1899). 1.1.2 Neural Underpinnings of Movement In this section, literature pertaining to the brain structures that underlie goal-directed reaching movements is covered. This will provide the necessary background on the neural structures that are involved in the planning and execution of the reaching movement, and will offer some context as to how and where the observed neural activity at the scalp may be coming from. At 7  present, the most well rounded explanation for the structures involved with movement planning and execution lies with the perception/action explanation put forth by Milner & Goodale (1995). Hence, the neural structures in the perception/action streams will be covered and applied to a real-world scenario.  In parallel to the observations of the behavioural features of reaching movements, a body of research has emerged over the last 30 years that depicts the neural underpinnings of goal-directed reaching movements – specifically when under the control of vision. For instance, single cell examinations of the neurophysiological anatomy of non-human primates, Ungerleider and Mishkin (1982) uncovered specialized brain areas that subserve different aspects of the visual processes associated with action control. These areas, which lay downstream from the primary visual cortex are separated into two main divisions: 1) a ‘where’ division that contains structures in the parietal region of the brain and 2) a ‘what’ division that contains temporal structures. The ‘what’ stream generally enables the visual system to recognize objects and places and contains brain structures specialized for colour processing (area V4) and pattern/shape recognition (i.e., area LO). Areas of the ‘where’ stream generally subserve the functions of locating and guiding movements towards objects in the environment, and contain areas that specialize in motion detection (i.e., MT: Middle Temporal), and areas dedicated to the planning and control of movement (i.e., PPC: posterior parietal cortex). These two streams were later identified in humans as the ventral (i.e. perception, ‘what’) and dorsal (i.e., action, ‘how’) stream (Goodale & Milner, 1992; Milner & Goodale, 1995). Seminal patient work by Goodale and Milner (1992) documented the deficiencies of visual perception with patient DF who acquired brain damage to the ventral stream. She was diagnosed with visual form agnosia, which was characterized by deficits in identifying objects but still retaining the ability to use objects. Conversely, patient RV 8  who sustained damage to the dorsal stream displayed optic ataxia; she retained the ability to identify objects, but when asked to act with the object (e.g., if asked to put a card into a slot: mailing task) she was unable to successfully complete the task.  The dissociations between vision for action and perception are also apparent in normal healthy subjects when exposed to visual illusions (e.g., Muller-Lyer, Titchener circles; Aglioti, DeSousa & Goodale, 1995). The perceptual system is posited to make relative size comparisons between objects in the environment (i.e., allocentric encoding), as a result when individuals were asked to judge the size of the target that was embedded in an illusory array; they had a tendency to report the incorrect size of the object. However, when individuals are asked to act upon the target object (e.g., reaching and grasping to it) the parameters of the participants grip aperture scaled to the size of the object regardless of whether the target was embedded in the illusory array or not. During the act of reaching, individuals are hypothesized to rely on the dorsal stream to make hand/target comparisons (i.e., egocentric encoding), which is metrically accurate, resulting in actions that are scaled to the correct size of the object. In all, the dissociation between action and perception provides a unifying story on how visual information is differentially used to perceive and act, and how these functions are supported by unique anatomical substrates. 1.1.3 Using a Visual Memory Representation as a Proxy for Vision To this point, the literature covered pertained to goal-directed reaching under conditions of full vision; however there is also a body of work that has sought to examine how goal-directed reaches are performed in the absence of real-time vision (i.e., memory-guided reaching). While 9  there have been several attempts by researchers to uncover the mechanisms underlying memory-guided reaching (e.g., Elliott & Madalena, 1987; Westwood, Heath & Roy, 2003), a consensus has not yet been reached in terms of the attributes and processes that characterize memory for goal-directed movement. Early work by Elliott and Madalena (1987), suggested that individuals store highly accurate visual memory representations of the environment (iconic memory, Sperling, 1960), that are useable for movements up to 2 seconds following the occlusion of vision. That is, the sensory information about the target can be held in memory until individuals are cued to move, wherein the information is then converted to a movement plan. The authors found that as long as movements were initiated within the 2 s window after the target disappeared, the endpoint accuracy was not significantly different than movements performed with the visual target still present. However, poor movement accuracy and precision was observed when the delays were longer than 2 seconds. While this work falls under a logical framework of memory decay of visual representation, other scientists (e.g., Westwood & Goodale, 2004) have taken on a different perspective on how memory-guided movements are generated and the brain areas that underlie this function. Contrary to the findings of Elliott and Madalena (1987), Westwood and Goodale (2004) observed that goal-directed movements performed following a delay and without vision were always less accurate than movements performed with full vision. The authors, who referred to the perception/action hypothesis (Milner & Goodale, 1995) in their explanations, reasoned that accurate reaches were facilitated by the action/dorsal visual stream (i.e. pathways extending to parietal cortical structures), which required vision to be available during the course of the movement and/or at least during the initiation of the movement. If vision was not available during these periods of time, individuals could only rely on the perceptual/ventral visual stream 10  (i.e. pathways extending to temporal cortical structures) to guide their limb movement. And while the ventral stream is effective for perception of objects and faces, its ability to discriminate the veridical size and distance of objects in the visual field is rather poor. As a result, the authors suggested that aiming without vision (no vision - NV) would always be less accurate than aiming in full vision (FV), due to the reliance on different visual streams.  In an attempt to reconcile the differences between Elliott and Madalena (1987) and Westwood and Goodale (2004), Binsted et al. (2006) had participants perform a series of reciprocal aiming movements for 10 seconds and manipulated visual feedback availability. Specifically during the first 4 seconds of the trial, individuals performed the aiming task in FV followed by the remaining 6 seconds in NV. Paradoxically, results supported both lines of contentions (see Figure 1.2). That is, when vision was initially removed, there was an increase in endpoint variability associated with the loss of vision, which supports Westwood and Goodale (2004).  Following this initial increase however, the endpoint variability stabilized for the following 2-3 seconds which may reflect a real-world representation as suggested by Elliott and Madalena (1987). Therefore, the authors concluded that indeed movements performed with vision are more accurate; however the use of a visual memory representation cannot be entirely disregarded.   11   Figure 1.2. From Binsted et al. (2006) and reproduced with permission. Standard deviation of movement endpoints on the Y- axis and time on the X-axis. Individuals were aiming to targets to the left and right until vision was removed, as indicated by the dashed line. What is noticeable is the initial increase in error when vision is first removed, followed by a plateau for 5 seconds followed by another increase in error.  1.2 Neural Processes and Anatomical Substrates for Movement  When taken together, the behavioural characteristics, clinical findings, and anatomical connectivity can be combined to generate a good framework for the flow of information between brain areas (and associated processes) active when a voluntary goal-directed reaching movement is planned and executed. Here we use an example of an individual grabbing a bottle off a counter top. 12  1.2.1 Visual Encoding of the Environment First, the eyes foveate the bottle located on the counter top. Visual information pertaining to the bottle is registered by receptors in the eye, processed and transmitted to the primary visual cortex of the brain. This visual information is then streamed ventrally towards inferotemporal areas of the brain and dorsally to the posterior parietal regions of the brain (Figure 1.3). The ventral ‘what’ stream contains structures dedicated to object recognition and processes the visual stimulus for colour (V4), texture (CoS), shape (LO) and size (Cant et al., 2008). The ventral stream’s close ties to memory structures such as the hippocampus and amygdala, allows for the comparison of the observed attributes of the stimulus with the memory representations of previously observed objects. This ultimately leads to the identification of the object (Ungerleider, Courtney & Haxby, 1998). From the ventral stream processing, the individual will recognize that the object on the counter is indeed a bottle. At the same time, the dorsal ‘where’ stream functions to localize the object in the environment and also accounts for the size of the object. Namely visual information about the objects in the environment is processed in the dorsal stream for size, location, orientation (V3a), motion (MT); and is combined with incoming somatosensory information about the position of the body in space (i.e., area MST, 7). Then information from the dorsal stream is relayed toward the frontal lobe, through parietofrontal loop involving the ventral infraparietal area (VIP) in the parietal lobe and sector F4 in the frontal lobe (Rizzolatti & Gentilucci, 1988), where executive control makes sense of this information, and the individual can make a decision on whether or not to make an action.  13   Figure 1.3. A schematic of the brain areas sub-serving the dorsal and ventral stream, adapted from Milner and Goodale (1995). The flow of information starts at the eye, goes through the lateral geniculate nucleus (LGN) and to the primary visual cortex (V1). From V1, the visual information is either streamed dorsally to parietal regions of the brain or ventrally towards temporal regions.    1.2.2 Movement Planning When a decision is made to reach for the bottle, the movement goal needs to be translated into motor action. To do this pre-frontal areas of the brain signal to frontal areas of the brain the goal of the movement. In response to this, premotor, supplementary motor and motor areas contralateral to the reaching limb become active. As such when an individual is reaching with the right hand, the brain areas contralateral to the limb (i.e., right premotor and motor structures will 14  be active). First, the premotor areas generate a motor plan based upon the available information about the body in space that is gleaned from the fronto-parietal network (Praamstra et al., 2009). The supplementary motor areas help with posture and coordination of the action (Massion, 1992; Viallet, Massion, Massarino & Khalil, 1992) The motor plan is then sent to the primary motor cortex, which specifies the direction of the movement in greater detail, through population vector coding (Georgeopoulos, 1986). Once the individual is ready to act, the basal ganglia disinhibits the thalamocortical pathways (via the direct inhibitory pathway), leading to an increase in thalamocortical activity which results in an output volley of motor activity down the corticospinal tracts to the muscles involved with the reaching movement (p. 857: Kandel, Schwartz & Jessel, 2000).  1.2.3 Movement Execution and Control When the movement is triggered, motor commands are sent to the muscles of the acting limb. Also, an efference copy of the motor commands, which is sent to the somatosensory cortex, parietal reach regions (area MT, ventral infraparietal sulcus: VIP), cerebellum (Wolpert & Ghahramani, 2000). These brain structures use the efference copy to make predictions of the movement outcome of the action by way of forward modelling, through comparing the position of the limb at its current position with the efference copy. This forward prediction is based on priors, which are previous experiences with executing reaching and grasping movements to the bottle. Take for instance, that the bottle has a lid on it with water droplets on its surface, one would assume that there should be liquid in the bottle. As such, the forward model would predict from this context that the bottle is not empty and require a fair amount of force to grip to lift. If the forward model predicts that the outcome of reaching and lifting the bottle will be achieved, 15  the movement will carry on unchanged. However, if the predictions suggest that the outcome will not be achieved, due to insufficient force specified to reach and grasp the bottle or that the initial limb position was inaccurate, alterations to the motor command will be made shortly after movement initiation.  Further along in the reaching movement, real-time visual and proprioceptive information of the reaching limb and target are received through visual and somatosensory centres of the brain (Goodale et al., 1994). The visual and somatosensory information is then combined in the parietal reach region and relayed through a fronto-parietal relay loop (Connolly et al., 2003; Westwood et al., 2003). If there are any errors observed between the trajectory of reaching limb and the intended target, a corrective impulse is sent through the primary motor cortex to initiate a limb trajectory amendment (Beggs & Howarth, 1972; Meyer et al., 1988). Once the movement is completed the actual and desired outcome of the limb movement will be compared. Feedback from the movement end will be compared with the forward prediction made earlier in the movement. If a noticeable error is incurred, activity in the anterior cingulate cortex will occur, resulting in a reconfiguration of the motor plan through the basal ganglia which then amends the motor plan for the upcoming movement (Krigolson & Holroyd, 2007). Also, the prior, from which the forward prediction is based will be updated to also reflect the most recent attempt at performing this action.   Overall, behavioural (e.g., Milner & Goodale, 1995) and imaging studies (e.g; Praamstra et al., 2009) have enabled us to make these speculations on the neural processes and corresponding brain areas that underlie human movement. However, we are still not entirely sure how these brain areas interact with each other on a moment-to-moment basis – a feature examinable through electrophysiological approaches (e.g., electroencephalography, 16  magnetoencephalography). In the next section, we will introduce the basic features of electroencephalography (EEG) as well as more specific literature pertaining to EEG and motor behaviours. 1.3 Evoked Brain Potentials Electroencephalograms are often used to observe neural activity associated with the occurrence of external stimuli, ranging from visual, auditory, somatosensory, as well as more endogenous cognitive processes that occurs as a consequence of different task requirements (Luck, 2005). This section begins with a review of event-related potentials, which are electrical components elicited by the brain in response to the presentation of a stimulus (i.e., visual or auditory stimulus) in the environment. First the ERP components associated with the encoding of visual stimuli in the environment (i.e., N2 and P3), followed by the components relating to the manual reaching towards these objects (i.e., MP and N4) are discussed. Then, literature that have demonstrated variations in these components when encoding of targets of varying difficulty and visual feedback availability, as well as when performing reaches under these varied conditions of difficulty and visual feedback availability, will be presented.  1.3.1 ERP Components Evoked by Visual Stimuli When a visual stimulus is presented, the eyes first register the appearance of the target. From the eyes, the visual information of the object travels towards visual centres of the brain located in the occipital cortex and towards higher visual areas, where the visual information undergoes further processing. Several ERP components that are generated in response to the onset of the visual target are believed to reflect a variety of neural processes. The first component observed is the C1, with a latency of approximately 80 to 100ms after a visual stimulus is presented. This 17  component can be either a positive or negative component depending on the location on the visual stimulus in the visual field and is thought to reflect activity in the primary visual cortex (V1) around the calcarine fissure (Di Russo et al., 2001). Following that is the P1, which is a positive peak occurring 100ms to 130ms after the onset of a visual target and with largest amplitudes over occipital electrodes. P1 amplitudes vary with different visual parameters such as stimulus contrast and also vary with top down processes such as selective attention (Vogel & Luck, 2000); these components are thought to reflect activity occurring in the extrastriate cortex (Spehlman, 1965; Jefferys & Axford, 1972; Di Russo et al., 2001). Finally, the N1 is a negative component that occurs around 100-200 ms post stimulus and has been found to vary spatial attention and discriminative processing (Hopf et al., 2002).  1.3.2 N2 and P3 Components While early components evoked by the visual stimulus (e.g., C1, P1), can be affected by variations in stimulus contrast and location, the components that follow (e.g., N2 and P3) are associated with further processing of the sensory signals by more specialized structures as well as the involvement of higher centres of the brain.  The N2 is a negative component that occurs 150-200 ms post stimulus and has been found to reflect stimulus identification (Sutton et al, 1965; see Figure 1.4), but can also reflect an array of other effects such as reward, motor planning depending on the site and the neural generator (see Patel & Azzam, 2005). Furthermore, there is also variant of the N2 component, called the N2pc, which is a negative component that is observed at posterior electrode sites that are contralateral to the location of the visual target and reflects spatial attention (Luck & Hillyard, 1994). Of the P3 peaks there is the P3a, which is a positive-going component that is observed at approximately 250-280 ms post-18  stimulus and occurs over fronto-central areas of the scalp. The P3a is thought to reflect the orienting of attention to the task, as objects being ignored have elicited smaller P3a amplitudes as compared to stimuli being attended to (Squires, Squires & Hillyard, 1975). Also a P3b peak is observed over parietal regions of the scalp which occurs anywhere from 280-600 ms post stimulus. The P3b relates to the probability of an occurrence of an event; that is when an unlikely but expected event occurs, the amplitude of the P3b component is greater in response to this unlikely event as compared with the likely event.  The most widely used explanation for the variation in the P3b amplitude, is that the P3b reflects context updating of working memory (Donchin & Coles, 1988). Specifically, when faced with an expected but less frequently encountered stimulus, individuals need to update working memory to represent the current new environment. As a result of this updating process, larger P3b amplitudes are observed when one encounters the less frequent vs. frequently encountered stimuli. Whereas when encoding a frequently encountered stimulus, attributes in the environment remain largely unchanged from one event to the next; therefore, the visual representation of the environment will be maintained in memory. Others authors such as Isreal et al., 1980 and Kida et al., (2004), have suggested that the P3b also relates to the resource allocation of attention. This was based on the observations made when oddball tasks were performed concurrently with other tasks in dual task paradigms, the amplitudes of the P3 that were elicited by the oddball task were greatly reduced in the dual-task, as compared to the oddball only condition (Isreal et al., 1980; Kida et al., 2004). As a result, it appears the increased cognitive demand of the secondary task affects the ability of the individual to evaluate the sensory environment and update working memory when attending to the oddball task, which is manifested as a reduction in the P3 component. 19  A more recent hypothesis forwarded by Polich (2007), has used the idea of neural inhibition to unify the explanations surrounding the P3. Namely, neural inhibition allows for the enhancement of focal attention when transmitting stimulus/task information. As a result any extraneous information would be minimized, such that information presented in the environment can then be compared with the information stored in memory, which is why a larger P3 is manifested more when the less likely event occurs. Furthermore, the neural inhibition hypothesis would also explain why the P3 is affected by cognitive load, as less attentional resources would be available for resisting inhibitory control. Also, it would explain why the effect of oddball tasks on P3 amplitudes decreases with age, as the cortical processes underlying inhibitory signals decreases with age (Juckel et al., 2012). While the explanations cited above have largely tried to implicate the processes underlying the P3 solely as perceptual, as further corroborated by observations of P3 modulations during a passive viewing task (Jeon & Polich, 2001), they often leave out the fact that oddball tasks often require individuals to generate a motor response to the stimuli presented. These motor responses are often elicited by key presses or verbal responses. More recently, there has been an attempt to tie the P3b with subsequent action. Namely, Verleger, Jaśkowski and Wascher (2005) suggested the P3b mediates the perceptual processing of the stimulus with the preparation of the appropriate response. This was based on the observation that the P3 often reaches its peak as the onset of the motor response occurs. From this observation, it was suggested that the P3 might reflect a monitoring process that assesses whether the selected action, which occurs prior to movement onset, is appropriate for the stimulus that is presented. Hence when an infrequent stimulus is presented, as encountered during an oddball task, this leads to a change in response. As a result of this change in the response, there needs to be a 20  monitoring process to assess whether the appropriate response has been chosen. As a result of this monitoring process, larger P3b amplitudes are observed when responding to the infrequently encountered but expected stimulus during an oddball task. While this explanation by Verleger, Jaśkowski and Wascher, (2005) have a viable argument, they fail to explain why variations in P3 still occur when motor responses are not required (i.e. passive listening/watching; Bennington & Polich, 1999; Jeon & Polich, 2001). This suggests that the P3 may occur somewhere between the perceptual process and the corresponding action (if required). This leads us to the explanation by Nieuwenhuis and Cohen (2005), who used the adaptive gain theory to explain the role of the P3 (Aston-Jones & Cohen, 2005). The adaptive gain theory posits that the locus-coeruleus (LC), located in the brainstem exerts an excitatory effect on the brain. This excitatory response follows the assessment of the significance of the stimulus and decision making process by brain areas such as the anterior cingulate cortex and orbitofrontal cortex. If the stimulus is deemed to be motivationally significant (i.e., task relevant), a phasic increase in LC activity occurs, leading to the generation of the P3. The LC projects to various cortical and subcortical structures via norepinephrine release (Samuel & Szabadi, 2008). This NE causes a heightened arousal in the structures, which ultimately facilitates and optimizes the selection and execution of responses to the presented stimuli.  Thus while these target evoked components have been extensively researched, particularly the P3, the explanations behind these processes are still being refined. Recent experiments however, have been investigating the ties between the P3 component with encoding a target or pre-cue (Kourtis et al., 2012; Krigolson et al., 2012), and the subsequent reaching movement to the target. 21  1.3.3 Variations in the P3 Component and Impending Movement Difficulty While the P3 components mentioned in the previous section have been shown to be modulated by the likelihood that a visual stimulus is presented (i.e., P3b) or the orienting of attention (i.e., P3a) more recent research has looked into how these components are modulated as a function of the level of difficulty of an impending limb movement. In a recent publication, Kourtis et al. (2012) paired electroencephalogram recording methods with a Fitts’ aiming task. In the experiment, individuals were asked to perform reaching movements to targets of various sizes and amplitudes. At the beginning of each trial, participants were first informed of the target they would have to reach to, by way of a visual pre-cue. The pre-cue was composed of an arrow of a given length; for instance if a long arrow was presented individuals would prepare to reach towards the far target, whereas in the case of a short arrow, the close target. Also, target size was varied across blocks of trials; therefore within a block individuals would aim to targets of a certain size, while the amplitude was varied within the block. Here the authors were interested in the components associated with the encoding of pre-cue, notably the N2 and P3 observed over parietal regions of the scalp would vary as a consequence of movement task difficulty. It was suggested by the authors that the variations in the N2 and P3 could be a result of activity from motor areas underlying the preparation of the movement (Kourtis et al., 2012). For instance, if the task was difficult (e.g., long amplitude, small target size) the actor would be less certain that the selected motor plan would successfully achieve the movement outcome, as compared to an easier task difficulty (e.g., short amplitude, large target). Therefore they predicted that the N2/P3 components observed would be larger in 22  tasks that were of greater difficulty. Indeed, Kourtis et al. (2012) found that when there was greater task difficulty, the magnitude of the N2 and P3 that was associated with the encoding of the pre-cue, was larger than when the cue indicated an easier aiming movement (see Figure 1.4). From these findings, the authors suggested the greater N2 amplitude is a consequence of alterations to the motor plan that are specific to the impending movement (Connolly, Andersen & Goodale, 2003). While the variations in P3 amplitudes were thought to reflect context updating processes during the forward modelling process (Polich, 2007; Verleger, Jaśkowski & Wascher, 2005). Therefore, these findings support the concept that individuals plan their movements at the time of being informed of the difficulty of the impending movement. Furthermore these components reflect the planning and predictive processes occurring in the motor and sensorimotor areas of the brain vary based upon the perceived difficulty of the movement.  Figure 1.4. From Kourtis et al. (2012) and and reproduced with permission. The figure shows grand averaged VEP waveforms derived from pooled electrode sites Pz, P1, P2, POz, PO3 and PO4. Notice the amplitude of the P3b was greater when individuals were 23  observing the cue that indicated greater task difficulty (i.e., thick black line) as compared to the cue indicating lesser difficulty (i.e., dashed line).  1.3.4 Movement-Related Cortical Potentials There are several large motor-related cortical potentials that occur shortly before and concurrent with an actual limb movement (Brunia, 1987, Kirsch & Henninghausen, 2010). The first potential that occurs in preparation for a movement is the readiness potentials (RP). This includes the Bereitschaftspotential (BP), which is a slow shifting negative deflection that occurs over fronto-central areas of the scalp at approximately 1-2 seconds prior to a self-generated voluntary movement (Kornhuber & Deeke, 1965). The BP is subdivided into two phases, an early BP that occurs over fronto-central regions of the scalp at approximately 2 seconds prior to the initiation of the limb movement. This early BP is thought to be generated by the supplementary motor area (SMA; Shibasaki & Hallet, 2006). Following this early BP, and increase in negative activity over areas contralateral to the reaching limb that occurs approximately 400 ms prior to movement initiation. This late BP is thought to be generated by primary motor areas of the brain (Shibasaki et al., 1980; Shibasaki et al. (1981). For movements that are initiated in response to an imperative stimulus, a similar slow shifting negative deflection is also observed over frontal areas of the scalp shortly before the stimulus, called the contingent negativity variation (CNV). Both these readiness potentials (BP and CNV) are thought to reflect motor preparation processes, as imaging studies have shown that the supplementary motor area, cingulate cortex and the thalamus are active during these periods where the RP occurs, prior to a movement (Nakai, 2004; Rohrbaugh & Gaillard, 1983; Yazawa et al., 2000). Furthermore, if the 24  individual is informed about the limb that will be used to perform the action as well as the direction of the movement, then the RP becomes a lateralized readiness potential (LRP), as the RP is observed at sites contralateral to the acting hand (Kutas & Donchin, 1980). Following the readiness potentials is a positive potential called the premotion positivity (PMP) and finally a negative potential called the motor potential (MP) occurring 10-50 ms prior to the onset of muscle activity. The MP is believed to reflect the descending discharge of activity from the motor areas of the brain (Brunia, 1987; Kirsch & Hennighausen, 2010; See Figure 1.5). Once the movement is underway, several additional components have been observed. These components are the P2, N4 and P3 components (Brunia, 1987; Kirsch & Hennighausen, 2010). While the function of the P2 is relatively unknown, it has been suggested that the N4 and the P3 relates to a second descending discharge of activity from the motor cortex, which contributes to decelerative control (Brunia, 1987; Kirsch, Hennighausen & Rosler, 2010; Kirsch & Hennighausen, 2010).  Figure 1.5. From Kirsch, Hennighausen & Rosler, 2010; and reproduced with permission. The figure shows a guided limb movement to a physical stop on the left. This was followed by a replication of the limb movement without the stop on the right, where individuals 25  attempted to replicate the previous movement. Labelled on the graph to the left are the peaks associated with the limb movements.   To better understand the brain processes that underlie goal-directed manual reaches, Kirsch, Hennighausen and Rosler (2010) conducted a study in which EEG was used to observe the brain activity (i.e., motor-related cortical potentials MP and N4) elicited by participants when they performed a one-dimensional motor replication task.  The components they were interested in examining were MP, which peaked shortly following initiation of the reaching movement; and N4, which peaked before peak deceleration. For the task, participants first moved a linear tracker to a mechanical stop and returned to the start position. This was followed shortly by a replication movement, where participants attempted to replicate the previous movement but without the stop. The authors found that these two negative components in the EEG waveform that were observed over fronto-central regions of the brain (i.e., electrode FC1). These components, the MP and N4 peaked shortly before peak acceleration and before the decelerative phase of the limb movement, respectively. Furthermore, the latter of the two components, the N4, scaled with the amplitude of the movement, as greater magnitudes of the N4 were observed for movements of greater amplitude (Figure 1.6).  26   Figure 1.6. From Kirsch, Hennighausen & Rösler (2010) and reproduced with permission. The above figure depicts the displacement (p) velocity (v) and acceleration (a) profiles of reaching to different movement amplitudes. The bottom figure shows the motor related cortical potential related to the above conditions. Notice the amplitude of the second component (N4), scales with the amplitude of the target, with reaches of longer amplitudes yielded larger N4 amplitudes.  From these findings the authors suggested the MP and N4 components reflected the activity in the motor areas of the brain as the motor command is being sent down the pyramidal tracts and down to the acting limb. Furthermore, findings support the idea that rapid goal-directed aiming movements are marked by two phases of cortical excitability (Sergio & Kalaska, 1998), one which helps to initiate the movement while the other is to help brake the movement as it nears the movement end. 27  1.3.5 Motor-Related Cortical Potentials and with Error Detection First, the early research into the neural correlates of error detection were examined, notably the error-related negativity (ERN) as observed by Falkenstein, Hohnsbein and Blanke (1991). Also, the correlates of error detection and corrective feedback use during goal-directed reaching movements were studied. When we perform an upper limb movement, errors inevitably occur due to the variability of the initial submovement towards the target (Schmidt, Zelaznik, Hawkins, Frank, Quinn, 1979; Meyer et al., 1988). As a result, error detection and feedback mechanisms are required to identify these errors and allow us to take corrective action, either through online control mechanisms that correct the limb trajectory in real-time or through alterations in the motor plan for a follow-up attempt (Krigolson & Holroyd, 2007). The error-related negativity, refers to a negative potential over fronto-central regions of the brain that occurs 80-150 ms following incorrect motor response errors (Falkenstein et al., 1991; Gehring, Goss, Coles, Meyer, & Donchin, 1993). Typical tasks used to elicit response errors and the consequent ERN, are the Stroop task, task switching protocols, flanker tasks, where frequent errors are encountered. Also notable is the feedback related negativity (FRN), which instead of being time-locked to the response (i.e., ERN), occurs 250-300 ms following the negative feedback is presented to the individual (Miltner, Braun & Coles, 1997). As compared with ERN, which has been suggested to be an evaluation of the efference copy of the movement, occurring with speeded responses (i.e., before feedback can be used), the FRN occurs after external feedback processed. Tasks used to elicit ERNs involve external feedback such as monetary losses (Zottoli & Grose-Fifer, 2007) or reaching tasks involving perturbations to the limb movement (Torrecillos et al., 2012) or perturbations to the target (Krigolson & Holroyd, 2007; Krigolson et al., 2008). Although the 28  method of inducing ERNs and FRNs are different, the brain structures underlying these potentials are believed to be one in the same: the dorsal anterior cingulate cortex (ACC). Support for the ACC’s role in the ERN/FRN has been supported by research using fMRI (Ito et al.,2003), ERN/source localization (Dehaene, Posner & Tucker, 1994), and lesions studies of the prefrontal cortex (Stemmer et al., 2004) It is believed that when an error occurs, the ACC detects the occurrence of an error and acts as a beacon, signaling to other brain areas such as the basal ganglia, motor areas (e.g., SMA, PFC) to make alterations to the motor plan (Krigolson & Holroyd, 2007). Indeed, studies observing patients with sustained lesions to the ACC revealed impairments in one’s the ability to correct errors, as well as a decrease in amplitude of the ERN when compared with age matched controls (Turken & Swick, 1999). However, given that damage to the ACC did not eliminate the ERN entirely, this has led investigators to suggest that the ACC may not be the sole generator of the ERN (Paus, 2001). As for the FRN/ERN’s significance as a neural process, several theories have been put forth. First, the error detection theory posits that the ACC is the site where comparisons between the desired outcome with an efference copy of the movement that is being executed. In the case of the FRN, this comparison is made between the desired outcome and actual outcome, stemming from long loop feedback processes As for the significant of FRN/ERN, authors have suggested these potentials reflect the magnitude of error that has occurred, as research has found that the amplitude of the ERN/FRN often scaled with the magnitude of error that was encountered (Goyer, 2008; Torrecillos et al., 2012). Others have suggested that the ERN is simply a binary signal, indicating that an error has occurred or not (Hajcak et al., 2006)  The second theory put forth is the reinforcement learning theory, which suggests that the ACC uses reward and error signals to identify and select appropriate motor responses. Namely, 29  the error and reward signals are projected from the mesencephalic dopaminergic system (i.e., basal ganglia) onto regions of the ACC, which is responsible for resolving the conflict. Generally, the idea is that individuals attempt to increase the amount of reward (i.e., correct responses), while also attempting to decrease the instances of negative reinforcement (i.e., error/ERN). The reinforcement learning theory posits the basal ganglia monitors the desired and actual sensory feedback and assesses whether an error has occurred. This comparison of activity is assessed as being successful or unsuccessful, results in a positive or negative reinforcement signal and is conveyed to the ACC. This negative reinforcement signal occurs through a phasic decrease in the dopaminergic activity from the basal ganglia to the ACC (Holroyd & Coles, 2002). When an unwanted movement outcome (i.e., error) is detected, this negative reinforcement signal is projected from the basal ganglia to the ACC. The ACC evaluates and resolves the error by reallocating control over various motor controllers in order to modify or choose a different and more appropriate motor response. This reallocation of control is what they believe to manifest as the ERN, with greater ERNs observed when a larger response is required in order to mitigate the error that is encountered. As a result, the newly chosen motor response would be more likely lead to the desired outcome (i.e., reward).  1.3.6 fERN and Reaching Movements Recent work involving fERN and goal-directed manual aiming has implicated the second negative component in the motor-related cortical potentials (i.e., the N4 peak) with error detection during a reaching movement. In the study, Torrecillos, Albouy, Brochier and Malfait (2014) had participants perform a shooting task, where a cursor was guided through a perpendicular force field to a target circle at peak velocity. Once the cursor crossed the target 30  circle it would continue on with its trajectory without guidance from the participant. In the experiment, most trials were performed with a perpendicular force field that participants were accustomed to. On a small percentage of the trials however, the force field was reduced or entirely removed, resulting in aiming errors. The authors wanted to examine the neural implications of encountering such error. The results of the experiment showed that the N4 component occurred when there was an error in the reaching movement. Furthermore, the amplitude of N4 component scaled with the amount of error encountered (Figure 1.7). Finally, as this component was expressed over fronto-central electrodes, the authors concluded that the activity was a result of activation of the ACC, a consequence of the error detection process. Therefore, in comparison to early experiments that examined ERNs using simple key presses tasks (Falkenstein, Hohnsbein, & Blanke, 1991; Gehring, Goss, Coles, Meyer, & Donchin, 1993) this recent work has shown that FRNs are also expressed during goal-directed reaching tasks.    Figure 1.7. Motor related cortical potentials associated with the shooting task. Note the modulation of the second negative component occurring approximately 250-300ms 31  following movement onset. Figure from Torrecillos et al. (2014), reproduced with permission.  1.3.7 Motor-Related Cortical Potentials and Visual Feedback Use In addition to neural correlates associated with error detection, attempts have been made to try to identify neural correlates associated with visual feedback use during goal-directed aiming. Krigolson and Holroyd (2007) examined neural correlates associated with the use of visual feedback using a target jump paradigm and EEG. The participants were engaged in an aiming task to a target with perturbations occurring to the target on a small percentage of trials (i.e., target jump upon movement initiation). The authors found that on trials where there was a target jump, a negative peak (N100) and a positive peak (P300) occurred shortly after the jump, over parietal-occipital regions of the scalp (Figure 1.8). These peaks were not present on trials where the target did not jump. Furthermore, when comparing the latency of the N100 and P300 peaks with the analysis of limb trajectory amendments, the authors found that the corrective submovements to the limb trajectory occurred shortly after N100 and P300 had ceased. The latencies of the events (i.e., P3 onset before initiation online corrections) led the authors to suggest that the N100 and P300 reflect the retrieval and use of visual feedback, respectively. However, in their follow-up study (Krigolson, Holroyd, van Gyn & Heath, 2008), the authors found that the peak latency of the P300 did not always occur before the onset of online corrections. This finding led the authors to refine their stance on the significance of the P300, suggesting that the P300 reflects the phasic increase in brain activity stemming from the release of norepinephrine from the LC-NE system (Nieuwenhuis, 2005). The release of NE would help 32  upregulate the motor system and help facilitate the updating of the internal model and motor plan for future reaches. While a frontal component may signal error detection (FRN, Krigolson & Holroyd, 2007; Torrecillos et al., 2014), as indicated by the previous section, there is likely corresponding activity in parietal regions that reflects the use of visual feedback (Krigolson & Holroyd, 2007). Therefore activity in both frontal (i.e., MP and N4 peaks) and parietal regions (i.e., N100, P300) will be of interest to us in the studies that will be presented in this thesis.  Figure 1.8. Taken from Krigolson & Holroyd (2007). This figure depicts the grand averaged waveform generated over electrode Pz, over the parietal region of the scalp during a reaching movement. Notably the large dashed control condition waveform depicts the parietal activity observed during reaching movement (i.e., guiding a cursor) without a target perturbation. The correctable and uncorrectable conditions are trials where the 33  target perturbation occurs with and without the ability to correct the trajectory of the cursor, respectively.   1.3.8 Hierarchical Model of Error Detection From the pattern of findings cited above, Krigolson & Holroyd (2006, 2007) forwarded the Hierarchical model of error processing. In this model, the authors suggest that there are two types of errors that one encounters when performing reaching movements. These different types of errors consequently are detected and contended with using different error detection systems. First are low level errors that occur from changes in the environment such as a change in target location, or errors in movement trajectory when performing a reach. To amend these low level errors, individuals rely on the visuomotor system to update their responses in real-time through the use of forward modelling and visual feedback from the limb and target. These corrections are largely mediated by the posterior parietal cortex which integrates visual and proprioceptive information in real-time (Desmurget & Grafton, 2000). However, if these low level errors fail to be amended, they become high-level errors. Such high-level errors are dealt with by a fronto-central system involving the anterior cingulate cortex, which is reflected by the fERN when errors of movement outcome have occurred (Holroyd & Coles, 2002; Holroyd, 2004). To deal with the errors of movement outcome, structures in the fronto-central cortex detect that an error has occurred and relays to brain structures involved with the motor planning to alter the motor plan for subsequent attempts. This model has been backed by an experiment conducted by Krigolson et al., (2007). In this experiment participants moved a cursor, which was controlled by a joystick, to a visual target. On a subset of the trials, the target jumped, eliciting N100 and P300 34  peaks, which required participants to move the cursor to the new location. This target jump, which would be classified as a low level error, required participants to modify their limb trajectory so that the cursor would move to the new target location. On some trials however, the participants could not modify the position of the cursor. The results showed that when the target jumped but the participants had control over the cursor, no ERN was observed. However, when participants did not have control over the cursor, which limited their ability to correct these low level errors, an ERN was observed. In all, the Hierarchical model of error detection proposed by Krigolson et al., (2007), unifies the different ways in which the motor system contends with movement error, and allows for the achievement of the desired movement outcome. 1.3.9 Modulation of Cortical Potentials during Memory-Guided Reaches While the neural correlates observed during the visual control of an aiming movement have been covered above, in this section literature pertaining to the neural correlates of memory-guided reaching and how the observed motor cortical potentials change when visual feedback of the target is removed, is presented. Recent work by Krigolson and colleagues (2012), sought to examine how individuals use memory to guide limb movements. Notably they were interested in testing two hypotheses (i.e., sensory vs. motor) of how memory-guided reaches are planned and executed. The first hypothesis (i.e., sensory) posits that when individuals observe a visual target prior to a memory-guided reach, they encode and store the representation of the target in visual working memory up until the point at which the movement will be generated. Right before the movement is initiated, individuals will use the representation of the target to generate a motor plan that will be subsequently executed. On the other hand, the second hypothesis (i.e., motor) 35  posits individuals generate a motor plan at the time of observing a target in space and store the motor plan in memory until the time of movement initiation. Therefore instead of simply observing and holding the coordinates of the target in memory, as suggested by the sensory hypothesis, individuals generate an appropriate motor plan that will reach the specified target upon first seeing it.   To test these hypotheses the authors used EEG and the event-related potential technique (ERP) to observe the differences in brain activity at two different time points prior to, and during a reaching movement. The first event was when individuals first observed the target that they were going to reach to (i.e., target encoding), and the second event was when individuals performed the reaching movements to the target under visually-guided vs. memory-guided reaches (i.e., movement execution). In the experimental protocol, participants were provided with a preview of the target for 2 s. For visually guided trials, an imperative auditory tone occurred at the end of the preview period, which informed participants to perform their aiming movement towards the target. For memory-guided trials, the target was occluded after the preview with the auditory tone occurring after a 1 s delay. Also important to note, in both visually-guided and memory-guided reaches, participants had positional information of their hand. The ERP epoch of interest was the preview period of the target, which allowed the experimenters to examine the neural activity (i.e., target evoked potentials) associated with the encoding of the target. Presumably, if individuals were using a motor-based memory representation to execute their memory-guided reach, the authors expected participants to generate a motor plan at the time of target encoding for the subsequent movement that will not have visual feedback (Ghafouri, McIllroy & Maki, 2004; Krigolson et al., 2012). That is, the authors suggested that activity over motor areas associated with the generation of the motor plan 36  would occur during target encoding, which would alter the target-evoked potentials for memory-guided vs. visually-guided reaches. On the other hand, if individuals were using a sensory memory representation (i.e., Elliott & Madalena, 1987) the authors predicted there would be no differences in the encoding of the target between memory and visually-guided movements. The results of the experiment coincided with the motor hypothesis, as the ERPs associated with the encoding of the target for memory-guided movements exhibited reduced potentials as compared to visually-guided reaches (Figure 1.9). This is not to say that the motor plan generated at the time of encoding for a memory-guided reach is different from the motor plan generated prior to a visually-guided reach. The authors suggest that motor planning occurs earlier (i.e., upon seeing the target) when preparing for a memory-guided reach, while the motor planning for visually-guided reaches do not have to occur until shortly before the execution of the reach.  Figure 1.9. Taken with permission from Krigolson et al. (2012). Grand averaged ERP waveforms locked to target presentation for the proximal target. This graph illustrates that the memory guided waveform is markedly suppressed as compared to the visually guided waveform. The authors suggest that the suppression of the memory-guided trace may have 37  to do with the involvement of motor planning processes when individuals are encoding the target, which is supports the motor hypothesis.     Furthermore, the authors examined the movement-related cortical potentials associated with the act of reaching under visually-guided vs. memory-guided reaching, where they found differences. Namely, for the second component (N4) occurring approximately 350 ms after movement start, memory-guided reaches had significantly smaller amplitudes than visually-guided reaches. The authors suggested the lower amplitudes could have to do with lower forces used to control the limb as it homed in onto the target (Figure 1.10). Notably there was a significant correlation between N4 amplitude and magnitude of peak deceleration, as trials with vision had higher amplitudes and deceleration as compared memory-guided reaches with lower amplitudes. 38     Figure 1.10. Taken with permission from Krigolson et al., (2012). ERP average waveform time locked to movement start. Notice the second peak observed at approximately 300 ms post movement initiation is lower in magnitude than in the visually guided condition. This coincided with lower deceleration values in the memory guided reach condition.  While the assertion by Krigolson et al., (2012) that the N4 component is simply tied to limb force is at odds with those made by Torrecillos et al. (2014), who speculated that the N4 component has to do with error processing, it can be suggested that both ideas are complementary. Notably, in the Krigolson et al. (2012) experiment, individuals still had vision of the position of their limb in the memory-guided condition. Therefore, the N4 peak should still exist as the participants would still be able to detect errors in the limb trajectory and make online corrections. The lowering of the N4 peak however may have to do with the lower forces or lower amount of corrections that are occurring to the limb in the memory-guided vs. the visually-guided conditions. While it is possible for us to continue to speculate what these waveform 39  components mean, the purpose of this thesis is to further address the meanings of these components with respect to movement.  In summary, the current literature examining the neural correlates associated with human movement has been covered. As seen, there remains uncertainty in the underlying processes that these cortical potentials reflect. Hence, two overarching questions have been used as a guide for the devising of three studies that examine the neural processes and the correlates associated with reaching movements. 1.4 Overarching Questions 1.4.1 Target Encoding Does the correlate P3 observed with target encoding vary depending on the reaching conditions towards said target and what cognitive process does this correlate reflect? As mentioned above in the introduction, the P3 component has traditionally been elicited through oddball paradigms (Squires, Squires & Hillyard, 1975). While the main hypothesis used to explain the modulation of the P3 was thought to reflect context updating processes, as the memory representation of the environment is being updated (Donchin & Coles, 1988), more recent explanations have implicated the P3 as being involved in the assessment of action, occurring at the time of the onset of the reaching movement (Verleger, Jaśkowski & Wascher, 2005). Indeed a modulation of the cognitive P3 has also been observed when individuals were encoding an endogenous cue (Kourtis et al., 2012), as well as an exogenous target (Krigolson et al., 2012) prior to goal-directed reaches. From these results, it was suggested from Kourtis et al. (2012) that the P3, which scaled with movement difficulty, reflected forward modeling 40  processes, with greater difficulty resulting in greater uncertainty of the movement outcome. An alternate hypothesis put forth by Krigolson et al. (2012) suggested the modulation of the P3 in their experiment had to do with the engagement of motor planning processes when individuals were encoding in preparation for memory-guided vs. visually-guided reaches. This thesis investigates the processes underlying the modulation of the P3 when individuals encode the target prior to a reaches under varying task difficulty and visual feedback availability. 1.4.2 Movement Execution What are the neural correlates associated with movement execution and what neural processes do they reflect?  Studies examining manual reaching amplitude using EEG have observed two cortical components that are observed over fronto-central areas of the scalp. The first peak (i.e., MP) is observed shortly after movement initiation and a second peak (i.e., N4) is observed towards the latter portion of the movement, (Brunia, 1987; Kirsch, Hennighausen & Rosler, 2010).  The first hypothesis put forth is that these two components reflect two phases of motor activity in the primary cortex that are necessary to propel the limb toward the target and to control the limb as it nears the movement endpoint (Kirsch, Hennighausen & Rosler, 2010; Sergio & Kalaska, 1998). Indeed Kirsch and Hennighausen (2010) showed the peak amplitudes of MP and N4 scaled with peak acceleration and deceleration of the limb when reaching to different movement amplitudes. That is, when executing long amplitude reaches, more force is required to propel and stop the limb as it reaches movement end, as compared to short amplitude reaches. The greater force output is also reflected by a larger magnitude of both the MP and N4 peaks, as compared to movements of smaller amplitudes. In turn reaches of shorter movement 41  amplitudes require less force to move to the desired movement outcome. This hypothesis was corroborated by comparisons of visually-guided and memory-guided reaches (Krigolson et al, 2012), as memory-guided reaches were performed with lower forces, which was reflected in lower MP and N4 amplitudes, as compared to visually-guided reaches.  The second hypothesis put forth suggests that the second elicited peak during a reaching movement (i.e., N4) may have to do with error processing stemming from the fronto-central cortex. As found previously by Torrecillos et al., (2014) the amplitude of the second peak coincided with the amount of error encountered during a shooting task. Overall, this could be explained by the Hierarchical error detection model (Krigolson et al., 2007), which proposes that there are two error detection systems that are used to detect and amend error. That is, when low-level errors are encountered (i.e. in movement trajectory or target location), the visual-motor system mediated by the PPC corrects for these errors through forward modeling and online visual feedback. However, when these errors fail to be corrected in real-time they are mediated by the fronto-central system, which manifests as a larger N4 peak during the movement as seen in Torrecillos et al., (2014). Therefore, the MP and the N4 cannot be totally explained as just being two phases of motor output (cf. Kirsch & Hennighausen, 2010), but rather the second peak may also reflect activity of the fronto-central error detection system. In all, the purpose of this thesis was to further examine what underlying processes these components reflect, through examinations of movement-related potentials under multiple variations in reaching conditions. 42  1.5 Experimental Approaches 1.5.1 Study 1: The Neural Correlates of Visually and Memory-Guided Reaches Under Varying Task Difficulty In this first study, the neural correlates associated with reaching movements to targets of varying task difficulty (i.e., Fitts’ Law) were examined. The research question was whether the components of task difficulty (i.e., target size and amplitude) and visual feedback availability during the reaching movement (i.e., target or no target) affect cortical motor output. This aim of this study was to generate waveforms relating to the encoding of a target, as well as the execution and control of the reaching movement to said target. To do this, an experimental protocol was devised to observe variations in brain activity associated with the encoding and execution of movements towards targets differing in target amplitude, size and visual feedback availability. Previous work conducted by Kourtis et al. (2012) observed a modulation of the P3 amplitude for electrodes over parietal regions of the scalp when participants were observing a central covert pre-cue that indicated the upcoming task requirements (i.e., arrow of varying size). Therefore in this thesis, attempts were made to examine whether these effects would also occur with an overt cue. Thus a preview of the actual target was presented to the participant before they were required to reach to it. Following the preview was a brief delay period of 2 s, and imperative tone where participants performed a reaching movement to the previewed target. Event-related recordings of the actual reaching movement enabled us to examine whether the movement-related potentials were modulated by the components of task difficulty (i.e., target size and target amplitude). Previous research examining movement-related cortical potentials (Kirsch, Hennighausen, Rosler, 2010) suggested that two negative components observed over 43  fronto-central regions of the scalp following movement initiation (i.e., MP and N4), are associated with distance-specific scaling of motor command. Modulations of these components are believed to reflect the motor output commands to the acting limb and vary depending on the forces required to achieve different movement amplitudes (Kirsch and Hennighausen, 2010). In this study attempts will be made on the one hand, to try and replicate previous findings from other studies (i.e., P3 and target amplitude; MP and N4 with PA and PD). In addition, other factors that have not been explored yet (i.e. target size, and visual feedback availability) were examined. 1.5.2 Study 2: The Neural Correlates of Reaching Under Varying Delay and Visual feedback The goal of this second study was to examine the brain potentials associated with goal-directed reaching under conditions of varied visual feedback and movement delay conditions. The protocol used in this experiment was similar to the one used by Krigolson et al. (2012); however, in addition to using a brief delay period between the preview of the target and the execution toward it, a longer delay condition was incorporated in order to examine the behavioural and electrophysiological consequences of memory representation decay. Therefore, in this experiment delay periods of 2 and 5 s were used. Based upon the previous findings of Krigolson et al. (2012), a modulation of the brain potentials associated with encoding the target was expected. Notably a suppression of the brain potential (i.e., P3) was expected over fronto-central electrodes of the scalp when individuals were preparing for a movement without vision.  Also, during the execution of the movement, a suppression of the movement potentials (i.e., MP and N4) over similar scalp regions is expected for reaches performed without vision compared to 44  reaches performed with the aid of visual feedback. This study will provide an indication of the strategies individuals adopted when performing reaching movements under full vision and under visual occlusion of the target.  1.5.3 Study 3: The Neural Correlates of Real-Time Vision of the Hand and Target During a Goal-Directed Reaching Movement In the third study, the goal was to examine the neural correlates associated with visual feedback availability of the hand during goal-directed reaching. Namely, there was a desire to examine whether differences in brain activity were observed when planning and executing movements with and without vision of the hand and target. Both behavioural and neurophysiological evidence (i.e., Action/Perception, Heath, 2005; Milner & Goodale, 1995) has shown that when vision of the hand is available vs. when it is not, there is a difference brain structures involved with the planning and execution of the reaching movement. Indeed, if different visual pathways are used in the planning and execution of movements when vision of the hand is varied, then differences in brain activity should be observed between hand visible and hand occluded reaches. In studies 1 and 2, removal of visual feedback was largely done through removal of vision of the target, while vision of the hand and stylus remained available. However, considering that a large part of the online feedback process for a reaching movement involves tracking and correcting the position of the limb (Desmurget & Grafton, 2003), manipulating vision of the reaching limb may have larger implications on the motor-related cortical activity than the simple removal of vision of the target only. Notably, it has been demonstrated that if vision of the target was removed but vision of the hand remained, individuals exhibited similar degrees of limb trajectory amendments 45  as conditions where both vision the hand and the target were available (Heath, 2005). As such it was suggested that having vision of the reaching limb allows for real-time visual processing by way of dorsal stream processes, which in turn allows for the amendments to occur to the hand leading to a more precise movement (Westwood, Heath & Roy, 2003; Milner & Goodale, 1992). As such, the components MP and N4 associated with the execution of the movement under conditions where the limb is visible should yield greater activity over the parietal and occipital centres that are contralateral to the reaching limb (Krigolson et al., 2008). When vision of the limb is not available during the reaching movement, the level of activity in these areas were not expected to differ. Furthermore, the use of EEG in this study allowed for the observation of the time course of the neural events associated with visual feedback retrieval and use. In all, this study provided a better understanding of how visual feedback is retrieved and used in a closed-loop system.    46  2 Study 1: The Neural Correlates of Visually and Memory-Guided Reaches Under Varying Task Difficulty 2.1 Introduction When reaching toward objects in the environment, some tasks are more difficult to perform than others. As such, we often find inserting a key into a lock to be a much slower and engaging process than simply adding sugar to your coffee in the morning. The difficulty of a task entails the distance the limb must traverse in order to reach the target, as well as the size of the target itself. These two parameters (i.e., movement amplitude and target size) contribute to the index of difficulty of a movement, as first proposed by Fitts (1954). To accommodate for variations in task difficulty, the central nervous system may dedicate more time for the planning of the movement and online control, during the execution of the reaching movement. Specifically, reaction times, as defined by the time it takes for a movement to be initiated, are often longer when the task difficulty is high (Henry & Rogers, 1960; Hick, 1952; Klapp, 1975). This increase in reaction time reflects a greater reliance on movement planning processes prior to undertaking the task. In addition to reaction time, movement time, defined as the time from when the movement is initiated until the movement is terminated, is also longer for tasks of greater difficulty (Fitts, 1954). The longer movement time for tasks of greater difficulty, such as threading a needle, results from fine-tuning of the limb trajectory through the use of visual feedback, as it homes in on the target (Fitts, 1954; Meyer et al., 1988; Woodworth, 1899; Zelaznik, Hawkins & Kisselburgh, 1983). While behavioural studies have provided insight into the processes underlying human movement, more recent studies using electroencephalographic techniques have allowed us to further investigate the intricacies of these functions.  47  Recent work by Kourtis et al. (2012) examined the movement planning processes prior to the execution of reaching movements in a Fitts’ task using electroencephalography (EEG). To probe the movement planning process, participants were presented with two sources of information about the impending movement. The first was an informative visual pre-cue at the start of each trial, which informed participants of the amplitude of the impending movement. The pre-cue consisted of an arrow of a given length, which specified the amplitude of the upcoming movement. A long arrow indicating to reach to a farther target and a short arrow specifying to reach to a closer target, relative to a home position. The second source of information was the size of the target that participants were aiming to, which remained the same within each block of trials but differed between blocks. This allowed participants to be certain of the target size they were aiming to on each trial. The authors were interested in examining whether the movement planning process involves an assessment of the index of difficulty of the movement from these two sources of information. To do this the authors examined the event-related potentials evoked by the onset of the pre-cue, namely the N2 and P3b components that were observed over the parietal regions of the scalp. Some support for such an approach was available from previous studies, which have also implicated the parietal N2 component as being associated with activity in the superior parietal lobule and the elaboration of a motor plan (Naranjo, Brovelli, Longo, Budai, Kristeva, & Battaglini, 2007), while the P3b component can reflect the context updating of the internal model associated with motor planning process (Donchin & Coles, 1988; Polich, 2007; Verleger, Jaśkowski & Wascher, 2005). It was hypothesized that if the N2 and P3b components were modulated by index of difficulty, it would support the idea that the motor system represents motor plans and forward predictions of the plan based on the index of difficulty (i.e., movement amplitude and target size) of the impending movement. The results 48  showed that when the task involved reaches of greater difficulty (e.g., long arrow & small target), the amplitudes of the N2 and P3 amplitudes were greater than reaches of lesser difficulty (e.g., short arrow & large target). From these findings the authors suggested the scaling of the N2 reflected alterations in the motor plan to correspond with the movement parameters. The scaling of the P3b component was thought to reflect forward modeling mechanisms that predict whether the motor commands being executed will achieve the desired movement outcome. Specifically, movements of greater difficulty would have greater uncertainty when it comes to judging whether the desired outcome will be achieved. As a result, a larger amount of resources may be dedicated to predicting and possibly correcting errors using forward modeling, which can explain why larger P3 amplitudes were observed. In contrast, for easier tasks, the level of uncertainty would be less and lead to a smaller P3 amplitude. In all, the authors concluded that movement planning processes are subject to the influence of perceptual encoding of an informative pre-cue prior to movement execution. While the use of an endogenous pre-cue (i.e., arrows of varying length) requires top down interpretation and categorization the meaning of the cue, it remains to be seen whether presenting the actual target to the actor prior to a reaching movement would also elicit the same effects.  Furthermore, Krigolson et al., (2012) found that the P3 was subject to modulation by the availability of visual feedback available of a subsequent reach. Specifically, the P3 evoked by a target was larger when individuals were encoding the target prior to a reach without vision of the target versus when it was available. This led authors to suggest, using a motor hypothesis (Ghafouri, McIllroy, Maki, 2004; Krigolson et al, 2012) that the larger P3 was due to the generation of a motor plan when individuals were encoding the target in preparation for a reach without vision of the target. The motor plan is then stored in memory until the reach without 49  vision of the target is executed. Whereas, for reaches with vision of the target, the generation of the motor plan does not need to be generated until shortly before initiating the reach to the target, as target vision is always available. In combination with the work cited above (i.e., Kourtis et al., 2012), it appears that the modulation of P3 in these manipulations reflect motor planning and forward modeling processes relating to the subsequent reaching movement. Although the motor-related cortical potentials associated with the act of reaching out towards the target were not a focus of the paper by Kourtis et al. (2012), there is also a body of research that has found a modulation of brain activity when executing reaches to targets of different movement amplitudes. Kirsch and Hennighausen (2010) had participants briefly preview a visual target, followed by a delay period where vision of the target was removed. Following the delay period participants executed reaching movements towards the location where the target was presented during the target preview. Of interest to the authors were the neural components associated with the execution of the movement, notably the motor MP and N4 components over fronto-central brain areas (i.e., electrode FC1). These two motor components were examined in previous studies (Kirsch & Hennighausen, 2010; Kirsch, Hennighausen & Rosler, 2010) and were implicated as correlates associated with activity in primary motor cortex. The first component, which was identified as the motor MP (Brunia, 1987), peaked shortly before the onset of the reaching movement. The motor MP was believed to reflect the motor output command from the primary motor cortex, which signals to the agonist muscles of the limb to initiate the reaching movement. The second component, identified as the motor N4 (Brunia, 1987), peaked shortly before peak deceleration of the limb trajectory. The motor N4 component was thought to reflect a second command by the primary motor cortex, signaling the antagonist muscles of the limb to slow down the limb as it homes in on the intended 50  target. For brevity purposes, both the motor MP and motor N4 components will be identified as MP and N4 although such terms can also be associated with non-motor components. Kirsch et al. (2010) were interested in testing if the amplitude of these motor components would scale with the amplitude of the reaching task. Supposedly, for reaches of greater amplitude, there should be greater force requirements for both the initiation of the movement as well as the halting of the limb at movement end. Hence, a scaling of the two components would coincide with the force demands specific to the movement amplitude that was to be attained. The results of the study revealed that the MP and N4 components indeed scaled with the amplitude of the reaching movements. Therefore, these results supported the idea that the MP and N4 modulations reflect two phases of motor activity in the primary cortex that are necessary to propel the limb toward the target and to control the limb as it nears the movement endpoint, respectively (Sergio & Kalaska, 1998). While the work by Kirsh and Hennighausen (2010) has provided a better understanding of the neural processes underlying reaching to different distances, questions remain as to whether other facets of the reaching environment also affect the way these reaches are executed. For instance, would target the size and visual feedback availability during the movement affect the way motor-related cortical potentials are expressed? Notably when reaching to different target sizes, the degree of control required in an easy task with large targets is significantly different than when compared to reaching to small targets (Fitts, 1954). With regards motor related cortical potentials and visual feedback availability during a reaching movement, Krigolson et al., (2012) found that the amplitudes of components MP and N4 were significantly smaller than reaches performed with vision of the target. In addition behavioural measures such as peak acceleration (PA) and peak deceleration (PD) were 51  significantly smaller when reaches were performed without versus with vision of the target. This led the authors to suggest that when vision was unavailable, individuals adopted a conservative reaching execution strategy which is not unlike results previously reported in behavioural experiments (e.g., Elliott, Hansen, Mendoza & Tremblay, 2004). Therefore, it was of interested to see whether there would be differences in motor related cortical potentials when reaches to targets of varying task difficulty were performed with and without visual feedback availability of the target. In the current study, the neural processes involved with planning and executing reaching movements to targets under varying task difficulty as well as visual feedback availability were examined. Specifically, of interest was the P3 evoked by the onset of the target prior to the reaching movement. That is, would the P3 vary based on task difficulty of the target and/or visual feedback availability of the target on the subsequent movement. Also of interest were the motor related cortical potentials, that is motor MP: relating to the execution of the motor plan, and motor N4, which reflect the late movement execution where online control mechanisms are used to influence the limb trajectory. 2.1.1 Target Encoding The current study examined whether individuals encode targets differently based upon the difficulty of the task and/or the availability of visual feedback of an impending movement. In order to probe the motor planning processes associated with the encoding of a target, participants were provided with a preview of the target prior to a reaching movement. Notably, this approach differs from that of Kourtis et al. (2012), who used an endogenous pre-cue (i.e., an arrow) to inform participants of the target amplitude. While such a method of using the actual target as the 52  pre-cue requires less cognitive processing—as exogenous cues are largely driven by bottom-up processing—the aim of the current study was to test whether the difficulty of the task was represented in a similar way as an endogenous cue during the motor planning process. It was hypothesized that the P3 associated with the encoding of the target pre-cue would be greater when the target presented was difficult (i.e., small target, long movement amplitude) as compared with an easier target (i.e., large target, short movement amplitude). Furthermore, the removal of vision on the subsequent reaching movement should lead to larger P3 amplitudes, as compared to trials where the subsequent reach was performed with vision. This was predicted because when participants did not expect to have vision during the reaching movement, individuals were expected to engage motor planning processes, which are stored in memory until the reaching movement is executed (see Krigolson et al., 2012). 2.1.2 Movement Evoked Potentials With respect to the motor related cortical potentials associated with movement execution, the components MP and N4 elicited over fronto-central areas of the scalp were of interest (see Brunia, van Boxtel & Bocker, 2011; Kirsch and Hennighausen, 2010; Kornhuber & Deeke, 1965; Krigolson et al., 2012;). As the fronto-central part of the scalp encompasses motor structures (i.e., primary motor cortex), this cortical activity may be related to motor processes. For the neural correlates associated with the early execution of the movement (i.e., the MP component), it was hypothesized that this component would be modulated by task difficulty. Specifically, the amplitude of MP was expected to be larger for reaches of longer amplitudes, as this reflects a greater amount of cortical motor output that is required to generate greater forces when reaching further. Also, it was expected that MP to would be smaller for reaches to the 53  smaller than reaches to the larger targets. Lastly, a modulation based on vision was also expected. As previously reported when movements are performed in the absence of vision of the target, there is a tendency for individuals to be more conservative with their reaches (i.e., slower movement velocities), and hence we expected less force and lower amplitudes of MP (Krigolson et al., 2012). For the second motor-related cortical potential (i.e., N4), the current study aimed to examine its relationship to the control of the limb trajectory as it nears movement termination.  It has been suggested that the N4 activity represents a second phase of motor output activity from the motor cortex, which signals the antagonist muscles to brake and clamp the limb as it reaches movement end (Hannaford & Stark, 1985; Kirsch & Hennighausen, 2010). Therefore, it was expected that the N4 amplitude would be larger when moving to a farther vs. a closer target. In addition to this, it was also hypothesized that the N4 component would also increase with target size. This was predicted because, as when individuals are reaching to a larger target, they would be more liberal with their reaching movements, hence exhibiting a more abrupt halting of the limb as it homes in on the target. Also, with the manipulation of vision of the target, it was hypothesized that the N4 component would vary based upon the availability of vision. Specifically, when vision was unavailable, it was predicted that there would be fewer amendments to the limb trajectory as it homed in on the target, thereby decreasing the need for any additional commands from the motor cortex. Therefore, it was expected that reaches without vision would yield lower N4 amplitudes vs. reaches with vision (Heath, 2005; Messier & Kalaska, 1997). 54  2.2 Methods 2.2.1 Participants Fourteen (14) participants (10 male, 4 female; average age: 23.6 ± 3.2 years) were recruited from the University of British Columbia to participate in the study. All participants were self-reported right-handers with normal or corrected-to-normal vision and had no known neurological deficits. Informed consent was provided prior to the experiment and with the sanction of the research ethics board of the University of British Columbia and the 1964 Declaration of Helsinki. 2.2.2 Task Goal-directed manual aiming movements were performed with a stylus on a graphics tablet in the horizontal plane to visual targets of various sizes and movement distances. Over the course of the experiment, continuous EEG was recorded, with events of interest (i.e., target encoding and movement execution) segmented and averaged to obtain event-related potentials (ERPs).  2.2.3 Apparatus Aiming movements were performed on an experimental setup similar to that used by Held and Gottlieb (1958). This apparatus consisted of a 487.7 x 304.8 mm digitizing tablet (Wacom Ltd Tokyo JP, Intuos4, PTK-1240, , with a sampling rate of 197 Hz), and a LCD monitor (DELL U3011, Round Rock, TX, USA; with a refresh rate of 75 Hz), with a half-silvered mirror positioned half way between the level of the tablet and the monitor. Participants rested their chin on a chinrest at the level of the mirror, such that their eyes were approximately 21 cm from the half-silvered mirror and 69 cm from the surface of the tablet. A strip of white LED lights was 55  also attached to the underside of the mirror, which allowed participants to view the position of the stylus and their hand. The stimuli were presented in white on a black background. A 1 cm diameter home position was presented on the left of the visual display with the fixation point presented 28 cm to the right and 3 cm distal to the home position in the orthogonal axis (see Figure 2.1). 2.2.4 Procedure On each trial, participants positioned the tip of the stylus onto the home position and fixated their eyes onto the fixation cross (Figure 2.1). Participants were then shown a preview of the target for 250 ms. Two possible target locations were used, which were 22 cm or 34 cm to the right of the home position, as well as three target sizes of 1, 2.5 or 5 cm. Following the preview, the target was extinguished and a 2 second delay was introduced. Such a manipulation was introduced to ensure that the processes related to target encoding would be taking place much earlier than the processes related to movement preparation. After the delay period, an auditory tone indicated to participants to initiate a reach to the target location as quickly and as accurately as possible by sweeping the stylus across the surface of tablet. For reaches under full vision (FV), the same target would reappear in the same location that it was previewed previously, while for reaches under no vision (NV) the target would not reappear. After executing the reaching movement, participants kept the stylus at the movement endpoint until a low-tone indicated for them to lift off the surface of the tablet and await the tone for the next trial. Visual feedback availability was varied across blocks of 60 trials, while target size and amplitude was randomly varied from trial to trial. The experimental design was a 2 vision by 3 target size by 2 target amplitude design with 60 trials per conditions resulting in 720 trials in total.  56  2.2.5 Data Reduction and Analysis The stylus displacement data were analyzed offline in MATLAB (The MathWorks Inc., Natick, MA). Each trial was first low-pass filtered (cut-off: 10 Hz) with a 2nd order dual-pass Butterworth filter. These data were then differentiated to obtain velocity profiles for each trial. Movement start was defined as when the resultant velocity exceeded 50 mm/s for 100 ms, and movement end was defined as when the resultant velocity dropped below 50 mm/s for 100ms (Krigolson et al., 2012; Westwood, Heath & Roy, 2003). In order to generally characterize the movement and to confirm that the task was performed in a normal fashion, several spatio-temporal measures were examined. Spatial variables of interest were constant error (CE), the average amount of bias of the movement endpoints relative to the target position and variable error (VE), the average amount of variability of movement endpoints, in the primary movement axis. Temporal variables, measured in milliseconds, were reaction time (RT), defined as the elapsed time from the presentation of the imperative go-tone to the initiation of the limb movement; movement time (MT), defined as the time from movement initiation to the termination of the movement; time after peak velocity (taPV): the period from after peak velocity to movement end (i.e., providing an indication of the amount of time dedicated for on-line corrections to be performed: see Chua & Elliott, 1993).  A R2 analysis of movement position relative to movement endpoint was performed at each decile of the movement trajectory. This analysis has been used to assess the influence of online control on the movement trajectory (Heath, 2005; Messier & Kalaska, 1997). Specifically, it has been suggested that high R2 values indicate a symmetrical movement trajectory, which is typical of reaches in NV. In contrast, low R2 are thought to indicate the influence of online control on the movement trajectory, which is characteristic of reaches in FV. For this analysis, 57  means were submitted to a 9 (Decile: 10%-90%) by 2 (Vision: FV, NV) by 3 (Target size: Large, Med, Small) by 2 (amplitude: close, far). Any significant interactions were decomposed using simple main effects with Bonferroni correction to alpha (p < .05).  Figure 2.1. The sequence of events occuring on a given trial. Participants first fixated on the fixation cross on the right and positioned the stylus over the home position on the left of the tablet. This was followed by the target preview of 250ms, when the target preview was marked as an event. Following the target preview, vision of the target was removed for 2 s. A ‘go’ tone was presented following the delay period, which informed participants to move to the target location. On FV trials, the target reappeared with the ‘go’ tone, while on NV trials, the target did not reappear. 58   The electroencephalograms (EEG) were recorded from 64 electrode locations using BrainVision Recorder software (Version 1.3, Brainproducts, GmbH, Munich, Germany). Electrodes were mounted in accordance to the 10-20 system to a fitted cap (Klem, Lüders, Jasper & Elger, 1999). In addition, vertical electrooculograms were recorded from electrodes placed above and below the right eye. Electrodes were also placed on the left and right mastoid processes (LM, RM) and used as reference electrodes, while AFz was used as the ground electrode. Electrode impedances were kept below 10 kΩ at all times. The EEG data were sampled at 500 Hz, amplified (Quick Amp, Brainproducts, GmbH, Munich, Germany) and filtered through a passband of 0.017 Hz—67.5 Hz (90 dB octave roll off). Following data collection, the EEG data were filtered through a (0.1 Hz—40 Hz passband) phase shift free Butterworth filter and re-referenced to an average reference. An independent component analysis (ICA) was performed using built-in routines from the Analyzer software to detect and remove artifacts from the data. Potentials accounting for stereotyped artifacts, including eye blinks, eye movements, and muscle activation, were identified and removed from the data. Trials in which the change in voltage at any channel exceeded 35 uV between contiguous sampling points were also discarded. Less than 10% of the ERP data were discarded for each participant. All ERP waveforms were baseline corrected using a 200 ms epoch immediately before target preview (-200 to 0 ms) and prior to movement start (-350 to -150 ms). For a statistical examination of the ERP data, the mean voltage for a 50 ms window centered on peaks of interest for each participant, electrode channel, and experimental condition (see Luck, 2005).  The ERP of interest with respect to target encoding was the P3 component. As done in the experiment by Kourtis et al. (2012), parietal and occipital electrodes were pooled to 59  derive an averaged parieto-occipital waveform for the P3. Such a pooling approach allowed for the detection the effects of manipulations of task difficulty on cortical activity as used in Kourtis et al. (2012). Motor-related cortical potentials were the MP and the N4 components. Both behavioural and ERP data were analyzed using a 2 (target vision: FV, NV) by 3 (target size: Large, Medium, Small) by 2 (target amplitude: Close, Far) ANOVA. Any significant interactions were decomposed using paired samples t-test with Bonferroni correction to alpha (p < .05).  2.3 Results 2.3.1 Behavioural Results For the RT analyses, there was a main effect of vision, F(1, 13) = 96.59, p < .001, which reflected that participants initiated their reaching movements in less time with than without vision of the target (FV: 283 ±15.5 ms; NV: 322 ±17.4 ms). Also, there was a main effect of target amplitude, F(1, 13) = 12.06, p < .01, which revealed that reaction times were shorter when reaching to the Close target vs. the Far target (299 ±15.8 ms vs. 305 ±16.9 ms). The analysis of MT (Figure 2.2) also revealed a main effect of target size, F(2, 26) = 5.26, p < .01, as MT was inversely related to the size of the target (Small: 460 ±28.5; Medium: 440 ±27.5; Large: 433 ±28.3 ms). As well, the MT analyses revealed a main effect of amplitude, F(1, 13) = 146.50, p < .001, as movements to the closer target were completed in less time than movements to the farther target location (414 ±25.6 and 474 ±29.7 ms).   60   Figure 2.2. Average movement time (ms) with standard error bars, as a function of vision condition, target size across the close and far targets  The analysis of taPV, yielded a target vision by target size interaction, F(2, 26) = 6.84, p < .001. Further inspection of the interaction revealed differences across target sizes in the FV condition between the Large (307 ±28.7 ms) and Small target (343 ±29.3 ms), while in NV the differences observed across target size failed to yield any significant effects (Large: 322±29.6 ms; Medium: 331±28.3 ms;  Small: 346±29.7 ms). The analysis of PA (Figure 2.3) yielded main effects of amplitude, F(1, 13) = 20.01, p < .01, (Close: 4812 ±555; Far: 5822 ±751 mm/s/s) and target vision, F(1, 13) = 6.54, p < .05 (FV: 5548±677; NV: 5086±646 mm/s/s). The analysis of PD yielded main effects of amplitude, F(1, 13) = 34.61, p < .01, (Close: 4680 ±553; Far: 5922 ±750 mm/s/s) and target size F(2, 26) = 4.72 ,  p < .05 (Large: 5566 ±694; Medium: 5429 ±661; Small: 4910±633 mm/s/s). 37039041043045047049051053022 (Close) 34 (Far)Movement time (ms) Amplitude (cm) NV SmlNV MedNV LrgFV SmlFV MedFV Lrg61   Figure 2.3. Acceleration (mm/s/s) profiles as a function of target amplitude (Close target and Far target).  The CE analyses failed to yield any significant effects, Fs < 3.93, ps >.41. In contrast, for VEx (Figure 2.4), there was a main effect of target vision, F(1, 13) = 24.61, p < .001, a main effect of target size, F(2, 26) = 4.49, p < .05, and a target vision by target size interaction, F(2, 26) = 4.23, p < .05. The breakdown revealed that endpoint precision was inversely proportional to the size of the target for reaches in FV, while there were no differences in precision for reaches in NV. -8000-6000-4000-2000020004000600080000 100 200 300 400 500Acceleration (mm/s/s) Time (ms) Close TargetFar Target62   Figure 2.4. Variable error in movement amplitude (mm) in the primary x-axis with standard error bars as a function of vision condition and target size (FV = target vision, NV = no target vision)  The R2 analysis of movement proportion relative to movement endpoint revealed a decile by target vision interaction, F(8, 112) = 24.77, p < .001. Breakdown of the interaction (Figure 2.5) showed significantly smaller R2 values for FV than NV trials at the 50-80% deciles of movement.    789101112131415Large Medium SmallVariable error (mm) Target Size FVNV* * 63   Figure 2.5. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement.  2.3.2 Electroencephalographic results 2.3.2.1 Target Encoding Visual inspection of the waveform averaged with respect to target preview identified the P3 component at approximately 290 ms post target onset (Figure 2.6). The analysis of the pooled parietal electrodes, which revealed a main effect of target vision, F(1,13) = 7.38, p < .01, as trials without target vision generated a greater positive P3 amplitude as compared to trials with target vision (FV: 1.8 ±.43 μV vs. NV: 2.5±.59 μV; Figure 2.7). 0.00.10.20.30.40.50.60.70.80.91.010 20 30 40 50 60 70 80 90R2 Proportion of Movement (%) FVNV* * * * 64   Figure 2.6. Grand average ERP waveforms locked to the onset of the target preview as a function of vision condition (FV: target vision; NV: no target vision). The P3, as indicated by the triangle, shows the location of the peak where the 50 ms window was taken for averaging of each condition.  -3-2-10123-200 0 200 400 600 800Voltage (μV) Time (ms) FVNV65   Figure 2.7. Average peak amplitudes of the P3 component observed at the pooled parietal electrodes. Analysis of the peak amplitudes revealed a main effect of vision.  2.3.2.2 Movement Evoked Potentials Visual inspection of the waveforms averaged with respect to movement start identified two negative peaks (Figure 2.8), namely component MP at 25 ms and N4 at 160-190 ms after movement initiation, at electrode Fz. Analysis of the MP (Figure 2.9) component revealed a main effect of vision, F(1,13) = 46.12, p < .01, as the MP amplitudes with target vision were greater than without, at -7.9 ±.76 and -6.5±.66 μV, respectively. The analysis of the second negative component (i.e., N4) observed at approximately 160 ms (FV) an 190 ms (NV) post movement initiation (Figure 2.10), yielded a main effect of target vision, F(1,13) = 4.22, p < .05, as the amplitude of the component was greater on trials where vision of the target was available as compared to trials without vision of the target, at -6.95 ±1.5 and -6.18 ±1.4 μV, respectively. Also, there was a main effect of movement amplitude, F(1,13) = 5.47, p < .05, as the N4 0123FV NVVoltage (μV) Condition * 66  amplitude was greater for reaches to the far target as compared to the close target, at -7.51 ±1.6 and -6.43 ±1.3 μV, respectively.    Figure 2.8. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition and target amplitude for electrode Fz. The N4 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition.  -9-8-7-6-5-4-3-2-101-400 -200 0 200 400 600Voltage (μV) Time (ms) FVNV67   Figure 2.9. Grand average amplitudes of MP as a function of vision condition and target amplitude with standard error bars   Figure 2.10. N4 grand average amplitudes as a function of target vision (FV: target vision; NV: no target vision) with standard error bars  -9-8-7-6-5-4-3FV NVVoltage (μV) -9-8-7-6-5-4-3FV NVVoltage (μV) * * 68   Figure 2.11. Grand average ERP waveforms, at electrode Cz, time-locked to the onset of the reaching movement as a function of vision, target size and target location. The peak of interest, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition.  -7-5-3-113-400 -200 0 200 400 600Voltage (μV) Time (ms) FV_Lrg_CloseFV_Lrg_farFV_Med_CloseFV_Med_farFV_Sml_CloseFV_Sml_far69   Figure 2.12. Averages for the negative component observed at approximately 170 ms post movement onset, as a function of vision, target size and movement amplitude with standard error bars.   Another set of observations yielded a modulation of a negative component observed at electrode Cz at 170 ms post movement onset (Figure 2.11). The analysis of this component yielded a vision by target size by target amplitude interaction, F(2,26) = 5.65, p < .01. Breakdown of the interaction (Figure 2.12) revealed that when vision of the target was available for reaches to large and medium targets, a modulation based on movement amplitude occurred. That is, reaches to -4-3.5-3-2.5-2-1.5-1-0.50Lrg_Close Lrg_Far Med_Close Med_Far Sml_Close Sml_FarVoltage (μV) FVNV* * 70  the far target yielded larger amplitudes of the second negative component as compared to reaches to the close target. Meanwhile, no differences were observed across reaching amplitudes for the small target, however both conditions yielded equally high amplitudes of the negative component, which were significantly greater than for reaches to the small target in NV. 2.4 Discussion The main goal of the current study was to examine the neural correlates associated with target encoding and goal-directed aiming under varying conditions of task difficulty and visual feedback. In order to adequately interpret the electrophysiological results, the behavioural findings were first inspected to first show that the experimental manipulations yielded the anticipated effects on the aiming performance. 2.4.1 Behavioural Results Analyses of RT yielded differences based upon vision and movement amplitude, as reaches without vision and to the far target resulted in longer RTs. This is indicative of the planning of the movement accounting for the availability of visual feedback and movement amplitude. In contrast, variations in target size may be more associated with online control. The analysis of MT revealed effects for both components of task difficulty (i.e., target amplitude and target size), as described by Fitts’ law (1954). As expected, movements to the far target resulted in longer movement times as the limb would have to travel further than the close target. As a consequence, reaches to the Far target also had higher peak velocities than reaches to Close targets. As well, movement times varied based on target size. When movement accuracy requirements were not as stringent (i.e. Large target), MTs were shorter as compared to when it was more stringent (i.e., Small target). While no differences in MT based on visual feedback 71  availability were found, differences were observed for time after peak velocity. That is, when individuals had vision of the hand and target, they generally spent more time after peak velocity to attempt to amend their limb trajectory as they were homing in onto the target (i.e., for the Medium and Large targets). Accordingly, endpoint variability for reaches with target vision were modulated by target size, as reaches to the Small target had less endpoint variability as compared to Medium and Large targets. In contrast, endpoint variability for reaches without target vision did not differ across all target size conditions.  Variations of MT were present across all target sizes for reaches with target vision. In the current experiment, individuals were still able to see their limb, thereby allowing individuals to perform online corrections to the limb trajectory. However, individuals were still more accurate when they had vision of both the limb and target. Overall, these results are typical of what is found with a standard discrete Fitts’ law reaching task, thus the brain activity components associated with these behaviours should similarly reflect differences yielded by variation in task difficulty and target vision availability. 2.4.2 Target Encoding Variations in the component P3 in response to the size and location of the target and the visual condition were examined (Koutis et al., 2012; Krigolson et al., 2012). The analyses yielded a main effect of vision condition, as component P3 was greater when participants were encoding the target prior to a reach without vision of the target as compared to with target vision. As such, the results were consistent with the findings of Krigolson et al (2012), who suggested that the larger P3 amplitudes reflect the recruitment of motor planning processes in preparation for a movement without vision of the target. Meanwhile, no differences were observed for P3 based 72  on target amplitude and size. Given that the most important source of information was vision of the target, it appears that individuals prioritized generating a motor plan and storing it in memory. 2.4.3 Movement Evoked Potentials The analyses of the movement related cortical components MP and N4 evoked by the execution of the reaching movement yielded differences based vision of the target, as the amplitudes of these components were greater for reaches with than without vision of the target. This finding was similar to that of Krigolson et al. (2012), who suggested that the smaller amplitudes of MP and N4 for reaches without target vision were a result of a conservative strategy adopted participants when vision of the target is limited (e.g., Elliott, Hansen, Mendoza & Tremblay, 2004). Indeed, Krigolson et al. (2012) found that reaches without vision of the target undershot the target location more than reaches with vision of the target, which corroborated their claim that individuals adopted a conservative reaching strategy. In the current study, however, no significant differences in endpoint accuracy between reaches with and without vision of the target were found. Instead, individuals were less precise when vision of the target was unavailable during the reaching movement, which could indicate that the lower MP and N4 amplitudes led to less precise control of the limb trajectory for reaches without vision of the target. Furthermore, the analysis of component N4 revealed it was sensitive to differences in movement amplitude, as reaches to the far target yielded greater N4 amplitudes than reaches to the close target, which was similar to the findings by Kirsch & Hennighausen (2010). The larger N4 for reaches to the far target, are presumed reflect a greater amount of cortical motor output sent to the limb antagonists to slow the limb towards the latter part of the reaching movement. In 73  sum, the findings highlighted above support the notion that components MP and N4 represent two phases of cortical motor output: the first phase propels the limb to the target and the second phase brakes and halts the limb as it homes in onto the target (Sergio & Kalaska, 1998).  Furthermore, a second negative component observed at electrode Cz yielded differences in component amplitude based on target location when vision of the target was available. That is, the amplitude of the component was larger for FV reaches to the far versus the close target location, which occurred for both the large and medium targets. While FV reaches to the small targets failed to yield significant differences across target location, the amplitude of these components were significantly greater than the amplitudes exhibited to reaches without vision of the target. As these component amplitude differences were only observed when vision of the target was present, the greater amplitude of this component could be attributed to motor output associated with on-line visual feedback use, as the limb is homing in onto the target.  2.5 Conclusion In all, this study provided a better understanding of how the motor system accomplishes reaching actions of varying task difficulty and visual feedback availability. Larger P3 amplitudes were observed during target encoding when individuals were encoding the target in preparation for a reach without vision of the target. This result supports the motor hypothesis (Ghafouri et al., 2000; Krigolson et al, 2012), which suggests that when individuals are encoding a target prior to a reach without vision of the target, a motor plan is generated at the time of target onset and stored in memory until the movement is executed. As for the movement-related cortical potentials evoked by the execution of the reaches to the targets (i.e., MP and N4), smaller component amplitudes were observed when vision of the target was unavailable during the reaching movement. Also, the second component N4, showed variations based on target location, 74  as reaches to the far target location yielded larger N4 amplitudes than reaches to the close target. Ultimately, these findings support the notion that MP and N4 reflect two phases of cortical motor output used to propel and halt the limb (Sergio & Kalaska, 1998). In addition, larger negative components observed at electrode Cz when reaching in FV, hints at differences in motor output relating to online visual feedback utilization.  75  Bridging Summary In study 1, components P3 and motor MP/N4 associated with the planning and execution of reaching movements during a discrete Fitts’ task (1954) were examined. In summary, the cortical events presumably indexed by P3 were most affected by visual feedback availability of the target on the impending movement, rather than movement difficulty (cf. Kourtis et al., 2010). While the findings do not replicate those of Kourtis et al. (2010), they do not necessarily go against the hypothesis that the P3 relates to the assessment of the motor plan and whether the selected motor plan will be successful (i.e., via forward modelling; Kourtis et al., 2010; Verleger, Jaśkowski & Wascher, 2005). Rather, when individuals are encoding the target in preparation to move, the availability of vision during the movement may take precedence over task difficulty. Therefore, the latter components associated with assessing whether the selected motor plan will achieve the desired outcome, would vary based on vision rather than difficulty. Furthermore, the early movement-related cortical component (MP, Brunia, 1987) showed modulations based upon vision of the target, while the late movement-related cortical component (N4) showed modulations based on movement amplitude and target size. From these findings, it can be suggested that the motor-related cortical components are not simply a result of the motor output associated with the amplitude of the movement, but also target size (cf. Kirsch, Hennighausen & Rosler, 2010). Notably, when vision of the target is available the visuomotor system uses vision of the target to refine the trajectory of the movement, which is reflected in the observed motor-related cortical components over motor areas and manifested in differences in movement behaviours. In study 2 the neural correlates of memory-guided reaching were examined (Elliott & Madalena, 1987; Westwood Heath & Roy, 2003). This behavior is often used as an index of the 76  degree to which the motor system is reliant on real-time visual inputs, and by extension, the fidelity of any held memory representation. In this study, we sought to refine and extend previous EEG work with memory-guided reaches (Cruikshank et al., 2014; Krigolson et al, 2012) by examining the processes required to overcome longer delays (2 s and 5 s), thereby gaining a better understanding of the relative salience of historical events within the cortex and gaining an understanding of the type of representation (motor vs. visual) held in memory.    77  3 Study 2: The Neural Correlates of Reaching Under Varying Delay and Visual feedback 3.1 Introduction Real-time visual feedback of the moving limb and environment improves reaching accuracy and precision (Westwood, Heath, Roy, 2005, Woodworth, 1899). This is in part because visual feedback of the limb trajectory is used to make corrections (Heath, 2005). Often times, however, there are situations where visual feedback of the environment is unavailable, such as reaching to the back of a cupboard for an object, or when we are reaching for a light switch in the dark. As a result we can also rely on our memory of the environment to guide our reaching movements.  Previous behavioural studies examining how memory is used to guide limb movements (e.g., Elliott & Madalena, 1987) were interested in examining the characteristics of memory-guided reaches. In Elliott and Madalena (1987), participants were presented with a preview of the target followed by a delay period of 0, 2, 5 and 10 seconds, during which vision of the target was removed. Following the delay period, participants were instructed to aim to the location of where the target was presented. Results of the total endpoint error revealed that reaches performed following 0 and 2 second delay periods were significantly more accurate than reaches performed following 5 and 10 second delay periods. This led the authors to suggest that an accurate representation of the environment can be used to plan and execute limb movements in the absence of vision for up to 2 seconds following visual occlusion. More recent studies using brain imaging techniques (i.e., electroencephalography: EEG; and the event-related potential technique: ERP) have provided more insight into the neural processes underlying visually-guided vs. memory-guided reaches. Krigolson et al. (2012) 78  showed differential brain activity during the encoding and execution of reaching movements to targets under visually-guided or memory-guided conditions. Following a target preview, an imperative tone was presented shortly after for visually-guided reach trials, which informed participants to reach to the target location. For memory-guided reach trials, a 1 s delay period without vision of the target was introduced between the target preview and the imperative tone. Following the delay period, the imperative tone was presented with participants reaching to the remembered target location. Overall, Krigolson et al. (2012) were interested in cortical components observed prior to and during the movement. One ERP of interest in Krigolson et al. (2012) for cortical activity observed prior to the movement was the positive component (P3) observed over parietal regions of the scalp, at approximately 300 ms following the onset of the target (i.e., during the preview period). This component was greater when individuals were encoding the target prior to a memory-guided vs. a visually-guided reach. Krigolson et al. (2012) suggested that this P3 difference associated with target encoding could be due to the engagement of motor planning processes when preparing for memory-guided but not visually-guided reaches (see also Ghafouri, McIllroy & Maki, 2004). By extension, the motor plan generated during encoding for a memory-guided reach was argued to be stored in memory until the movement was required (i.e., motor hypothesis). For visually-guided reaches, the motor plan was argued to be generated shortly before the commencement of the movement, because the target was visible. Critically, these assertions are contrary to previous suggestions that individuals store a sensory representation of the environment, which is used to generate a motor plan shortly before movement initiation (i.e., sensory hypothesis: Elliott & Madalena, 1987). 79  For the motor-related cortical components associated with the execution of the memory-guided vs. visually-guided reaches, Krigolson et al. (2012) observed a modulation of activity over fronto-central regions of the scalp. Notably the authors were interested in the MP component associated with the initiation of the movement and the N4 component associated with the control of the limb movement (see Kirsch & Hennighausen, 2010; Kirsch, Hennighausen & Rosler). Krigolson and colleagues observed that both components were suppressed in the memory-guided reaches as compared to the visually-guided reaches. The smaller magnitude of the components for memory-guided reaches coincided with lower peak acceleration and deceleration; two features consistent with a conservative control strategy when visual feedback is not available (Elliott, Hansen, Mendoza & Tremblay, 2004; Westwood, Heath & Roy, 2003). These findings support the hypothesis that the components (i.e., MP and N4) reflect two phases of cortical activity relating to outgoing motor commands sent to agonists and antagonists of the reaching limb. The first phase activates the agonists, which propel the limb towards the target and the second phase, which activates the antagonists to brake and clamp the limb as it nears the target (Kirsch, Hennighausen & Rosler, 2010; Krigolson et al., 2012; Sergio & Kalaska, 1998). While this study is one of the first examinations of the neural correlates associated with the planning and execution of visually-guided vs. memory-guided reaches, several issues should be noted. That is, with only a delay of 1 s in the memory-guided condition, it is difficult to explicitly disentangle memory for the visual environment (i.e., sensory hypothesis) from memory of the motor plan (i.e., motor hypothesis) that is generated at the time of target encoding. In particular, the use of the 1 s delay period for the memory-guided but not the visually-guided reaches may have led to some differences in motor preparedness (i.e., CNV; ), which may have confounded the differences observed with target encoding. In addition, previous examinations of 80  memory-guided reaches (i.e., Elliott & Madalena, 1987) have used various delay periods, to examine the effect of memory and temporal decay. In the current study, ERPs relating to the planning and execution of reaching movements under varying target vision, memory delay and presentation schedule were examined. The neural events of interest were brought on by: a) the onset of the target during the target preview and b) the execution of the limb movement to the target. The goal of examining whether variations in brain activity exist during these events was so that inferences could be made into the processes that underlie movement planning (i.e., target onset) and movement execution of the reaching movements. To evoke differences in planning and control of reaching movements, participants performed reaches to targets under various conditions of visual feedback of the target (i.e., FV and NV) and under two different delay periods of 2 and 5 seconds. The conditions were chosen so that individuals would have to rely on a memory representation to execute a reach to the target in NV with the 2 second delay period (Elliott & Madalena, 1987). While when reaching in NV with the 5 second delay period, individual would have to rely on a significantly decayed memory representation to guide their movement. In turn, reaches in FV across both delay conditions, provided a means of comparison when memory was not required in the execution of the reaching movement. Furthermore, the presentation schedule of the various reaching conditions was manipulated across experiments: Conditions were presented in a randomized and blocked order in experiments 1 and 2, respectively. Previous studies have observed disparities in the way the FV and NV are manifested when presentation schedule is manipulated (Elliott & Allard, 1985; Cheng et al., 2009). Specifically, Elliott & Allard (1985) found that when individuals engaged in a randomized reaching protocol, the differences between FV and NV were not as robust as the differences between the same conditions in a blocked protocol. Furthermore, there has been 81  suggestion that when switching from one vision condition to the next (e.g., NV to FV), the current reach movement (i.e., trial n in FV) may be reminiscent in both precision and trajectory amendments, of reach observed in the previous trial (i.e., NV trial n-1; e.g., Cheng et al., 2009). In all, the reminiscence across conditions could attenuate the overall differences in precision and accuracy between FV and NV randomized protocol as compared to the blocked protocol. If indeed differences across FV and NV occur in the behavioural measures in the random and blocked schedules, then it may be possible for differences brain potentials related to planning and movement execution will also occur. For experiment 2, the vision and delay conditions were blocked in order to maximize the behavioural differences and also potential ERP differences. 3.1.1 Target Encoding (P3) For target encoding, the P3 component evoked by the onset of the target during the target preview period was of interest. A greater positive magnitude was expected for the P3 on trials in NV compared to FV due to the involvement of motor planning processes at the time of target encoding (Ghafouri et al., 2000; Krigolson et al., 2012). However, if these differences fail to materialize, then these results would lend more support to the sensory hypothesis (Elliott & Madalena, 1987). Furthermore, if individuals are expecting the long delay period (5 s) between target encoding and movement initiation in NV, the P3 will be larger than the short delay period (2 s). This is because when individuals are expecting a longer delay, they may dedicate more effort into planning and maintaining the memory representation.  With regards to the effects of condition scheduling, the above-mentioned differences were expected at least in the blocked protocol. Under the blocked schedule, the larger P3 amplitude were expected for trials without vision compared to trials with vision because 82  individuals were certain of the vision and delay condition (i.e., expected to use visual feedback to their advantage because they will know when they will have it). In the random protocol, however, vision and delay conditions varied from trial-to-trial, resulting in a potential attenuation of the differences across vision and delay conditions . 3.1.2 Movement Evoked Potentials (MP and N4) Two movement-related components elicited by the execution of a reaching movement were of interest in this study (i.e., MP and N4). The cortical motor component (MP; Brunia, 1987) has been generally taken as a reflection of the cortical drive to the initial impulse of the reach. As such, MP was expected to vary with behavioural measures such as peak velocity of the limb, anticipating larger MP in FV vs. NV. This was the case in part because lower MP amplitudes have been found to accompany the higher magnitude of forces normally observed for reaches in FV vs. NV (Krigolson et al., 2012). For the movement-related cortical component observed towards the latter portion of the reaching movement (i.e., N4; Brunia, 1987), it was also expected to be larger for reaches in FV as compared to NV. The modulation in both these components is thought to reflect a conservative approach to movement execution that individuals adopt when performing reaches in NV. Furthermore, the decrease in magnitude of MP and N4 when performing in NV should be exacerbated by the longer delay periods (5 s). Also, it was expected that behavioural variables, such as peak velocity, would also be lower during NV reaching trials. With regards to scheduling of vision and delay conditions, it was again expected that differences in MP and N4 with respect to vision would only be observed at least in the blocked protocol. Specifically, in the randomized protocol, the neural correlates in the trials in FV were 83  predicted to be more reminiscent of those found in trials in NV, which could in turn lead to smaller or no significant differences across conditions. 3.2 Methods 3.2.1 Participants Thirty-three (33) participants (20 male, 13 female; age range: 18-40 years) were recruited from the University of British Columbia to participate in the study. In Experiment 1 (Randomized protocol), seventeen (17) participants (10 male, 7 female; average age: 25.3 ± 6.3 years) completed the study. In Experiment 2 (Blocked protocol), sixteen (16) participants (9 male, 7 female; average age: 24.6 ± 4.5 years) completed the study. All participants were right hand dominant with normal or corrected-to-normal vision and had no known neurological deficits. Informed consent was provided prior to the experiment, which was in accordance with the research ethics board of the University of British Columbia and the 1964 Declaration of Helsinki.  3.2.2 Task Goal-directed manual aiming movements were performed with a stylus on a graphics tablet in the horizontal plane, to visual targets under varying vision and delay periods. Participants were instructed to gaze a fixation cross (see Figure 26) and aim to the presented target after an imperative stimulus (i.e., auditory tone). 84  3.2.3 Apparatus Participants performed aiming movements on a setup similar to that used by Held and Gottlieb (1958). The experimental setup consisted of a 487.7 x 304.8 mm digitizing tablet (Wacom Ltd Tokyo JP, Intuos4, PTK-1240; with a sampling rate of 197 Hz), and a LCD monitor (DELL U3011, Round Rock, TX, USA; with a refresh rate of 75 Hz), with a half-silvered mirror positioned half way between the level of the tablet and the monitor. Participants rested their chin on a rest at the level of the mirror, such that their eyes were approximately 21 cm above the half-silvered mirror and 69 cm above the surface of the tablet. The stimuli were presented in white, while the background was black. A 1 cm diameter home position was presented to the left of the visual display with the fixation point 26 cm from the home position in the horizontal axis (see Figure 3.1). The targets were 1 cm in diameter and projected along an arc with the furthest targets spaced 5.5 cm from the fixation point. LED lights were attached to the underside of the mirror and provided enough lighting so that individuals were able to see their hand and stylus throughout the experiment. EEG was recorded from 64 electrode locations using BrainVision Recorder software (Version 1.3, Brain Products GmbH, Munich, Germany). Electrodes were mounted in accordance to the 10-20 system to a fitted cap and were referenced to a common ground (Klem, Lüders, Jasper & Elger, 1999). In addition, vertical electrooculograms were recorded from electrodes placed above and below the right eye. Electrodes were also placed on the left and right mastoid processes (LM, RM) and used as reference electrodes, while AFz was used as the ground electrode. Electrode impedances were kept below 10 kΩ at all times. The EEG data were sampled at 500 Hz, amplified (Quick Amp, Brain Products GmbH, Munich, Germany) and filtered through a passband of 0.017 Hz—67.5 Hz (90 dB octave roll off). 85   Figure 3.1. The visual display containing the home position on the left, fixation to the right as well as all possible target locations. From the top of the target array to the bottom, the targets are T3, T2, T1, B1, B2, B3) 3.2.4 Procedure 3.2.4.1 Experiment 1 (Random Procedure) At the beginning of each trial, the home position and the fixation were displayed (see Figure 3.2). The participants were instructed to position the stylus over the home position but not contact the surface until they fixated their eyes onto the fixation cross, which was presented in the right visual field. Individuals were instructed to remain fixated on the cross for the entire trial. Once participants touched the stylus onto the home position the trial was triggered to start. First, a cue was presented around the fixation cross for 250 ms. The cue could either be a square, circle, triangle or hexagon and indicated the vision/delay condition for the given trial. The cues informed participants of the vision and delay condition on each trial. The association of the cues 86  with the different conditions was counterbalanced across participants. After the offset of the cue, the target was presented for 250 ms, along with a signal sent to the EEG system indicating the onset of the target. This was followed by the delay period of either 2 s or 5 s. After the delay, the ‘go’ tone was presented by way of a piezo electric buzzer (1500 Hz, 25 ms), with another signal sent to the EEG system indicating the onset of the imperative tone. Participants were instructed to move as quickly and as accurately as possible.  Figure 3.2. A depiction of the sequence of events on a typical trial. The trial started after the participant foveated on the fixation cross and positioned the stylus onto the home position. A cue was subsequently presented, which informed the participant of the delay period and the visual feedback condition on the impending movement. Following the cue, there was a preview of the target for 250 ms and a delay period (2 s or 5 s) during which the target was removed. Following the delay period, an imperative tone was presented, 87  which informed the participant to move to the target location. On trials with vision, the target reappeared, whereas on trials without vision the target did not reappear.  For trials in FV, the target re-appeared with the ‘go’ tone, providing participants with visual feedback of the target. For NV trials, participants aimed to the location where they believed the target was presented. Participants were encouraged to aim as quickly and as accurately to the target in the FV condition or where the target location was in the NV condition. Once participants completed the movement, they were asked to keep the stylus stationary until a low tone indicated for them to move back to the home position and to prepare for the next trial. The conditions were randomized across blocks of 30 trials. The experiment thus represented a 2 vision x 2 delay design, with 60 trials per conditions, for a total of 240 trials. Breaks were given every 30 trials or at the request of the participant. 3.2.4.2 Experiment 2 (Blocked procedure) The protocol in this experiment was the same as experiment 1, except that the same vision and delay condition was used from trial-to-trial in the same block as opposed to a randomized presentation schedule. The conditions were performed in blocks of 30 trials in a given condition before moving onto the next block condition. 3.2.5 Data Reduction and Analysis Stylus displacement data were analyzed offline in MATLAB (The MathWorks Inc., Natick, MA). Each trial was low-pass filtered (cut-off: 10 Hz) with a 2nd order dual-pass Butterworth filter. These data were differentiated to obtain velocity profiles for each trial. Movement start 88  was defined as when the resultant velocity exceeded 50 mm/s for 100 ms, and movement end was defined as when the resultant velocity dropped below 50 mm/s for 100ms (Krigolson et al., 2012; Westwood, Heath & Roy, 2003). In order to characterize the movement and to confirm that the task was performed in a normal fashion, several spatio-temporal measures were examined. Spatial variables of interest were constant error (CE: the average amount of bias of the movement endpoints relative to the target position) and variable error (VE: the average amount of variability of movement endpoints), both in the primary (X) and secondary (Y) axes of the movement.  Temporal variables, measured in ms, were reaction time (RT: the elapsed time from the presentation of the imperative go-tone and the initiation of the limb movement); movement time (MT: the time from movement initiation to the termination of the movement); time to peak velocity (tPV: time taken for peak velocity to be reached following movement initiation, which provided us with an indication of the reliance on open-loop control); and time after peak velocity (taPV: period from after peak velocity to movement end, which provided us with an indication of the amount of time dedicated for on-line corrections to be performed). For this analysis, means were submitted to a 2 (Vision: FV and NV) by 2 (Delay: 2 and 5 s) ANOVA. Any significant interactions were decomposed using paired samples t-test with Bonferroni correction to the alpha value, which was set at .05. The R2 analyses of movement position relative to movement endpoint were performed at each decile of the movement trajectory. This analysis has been used to assess the influence of online control on the movement trajectory (Heath, 2005; Messier & Kalaska, 1997). Specifically, it has been suggested that high R2 values indicate a symmetrical movement trajectory, which is typical of reaches in NV. In contrast, low R2 indicate the influence of online control on the 89  movement trajectory, which is characteristic of reaches in FV. For this analysis, means were submitted to a 9 (Decile: 10%-90%) by 2 (Vision: FV and NV) by 2 (Delay: 2 and 5 s) ANOVA. Post experimental processing of the EEG data involved filtering the raw data through a (0.1 Hz—40 Hz passband) phase shift free Butterworth filter and re-referenced to an average reference. An independent component analysis (ICA) was performed using built-in routines from the Analyzer software to detect and remove artifacts from the data. Potentials accounting for stereotyped artifacts, including eye blinks, eye movements, and muscle movements, were identified and removed from the data. The events of interest (i.e., target encoding and movement execution) segmented and averaged to obtain event-related potentials (ERPs). Trials in which the change in voltage at any channel exceeded 35 uVs per sampling point were also discarded. Less than 10% of the ERP data were discarded for each participant. All ERP waveforms were baseline corrected using a 200 ms epoch immediately before target preview (-200 to 0 ms) and prior to movement start (-350 to -150 ms). For each experimental condition (visually-guided, memory-guided), ERP epochs were extracted from the continuous EEG and averaged with respect to target preview (200 ms before to 1000 ms after) and movement onset (500 ms before to 600 ms after). Grand average waveforms for the events of interest (i.e., target onset, and movement initiation) were also generated in order to visualize the component of interest (i.e., P3, MP, & N4). For a statistical examination of the ERP data, we calculated the mean voltage for a 50 ms window centered on peaks of interest for each participant, electrode channel, and experimental condition. The ERP of interest with respect to target encoding was the P3 component and the motor-related cortical components were the MP and the N4 components. The ERP data were analyzed using a 2 (vision: FV, NV) by 2 (delay: 5 s, 2 s) ANOVAs, with any significant interactions were 90  decomposed using paired samples t-test with Bonferroni correction to the alpha value which was set at .05.  3.3 Results 3.3.1 Experiment 1: Random Protocol 3.3.1.1 Behavioural Results For RT there was a main effect of vision, F(1, 16) = 77.48, p < .001, as participants initiated their movement in a greater amount of time on trials in NV: 324 ±13 ms, vs. FV: 290 ±13 ms. Also, there was a main effect of delay F(1, 16) = 7.91, p < .02, as reaches following a 2 s delay were initiated in a shorter amount of time as compared to than reaches following a 5 s delay (2 s: 314 ±14 ms; 5 s: 301 ±11 ms). Analysis of MT yielded a main effect of delay, F(1, 16) = 9.50, p < .01, as movements following a short delay were shorter in duration than movements following a long delay (438 ±21 vs 447 ±23 ms). Conversely, taPV yielded a main effect of vision, F(1, 16) = 9.57, p < .01, as there was a greater amount of time spent after peak velocity for reaches in FV as compared to reaches in NV, at 205 ±12 vs. 217 ±11 ms, respectively. There were no significant effects found for PV. For VE in the primary x-axis (Figure 3.3), there was a main effect of vision, F(1, 16) = 15.87, p < .001, as movements in FV were more precise than movements in NV (FV: 6.8 ±.31 vs. NV: 8.1 ± .35 mm). Analysis of VE in the secondary y-axis also yielded a main effect of vision, F(1, 16) = 18.49, p < .001, as reaches in FV more precise than reaches in NV (FV: 4.6 ±.10 vs. NV: 6.0 ± .29 mm). CE in the primary axis, (Figure 3.4) showed a main effect of vision, F(1, 16) = 20.01, p < .001, as trials in FV overshoot the target by 1.8 ± 1.0 mm, as compared to 91  trials in NV, which undershot the target by 1.1 ± 1.4 mm. Furthermore, a vision by delay interaction F(1, 16) = 16.38, p < .05. Breakdown of the interaction revealed that reaches in NV following a 2 s delay yielded lower constant error (-.32 ± 1.2 mm) than after a 5 s delay (-1.8 ± 1.5 mm), while these differences across delay condition were not observed for reaches in FV. While for CE in the secondary axis, no significant effects were found.   Figure 3.3. Variable error in movement amplitude (mm) with standard error bars as a function of vision condition (FV = full vision, NV = no-vision)   012345678910FV2 FV5 NV2 NV5Variable error (mm) Condition * 92   Figure 3.4. Constant error in movement amplitude (mm) with standard error bars as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay)  Finally, R2 analysis (Figure 3.5) of movement limb position relative to movement endpoint, yielded a main effect decile, F(8, 128) = 726.78, p < .001, vision by decile interaction F(1, 128) = 5.99, p < .001. Using simple main effects to compare means throughout each decile revealed main effects of vision at the 80% and 90% deciles, Fs > 9.49, ps < .01 -5 -3 -1 1 3 5FV2FV5NV2NV5Constant error (mm) Condition   * * 93   Figure 3.5. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement.  3.3.1.2 Electroencephalographic Results Visual inspection of the grand average ERP waveforms revealed a target evoked P3 component at approximately 380 ms at electrode PO3 (Figure 3.6). However, the analysis of the peak amplitude of P3 revealed no significant differences Fs< 4.1, ps > .06 associated with experimental manipulations. Visual inspection of the waveform, and in line with previous literature (Krigolson et al., 2012) revealed peaks of interest at 50 ms (MP) and 260 ms (N4) following movement initiation, maximal at electrode FCz. Analysis of the MP component revealed a main effect of delay, F(1,16) = 46.12, p < .01, as the amplitude of the component for reaches performed after a short delay were significantly smaller than for reaches after a long delay, at -6.12 ±1.3 and -6.97 ±1.3 μV respectively. The analysis of the N4 component, (Figure 00.10.20.30.40.50.60.70.80.910 20 40 60 80 100R2  value Movement proportion (%) FVNV* * 94  3.7, 3.8) yielded a main effect of vision, F(1,16) = 46.12, p < .01 with a more negative component apparent for trials with vision, at -8.21 ±1.2 vs. -7.33 ±1.1 μV respectively.    Figure 3.6. Grand average ERP waveforms of electrode PO3, time locked to the onset of the target preview as a function of vision condition. The P3 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition.     -2-1.5-1-0.500.511.522.5-500 -300 -100 100 300 500 700 900Voltage (μV) Time (ms) PO3_FV2PO3_FV5PO3_NV2PO3_NV595   Figure 3.7. Grand average ERP waveforms time locked to the onset of the reaching movement as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay). The N4 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition.  -10-8-6-4-202-400 -200 0 200 400 600 800Voltage (μV) Time (ms) FCz_FV2FCz_FV5FCz_NV2FCz_NV596   Figure 3.8. Averaged N4 peak amplitude as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay) with standard error bars. 3.3.2 Experiment 2: Blocked protocol 3.3.2.1 Behavioural Results For RT, there was a main effect of vision, F(1, 15) = 53.26,  p < .001, as reaction time for NV was shorter than in FV at 349 ±7 vs. 304 ±12 ms, respectively. Also for MT, there was a main effect of delay, F(1,15) = 5.15, p < .04, as movements following a short delay took longer to execute than movements following long delays (492 ±22 vs. 476 ±21ms). For PV there was a main effect of delay, F(1,15) = 4.86, p < .05, as reaches following a long delay attained higher PV than reaches following a short delay, at 1354 ±74 and 1311 ±70 mm/s2 respectively. The analysis of taPV did not yield any significant effects, Fs < 3.43, ps > .08. -12-11-10-9-8-7-6-5-4FV2 FV5 NV2 NV5Voltage (μV) * 97  For VE in the primary x-axis, analysis yielded a main effect of vision, F(1, 15) = 6.54, p < .02, as endpoint variability was greater in NV than in FV (7.60 ±.71 mm and 6.78 ±.71 mm,). VE in the secondary y-axis demonstrated a vision by delay interaction, F(1, 15) = 5.97, p < .03. Decomposition of this interaction revealed that the NV5 delay condition was the most variable of all aiming conditions followed by the NV2 condition, while there were no significant differences between the FV2 and FV5 conditions. Furthermore, there was a main effect of vision, F(1, 15) = 17.8, p < .001, reaches in FV were overall more precise than NV, 4.9 ±.17 mm vs. 5.9 ± .12 mm of error. Also a main effect of delay, F(1, 15) = 5.97, p < .01 revealed that movements with the shorter delay was more precise than movements with long delays. For CE in the primary x-axis (Figure 3.9), there was a main effect of vision, F(1, 15) = 10.11, p < .001), as trials in FV overshot the target by 1.4 ±.65 mm as compared to trials in NV which undershot the target by .8 ±.76 mm. CE in the secondary axis analysis yielded no significant effects. 98   Figure 3.9. Constant error in movement amplitude (mm) with standard error bars as a function of vision and delay condition (FV2 = full vision 2 second delay, NV2 = no-vision 2 second delay, FV5 = full vision 5 second delay, NV5 = no-vision 5 second delay)  The R2 analysis of movement limb position relative to movement endpoint (Figure 3.10), revealed a decile by vision interaction, F(8,120) = 5.42, p < .001. Using simple main effects to compare means throughout each decile revealed main effects of vision at 70 and 80% of the movement trajectory, Fs(1, 15) > 8.9, ps < .01.  -3 -2 -1 0 1 2 3FV2FV5NV2NV5Constant error (mm) Conditions * 99   Figure 3.10. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement. (FV = full vision, NV = no-vision)  3.3.2.2 Electroencephalographic Results The P3 component was maximal at electrode PO3 at 320 ms post stimulus onset; however, this component showed no significant effects attributable to the experimental manipulations, Fs<3.35, p >.08 (Figure 3.11). Visual inspection of the waveform revealed robust MP and N4 peaks at 50 ms and 150-300 ms following movement initiation (Figure 3.12). Analysis of the MP component yielded no significant differences, while the analysis of the N4 component yielded a main effect of delay, F(1,15) = 9.89, p < .01 (Figure 3.13). Trials performed after a short delay elicited smaller N4 amplitudes than trials performed after a long delay (2 s: -6.9 ±1.3 μV; 5 s: -9.3 ±1.3 μV). 0.00.10.20.30.40.50.60.70.80.90 10 20 30 40 50 60 70 80 90 100R2 value Movement proportion (%) FVNV* * 100    Figure 3.11. Grand average ERP waveforms locked to the onset of the target preview as a function of vision condition. The P3 peak, indicated by the arrows shows the location of the peak where the 50 ms window was taken for averaging of each condition.  -2-101234-500 0 500 1000Voltage (μV) Time (ms) PO3_FV2PO3_FV5PO3_NV2PO3_NV5101   Figure 3.12. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition. The N4 peak, indicated by the arrows shows the location of the peak where the 50 ms window was taken for averaging of each condition.  -12-10-8-6-4-202-400 -200 0 200 400 600Voltage (μV) Time (ms) FCz FV2FCz FV5FCz NV2FCz NV5102   Figure 3.13. Grand average N4 peak amplitude as a function of vision condition with standard error bars.  3.4 Discussion The present study examined key brain potentials associated with the encoding of visual targets and the execution of reaches towards these targets under various conditions of visual feedback of the target and delay. The differences in the brain potentials evoked during these events allowed for inferences into the processes underlying planning and execution of the limb movements. In addition, differences observed across random and blocked condition scheduling provided insight into the modulation of these potentials when trial-to-trial conditions were varied or maintained, respectively.   -14-12-10-8-6-4-20FV2 NV2 FV5 NV5Voltage (μV) * 103  3.4.1 Behavioural Findings The behavioural findings in both experiments generally revealed that reaches in FV were more precise (i.e., VE) than reaches in NV. This was similar to previous experiments examining visually-guided vs. memory-guided reaches (Binsted et al., 2006; Elliott & Madalena, 1987; Westwood, Heath & Roy, 2003), supporting the contention that having concurrent vision of the hand and target leads to more consistent endpoint distributions (Binsted et al., 2006; Elliott and Allard 1985, Westwood, Heath & Roy, 2003; Zelaznik, Hawkins & Kisselburgh, 1983). Also notable were movement endpoint biases in the primary reaching axis. Reaches in FV overshot the target (i.e., CE), which indicated participants were perhaps more liberal with the execution of their limb movements with than without vision (Westwood, Heath & Roy, 2003). As the behavioural results did not significantly differ between the random and blocked protocol, the findings were contrary to those of by Elliott & Allard (1985) who found that the differences between FV and NV were less apparent in the randomized conditions. The lack of differences across presentation scheduling could have been due to the use of a single index of difficulty, as movement amplitude was consistent across all target locations. As a result, individuals were able to prepare a similar motor plan from trial-to-trial, thereby decreasing the amount of uncertainty regarding movement amplitude, even though vision and delay were manipulated. 3.4.2 Target Encoding (P3) The first component relating to the encoding and planning of a movement towards the target was the P3 component. As individuals were aware of the upcoming vision and delay condition, it was 104  predicted that the P3 would be larger when encoding a target prior to a reach in NV than in FV. This was based on the motor hypothesis (Krigolson et al., 2012), which predicts that motor planning processes are involved when encoding a target in preparation for NV but not FV. The findings of the current experiment, however, failed to support this motor hypothesis because no significant differences in the P3 were observed across vision conditions. Furthermore, the effects of delay, which was expected to exacerbate the differences in P3 between FV and NV, also failed to materialize. In reconciling these findings with those of Krigolson et al., (2012), the discrepancy in findings could have been due to the experimental protocol used. In Krigolson et al. (2012), there were trial-to-trial variations in movement amplitude. These variations may have led participants to constantly modify their motor plan, to cope with the task demands and take into account visual feedback of the movement in the process. While in this study, the index of difficulty was consistent across target locations. Therefore participants could were able to prepare and reuse part of the motor plan to execute the movements to the different target locations, which only varied in direction. As a result, even though vision and delay conditions were manipulated across trials, they were not enough to necessitate a change in motor plan, and hence that can explain why no significant changes in P3 were observed. 3.4.3 Early Movement Potential (MP) The analysis of movement related cortical potential for the random protocol yielded differences for the early negative component MP based on the delay period, as MP amplitude was greater in the 5 s delay period versus the 2 s delay period. These ERP results were also complemented by the RT results, as reaches following the 5 s delay took less time to initiate than reaches following the 2 s delay. The differences could be attributed to variations in anticipation 105  that occur prior to the imperative tone as the fore-period progressed (i.e., aging fore-period effect: see Trillenberg, 2003). Normally, prior to a reach, a slow negative shift in neural activity is observed over motor areas of the scalp called the contingent negative variation (CNV: Brunia, 1987; Brunia, van Boxtel & Bocker, 2011). This negative shift normally starts at the warning stimulus to move and peaks at the imperative tone to execute the movement. In the randomized protocol, however, even though participants were made aware of the vision and delay condition, they were less certain of the delay period from trial-to-trial, as compared with the blocked protocol. Therefore, if participants were unsure of when the imperative tone would occur, the anticipatory CNV could not be maximally expressed prior to the start of the movement. Specifically, in the randomized protocol, the participants appeared to be more prepared for the imperative done during the 5 s vs. the 2 s delay period. Indeed, referring to the phenomenon of the aging fore-period (Trillenberg, 2003), participants could expect 2 potential delay periods (i.e., 2 or 5 s), early after the warning stimulus. As result of these two options, participants were less prepared for the imperative tone at the 2 s delay mark. However, as the delay period ‘aged’, participants became more certain of when the imperative tone would occur. Hence, the anticipatory CNV was more expressed at the 5 s vs. the 2 s delay period, thereby yielding larger MP amplitudes for the 5 s delay condition. In contrast, for the blocked protocol, no differences in the MP component were observed with respect to the delay periods. With a blocked protocol, individuals were certain of the vision and delay period encountered on every trial. The lack of differences in MP amplitudes could be because individuals could reuse part of the motor plan (re.: same amplitude and same index of difficulty) across different target locations. 106  The above findings were contrary to the hypotheses linking component MP with cortical motor output (Kirsch, Hennighausen & Rosler, 2010; Krigolson et al., 2012). Namely, it was predicted that when vision of the target was not available, individuals would become more conservative with their reaching movement and use less force. The decrease in force would be reflected in lower MP and N4 component amplitudes. This is not to mean that MP does not represent cortical motor output, however, the amplitude of this MP peak is susceptible to changes in anticipatory processes in the brain. 3.4.4 Late Movement Potential (N4) For the second motor related cortical component, it was hypothesized that the amplitude of N4 relates to the second phase of cortical motor output, used to brake and clamp the limb as it reaches movement end (Kirsch, Hennighausen & Rosler, 2010; Krigolson et al., 2012). In the randomized protocol, trials in FV exhibited greater N4 amplitudes than trials in NV. This result coincided with a greater amount of time spent after peak velocity for reaches in FV vs. NV. These findings supported the earlier hypothesis put forth, that reaches in FV were executed with greater amounts of force (i.e., Krigolson et al., 2012). In addition, the R2 analysis yielded significantly greater R2 values for reaches in FV than NV, indicating that the increase in magnitude in N4 for reaches in FV could be related to the use of online visual feedback. In the blocked protocol, the hypothesis that reaches in NV would yield lower N4 amplitudes as compared to reaches in FV was not supported. This hypothesis was based on the above notion that reaches in NV would be executed more conservatively yielding lower, MP and N4 amplitudes (Krigolson et al., 2012). Instead, the amplitude of N4 was greater when the reaches were performed following a delay of 5 s vs. 2 s. At the behavioural level, PV of the limb 107  was also higher for reaches following the 5 s versus the 2 s delay. Because higher PV requires stronger decelerative forces, the N4 amplitude could be related to the amount of motor output and force to slow down the limb as it is homing in onto the target, as suggested by Kirsch, Hennighausen & Rosler (2010). The explanation for the observed differences across delay conditions could be that once the memory representation of the target had encountered a significant amount of decay following a 5 s delay period (Binsted et al, 2006; Elliott & Madalena, 1987), individuals more likely relied on a strategy in which they simply executed a fast and stereotyped initial impulse to the target. By executing a stereotypically forceful and fast limb movement towards the target (i.e., high PV), individuals would increase the amount of variability in the limb trajectory (Schmidt, Zelaznik, Hawkins, Frank & Quinn, 1979). In order to slow down the limb as it neared the target, the N4 was thus also greater. Such an explanation is further supported by the fact that the larger N4 amplitudes and higher PVs in the 5 s vs. 2 s delay conditions were only observed in the blocked protocol. It is possible that, because the same vision and delay conditions were presented on each trial in the blocked protocol, individuals had a greater tendency to plan their movements in the blocked versus the randomized condition. 3.5 Conclusions In the current study, differences in target encoding were not between reaches in FV and NV, which was contrary to the predictions of the motor hypothesis (Krigolson et al., 2012). This result rather supports the sensory hypothesis (Elliott & Madalena, 1987), which suggested that the location of the target is stored in visual memory and is used to generate a motor plan shortly before the execution of reaching movements in both FV and NV. Furthermore, results for MP and N4 support the idea that these components are related to the two phases of cortical motor output used to initiate and brake the limb, respectively (Sergio & Kalaska, 1997). The current 108  study added to the idea of two phase of cortical motor output by showing that in the randomized protocol, MP amplitudes were susceptible fluctuations in movement anticipation as the foreperiod progressed. That is, greater MP amplitudes were observed when the timing of movement onset was predictable (i.e., MP greater in 5 s vs. 2 s) due to the aging foreperiod effects (Trillenburg et al., 2003). With respect to component N4, larger amplitudes paired with greater PV values were observed following the long foreperiods in the blocked but not in the random schedule. Such results were indicative of a stereotyped movement impulse, which were executed with greater forceful impulse, hence requiring more cortical motor output to brake and clamp the limb as it reached movement end. In all, not only did these results help reinforce the idea that MP and N4 reflect cortical motor outputs but they also highlight that movement anticipation and delay periods affect the way in which cortical motor components and ultimately the movement behaviours are expressed.   109  Bridging Summary In study 1, manipulation of vision of the target and task difficulty yielded larger P3 amplitudes in target encoding when target was not available during the movement. In study 2, task difficulty was held constant across reaching movements while manipulating vision of the target, delay period (2 and 5 s) and the scheduling of these conditions (random and blocked protocols). There were no differences in P3 amplitudes (re.: target encoding) based on any of these manipulations. This could have been because the task difficulty (i.e., target size and movement amplitude) was consistent across all trials. In comparing the results of study 2 to those of study 1, it appears that the modulation of the P3 as a function of vision of the target requires variations in task difficulty. Therefore, in study 3, vision of the hand, target and task difficulty were manipulated to further examine how these factors affected the way individuals encode the target prior to a reach and are reflected in P3 amplitudes. It has been shown that vision of the hand is the most salient source of information that an individual uses to successfully execute a reaching movement (Heath, 2005). As a result, removal of vision of the hand should lead to larger P3 amplitudes than the removal of vision of the target. Ultimately, such results would add to the knowledge gleaned about target encoding and how reaching conditions modulates P3 amplitudes. As for motor-related cortical potentials, Study 1 showed that the MP and N4 components were greater when reaching amplitudes were longer and targets were smaller (i.e., greater task difficulty). Also, reaches in NV resulted in attenuated MP and N4 amplitudes. These findings support the hypothesis that these components were associated with cortical output from the primary motor cortex. In Study 2, a similar suppression in MP and N4 were observed in the randomized vision and delay protocol. For the blocked protocol, larger N4 amplitudes were observed following the long delay period. Paired with a greater PV, the larger N4 for the 5 s 110  delay conditions indicated that there was a more forceful impulse following a long delay period, which required more force to brake and clamp the limb as it neared the target. Therefore, the findings of Study 2 reinforce the notion that MP and N4 reflect cortical motor outputs. For Study 3, vision of the limb was manipulated during the reaching movement. In the first two studies, vision was manipulated by way of varying visual feedback of the target. As a result when participants performed reaching movements without seeing the target, they still had vision of the reaching limb, which still allowed for online visual feedback corrections. However, removing vision of the limb should significantly compromise the ability of individuals to perform online visual feedback corrections. As indicated by Krigolson & Holroyd (2007), when visual feedback corrections cannot be made by the visuomotor system, then an alternate fronto-central error detection system will be recruited and manifest as a fERN. Hence, by removing vision of the limb in this study, the observed cortical potentials associated with encoding (i.e, P3) and movement execution (i.e., MP and N4) were expected to increase. These increases could be attributed to a greater reliance of movement planning processes during target encoding, and the addition of feedback error related negativity processes over the course of movement execution.    111  4 Study 3: The Neural Correlates of Real-Time Vision of the Hand and Target During a Goal-Directed Reaching Movement 4.1 Introduction It is generally known that actions performed in the presence of vision are more accurate and precise than actions performed without vision (e.g., Carlton, 1992; Woodworth, 1899). This is because vision provides the necessary information for corrections to be made to the limb trajectory as it nears the intended target. Scientific research has sought to uncover the processes that underlie the planning and control of human actions, through a variety of behavioural experiments. These experiments usually involve reaching to targets using the upper limbs under varying conditions of visual feedback of the hand and the target (Binsted et al., 2006; Cheng et al, 2010; Elliott & Allard, 1985; Elliott & Madalena, 1987; Heath, 2005; Westwood, Heath & Roy, 2003). While behavioural studies have accumulated a large body of knowledge about processes underlying the planning and execution of human limb movements, more recent experiments have implemented electroencephalography (EEG) to further elucidate how these processes work (Krigolson et al., 2012; Torrecillos et al., 2012). In the current study, electroencephalography was used to observe the brain potentials during a simple manual reaching task under different vision conditions. Recently Krigolson et al. (2012), observed differences in brain activity when individuals were planning and executing reaches to targets under various visual feedback conditions (i.e., visually guided vs. memory guided). Their investigations reported on the neural correlates of the visual encoding of a target during the planning of a reaching movement as well as the motor-related cortical potentials associated with the execution of the movement itself (Krigolson, Bell, 112  Kent, Heath, & Holroyd, 2012). For target encoding, the authors observed a positive component evoked by the onset of the target prior to the reaching movement. This positive component occurred approximately 280 ms following the onset of the target and was larger when individuals were preparing for a memory-guided vs. visually-guided movement. It was suggested that this larger P3 amplitude observed in the memory-guided conditions was due to the involvement of motor planning processes (i.e., motor hypothesis) that occur when encoding the target prior to memory-guided but not visually-guided reaches. For the neural correlates associated with the execution of the reaching movement, two negative components over the fronto-central regions of the scalp at approximately 50 and 300 ms following movement initiation were observed. These motor-related cortical components, named MP and N4, respectively (Brunia, 1987), have previously been implicated with activity in the primary cortex and outgoing motor commands (Kirsch & Hennighausen, 2010, Kirsch, Hennighausen & Rosler, 2010). Krigolson observed that the amplitudes of these motor-related cortical potentials (i.e., MP and N4) were significantly smaller in the memory-guided vs. the visually-guided reaches. To complement these findings, the behavioural results also yielded lower peak acceleration (PA) and deceleration (PD) values during memory-guided reaches. Taken together, it was suggested by the authors that differences in these observed components were a result of a more conservative execution approach, when reaching to a remembered target location (see also Elliott, Chua, Pollock, & Lyons, 1995 for similar behavioural findings). Furthermore, these results also reinforced the notion that the observed motor-related cortical potentials have to do with motor output, as the first deflection reflects output to the agonist muscles and the second deflection reflects a second volley of motor output to the antagonist muscles, which act to halt the limb as it nears the movement end (Kirsch & Hennighausen, 2010; 113  Kirsch, Hennighausen & Rosler, 2010; Sergio & Kalaska, 1998). Indeed these findings provided another glimpse into the way the motor system contends with variations in visual feedback of the target. However, considering that vision of the effector (i.e., cursor) was provided across all conditions, the study did not address how the central nervous system utilizes visual information regarding the limb. Heath (2005) examined the behavioural effects of variations in visual feedback of the hand and target on goal-directed manual aiming. Participants performed reaches to a target with either vision of: a) both the hand and the target; b) the hand only; c) the target only; d) neither the hand nor the target. Not surprisingly, the results of the experiment showed that individuals were most accurate when they had vision of the hand and target. Importantly, when individuals had vision of the hand only, they exhibited better movement precision than when vision of the limb was not provided. Furthermore, when comparing the R2 analysis of movement position at peak acceleration, peak velocity and peak deceleration relative to movement endpoint, reaches with vision of the hand and target and with vision of the hand only both yielded low R2 values. The low R2 values indicated that the position achieved at PA, PV and PD were not predictive of the endpoint position, as there were possible amendments to the limb trajectory during the movement. In contrast, reaches without vision of the limb (i.e., target only), exhibited higher R2 values, indicating less online trajectory amendments to the limb under these conditions. This led Heath (2005) to suggest that when vision of the limb was available, individuals used real-time visual information of the limb to perform trajectory amendments, which can be facilitated by the dorsal visual stream (Goodale & Milner, 1992; Milner & Goodale, 1995). However, when vision of the limb was not available during the reaching movement, individuals were forced to rely on a remembered and, possibly, less accurate ventral stream representation of the environment to help 114  guide their limb to the target (Westwood, Heath & Roy, 2003). With these types of behavioural studies, the function of vision for the control of ongoing movement has been established. However, without the application of appropriate neuroimaging techniques, conclusions are unable to be drawn with regards to the cortical processes and their time courses.  At present, no study has examined the effect of limb vision of the hand on the brain potentials expressed during movement planning and control. However, studies that have manipulated limb trajectory errors have provided some insight into how movement related potentials could differ based upon vision of the hand. Notably, the hierarchical error detection model put for by Holroyd & Krigolson (2007), suggests that the brain uses different error detection systems depending on the ability, type and degree of error that is encountered during the reaching movement. Namely the model hypothesizes that there are two types of errors that occur during reaching movements, low and high level errors, each of which are dealt with by different areas of the brain. Low level errors, such as errors in movement trajectory, or a jump of the target, results in a discrepancy between the actual and desired outcome. These errors however are usually correctable over the course of a reaching movement through visual feedback use, which is mediated by the posterior parietal cortex (PPC; Milner & Goodale, 1995). Evidence supporting the PPC’s role in mediating limb/target corrections was provided by Krigolson, Holroyd, Van Gyn and Heath (2008). In that experiment, participants performed reaching movements by guiding a cursor on a screen from a home position to a visual target. On a portion of the trials, the location of the target jumped following the initiation of reaching movement, inducing a low level error. As a result, participants would have to minimize this low level error by correcting the limb trajectory to the new location. On half of the trials with a target jump, participants had the ability to move the cursor to the new target location (i.e., correctable), while 115  on the other half of the trials, participants did not have the ability to do so (i.e., uncorrectable). If participants were unable to modify the position of the cursor following the target jump (i.e., uncorrectable), the low level error would spiral into an outcome error. As a consequence a feedback error related negativity (fERN) would be observed over fronto-central cortex structures, such as the ACC. From the fERN, a reworking of the motor plan would take place, so that subsequent attempts at the task would be more successful. Indeed, these results were corroborated by Torrecillos et al., (2012), who observed that the second negative component generated by the reaching movement (i.e., N4), scaled with the magnitude of error incurred following a perturbation to the limb position during a reaching movement.  As a result, there is reason to believe that if visual feedback of the limb is removed during the reaching movement, the low-level errors that are typically mitigated with vision of the limb would precipitate into high-level outcome errors. As a result, the fronto-central error detection system would be recruited to detect these high level outcome errors, thereby generating a fERN that could influence the motor related cortical potentials (i.e, MP and N4) observed over the course of the reaching movement. Therefore, in the current study the effects of limb and target vision manipulations on the brain potentials exhibited prior to and during a goal-directed reaching movement were examined. Two events of interest using the event-related potential technique were: 1) The preview of the target to the participant (i.e., target encoding), and 2) The initiation of the reaching movement (i.e., movement execution). Similar to Heath (2005), four different reaching conditions deployed in this experiment were: Hand & Target, where vision of the hand and target was available during the reaching movement; Hand only, where vision of the hand was available but with vision of the target removed; Target only, where vision of the hand is removed but vision of the target is available; and No hand/no target, vision of the hand and 116  target are both removed. Hypotheses were elaborated separately for target encoding and movement execution. 4.1.1 Target Encoding For target encoding, the ERP component of interest was the P3 component observed approximately 280-300 ms following the onset of the target. Krigolson et al. (2012) observed larger P3 amplitudes for memory-guided reaches (i.e., Hand only) as compared to visually guided reaches (i.e., Hand and Target). Therefore, one of the goals of this study was to replicate the findings by Krigolson et al., (2012) with the Hand and Target and Hand only conditions. In addition, it was of interest to see whether P3 would also vary based on vision of the hand (i.e., Target only, No hand/no target).  Using the motor hypothesis (Ghafouri, McIllroy & Maki, 2004; Krigolson et al., 2012) to predict the results of the current experiment, the no hand/no target condition was expected to yield the largest P3 amplitude, as reaching under these conditions would occur with the least amount of information. As a result a greater amount of the motor planning would be required and would by the largest P3 amplitude. This would be followed by the target only; and hand only conditions, as reaches under these conditions would still require some level of motor planning prior to movement execution as some information will be unavailable to the participants during the reaching movement. Lastly, the smallest P3 amplitude was expected to be observed in the full vision (i.e., the hand & target) condition, as individuals were not expected to recruit any motor planning processes during the target encoding phase. Instead the recruitment of motor planning processes can be left until shortly before the execution of the reaching movement towards the target. 117  4.1.2 Movement Execution  For the movement-related potentials, components MP and N4 were of interest, occurring shortly after the onset of the reaching movement and during the course of the reaching movement, respectively (Brunia, 1988). Similar to Krigolson et al., (2012), it was predicted that when vision of the target is removed during the reaching movement (i.e., Hand only) individuals would adopt a conservative approach to executing their limb movements. As a result of this conservative approach, amplitudes for MP and N4 were expected to be reduced in the Hand only condition relative to reaches with the Hand and Target. Such a conservative approach was also expected to result in lower PV amplitudes at the behavioural level.  In addition, given the importance of vision of the hand reaching movements, the presence or absence of real-time vision of the hand was expected to affect the way components MP and N4 are expressed. Using the arguments of the hierarchical error detection model (Holroyd & Krigolson, 2007), when vision of the limb was available during the reaching movement (i.e., Hand and target; Hand only), errors in movement trajectory were expected to be mitigated by PPC. However, once vision of the hand was removed, the visuomotor system’s ability to mitigate these trajectory errors using online control was expected to be compromised. As a result, the alternate fronto-central error detection system should be recruited and reflected by the presence of a fERN, which influences the magnitude of the N4 component (i.e., Torrecillos et al., 2012). Therefore, for reaches where vision of the hand was not available (i.e., Target only, No hand/no target) the N4 was expected to yield the largest amplitude. While for reaches with vision of the hand (i.e., Hand and target; Hand only), the N4 was expected to yield smaller amplitudes. Complementary to this, if parietal and occipital structures are indeed involved with mitigating low level errors in movement trajectory (Krigolson & Holroyd, 2007), 118  then for reaches where vision of the hand is available, there should be corresponding activity observed in parietal-occipital areas that are contralateral to the reaching limb. 4.2 Methods 4.2.1 Participants Sixteen (16) participants from the University of British Columbia community completed the study (12 male, 4 Female, average age: 24.2 ± 3.8 years). All participants were right hand dominant with normal or corrected-to-normal vision and had no known neurological deficits. Informed consent was provided prior to the experiment, which was in accordance with the research ethics board of the University of British Columbia and the 1964 Declaration of Helsinki. Upon completion of the study, participants were remunerated $50 for their time. 4.2.2 Task Goal-directed manual aiming movements were performed with a stylus on a graphics tablet in the horizontal plane to visual targets under various conditions of visual feedback of the hand and target. During the experiment, continuous EEG was recorded from a 64-electrode cap, with events of interest (i.e., target onset and movement execution) marked onto the EEG recording. These markers were later segmented and averaged to obtain event related potentials (ERPs).  4.2.3 Apparatus Participants performed aiming movements on a setup similar to that used by Held and Gottlieb (1958). The setup consisted of a 487.7 x 304.8 mm digitizing tablet (Wacom Ltd Tokyo JP, Intuos4, PTK-1240; with a sampling rate of 197 Hz), and a LCD monitor (DELL U3011, Round 119  Rock, TX, USA; with a refresh rate of 75 Hz), with a half-silvered mirror positioned half way between the level of the tablet and the monitor. Participants rested their chin on a rest at the level of the mirror, such that their eyes were approximately 21 cm from the half-silvered mirror and 69 cm from the surface of the tablet. White stimuli were presented on a black background. A 1 cm diameter home position was presented to the left of the visual display with the fixation point 36 cm from the home position in the horizontal axis (see Figure 4.1). The targets were 7 cm by 1 cm, with two possible target amplitudes used, Close: 40 cm; Far: 44 cm. Vision of the limb was manipulated by way of an LED that was attached to the tip of the stylus. When the LED was on, participants had position information of their limb while on the other hand when it was extinguished participants did not have vision of the position of the limb. 4.2.4 Procedure At the start of each trial, a home position and fixation cross was presented to the participants. When the participants were ready to start the trial, they positioned the stylus (with the LED on) at the home position and fixated onto the fixation cross. This was followed by a target preview, where the target was presented for 250 ms and the instructions were not to move to the target yet. Subsequent to the preview was a delay period of 2 seconds, where for Hand/Target trials the target remained on throughout the delay period. For Target Only trials the target also remained on however the LED on the stylus turned off. For the Hand Only trials the LED on the stylus remained on but vision of the target was removed. For the No Target/No Hand trials, the LED on the stylus and target were both not visible. Lastly, in the Control condition, participants remained on the home position and did not execute a movement to the target location. On half of the control trials the target was extinguished during the delay period, while on the other half the 120  target was available. Following the delay, a piezo electric buzzer (1500 Hz, 25 ms) signaled participants to move as quickly and accurately as possible to the target location that was previewed. Once at movement end, participants were asked to keep their hand at movement endpoint until a low-tone informed them to move back to the home position and blink.   Figure 4.1. The experimental procedure used in the study. Note that for target preview, the target remained on for Hand/Target and Target only conditions, whereas for the Hand Only, No hand/No target conditions the target was extinguished. Also two target amplitudes were used (Close: 40 and 44 cm)   Overall, the experiment involved a 5 (Preview condition: Hand/Target, Hand only, Target only, No hand & No target, Control: No movement) by 2 (Target amplitude: Close, Far) design, 121  with 60 trials per condition for a total of 480 trials. The conditions were run in a blocked fashion in 30 trial blocks, with breaks in between blocks or at the request of the participant.  4.2.5 Data Reduction and Analysis The stylus displacement data from each trial were low-pass filtered (cut-off: 10 Hz) with a 2nd order dual-pass Butterworth filter. These data were differentiated to obtain velocity profiles for each trial. Movement start was defined as when the resultant velocity exceeded 50 mm/s for 100 ms, and movement end was defined as when the resultant velocity dropped below 50 mm/s for 100 ms (Krigolson et al., 2012; Westwood, Heath & Roy, 2003). In order to generally characterize the movement, and to confirm that the task was performed in a normal fashion, several spatio-temporal measures were examined. Spatial variables of interest were constant error (CE), which is the average amount of bias of the movement endpoints relative to the target position and variable error (VE), which is the variability of movement endpoints, both calculated in the primary movement axis.  Temporal variables, measured in milliseconds, were reaction time (RT), defined as the elapsed time from the presentation of the imperative go-tone to the initiation of the limb movement; movement time (MT), defined as the time from movement initiation to the termination of the movement; and time after peak velocity (taPV): the period from after peak velocity to movement end, which is also thought to provide an indication of the amount of time dedicated for on-line corrections to be performed (e.g., Chua and Elliott, 1993).  An R2 analysis of movement position relative to movement endpoint was performed at each decile of the movement trajectory. This analysis has been used to assess the influence of online control on the movement trajectory (Heath, 2005; Messier & Kalaska, 1997). Specifically, high R2 values indicate that the limb positions at the given decile(s) are predictive of the position 122  of the limb at movement end. This suggests that fewer amendments to the limb trajectory were made and is typical of reaches in NV. On the other hand, low R2 values indicate that the limb positions do not predict the position at movement endpoint and is suggestive of online control influences on the movement trajectory, which is characteristic of reaches in FV. For this analysis, means were submitted to a 9 (Decile: 10%-90%) by 5 (Vision: Hand & Target, Hand only, Target only, No hand/No target) by 2 (amplitude: close, far). Any significant interactions were decomposed using simple main effects with Bonferroni correction to alpha (p < .05). The electroencephalogram (EEG) was recorded from 64 electrodes using BrainVision Recorder software (Version 1.3, Brainproducts, GmbH, Munich, Germany). Electrodes were mounted in accordance to the 10-20 system using a fitted cap and were referenced to a common ground (Klem, Lüders, Jasper & Elger, 1999). In addition, vertical electrooculograms were recorded from electrodes placed above and below the right eye. Electrodes were also placed on the left and right mastoid processes (LM, RM). Electrode impedances were kept below 10 kΩ at all times. The EEG data were sampled at 500 Hz, amplified (Quick Amp, Brainproducts, GmbH, Munich, Germany) and filtered through a passband of 0.017 Hz—67.5 Hz (90 dB octave roll off). Following data collection, the EEG data were filtered through a (0.1 Hz—40 Hz passband) phase shift free Butterworth filter and re-referenced to an average reference. Independent component analysis (ICA) was then performed to detect and remove artifacts from the data. Components accounting for stereotyped artifacts, including eye blinks, eye movements, and muscle movements, were identified and removed from the data through the ICA protocol with manual confirmation of their presence. Trials in which the change in voltage at any channel exceeded 35 uVs between two samples were also discarded. Overall, less than 10% of the ERP data were discarded for each participant. All ERP waveforms were baseline corrected using a 123  200 ms epoch immediately before target preview or 400 ms prior to movement initiation. For each experimental condition, ERP epochs were extracted from the continuous EEG and averaged with respect to target preview onset (200 ms before to 1000 ms after) or movement initiation (400 ms before to 700 ms after). For a statistical examination of the ERP data, the mean voltage of a 50 ms window centered on peaks of interest was calculated (target encoding: P3; movement execution: MP, N4) for each participant, electrode channel, and experimental condition. Behavioural and ERP movement data were submitted to a 4 (vision condition: Hand & Target, Hand only, Target only, No hand/No target) x 2 (target amplitude: Close, Far) repeated-measures analysis of variance (ANOVA). For the analysis of ERPs with respect to target encoding, a 5 (preview condition: Hand & Target, Hand only, Target only, No hand/No target, Control) x 2 (target amplitude: Close, Far) ANOVA design was used. Significant interactions were decomposed via paired samples t-tests. All statistical tests were evaluated at an alpha level of 0.05.  4.3 Results 4.3.1 Behavioural Results The analysis of RT yielded a main effect of vision F(3, 45)= 6.65, p < .02, as reaches with vision of the hand only (316 ±11 ms) were initiated in the shortest amount of time among all reaching conditions (i.e., all greater than 327 ms). For MT, there was a main effect of amplitude F(1, 15)= 243.53, p < .01, as reaches to the far target took a greater amount of time to complete than reaches to the close target, at 454 ±18 ms vs. 414 ±17 ms, respectively. The analysis tPV yielded a main effect of amplitude, F(1, 15)= 385.17, p < .01, as the time taken to reach peak velocity was greater when individuals were reaching to the far vs. the close target (241± 8.1 ms and 221 124  ±7.5 ms, respectively). For taPV (Figure 4.2), there was a main effect of vision, F(3, 45)= 6.65, p < .02, as reaches with the hand and target yielded the longest time after peak velocity (216 ±11 ms). This was followed by the hand only (203 ±12 ms) and the no hand/target reaches (196 ±14 ms), which did not differ from each other but both yielded longer taPV than for trials with vision of the target only (192 ±11 ms). Furthermore, there was a main effect of amplitude, F(1, 15)= 82.72, p < .001, as reaches to the far target required more time after peak velocity than reaches to the close target (far: 211 ±12 ms; close: 192 ±11 ms).   Figure 4.2. Time after peak velocity (ms) with standard error bars as a function of vision condition and target location.  The analysis of PV (See Figure 4.3) yielded a main effect of amplitude F(1, 15)= 19.21, p < .001, as reaches to the far target yielded greater PV than reaches to the close target  (far: 1574 ±98 mm/s; close: 1296 ±78 mm/s).   150170190210230250270290Hand/Target Hand only Target only No Hand/TargetTime after peak velocity (ms) Vision Condition * * * 125   Figure 4.3. Velocity profiles as a function of each reaching condition  For the analysis of VE in the primary movement axis (Figure 4.4), there was a main effect of vision, F(3, 45)= 26.78, p < .01. Notably, reaches with vision of the hand and target yielded the smallest variable error values (9.4 ±.58 mm), followed by the hand only condition (12.5 ±.59 mm). Reaches with vision of the target only and without vision of the hand and target yielded the largest variable error values (Target only: 16.8 ±1.2 mm; No hand/Target: 16.9 ±.76 mm). 0200400600800100012001400160018000 20 40 60 80 100Velocity (mm/s) Movement Proportion Hand/Target_closeHand/Target_farHand only_closeHand only_farTarg only_closeTarg only_farNo hand/targ_closeNo hand/targ_far126   Figure 4.4. Variable error in movement amplitude (mm) with standard error bars as a function of vision condition.  For CE in the primary movement axis (Figure 4.5), there was a main effect of vision, F(3, 45)= 8.7, p < .01, as reaches without vision of the hand undershot the target (i.e., Target only: -19 ±5.6 mm; No hand/No target: -18 ±5.2 mm), significantly more than reaches with vision of the hand (i.e., Hand/Target: -2.1 ±.79 mm, Hand only: -1.2 ±1.9 mm) 0510152025Hand & Target Hand Only Target Only No Hand/TargetVariable Error (mm) Vision Condition Hand & TargetHand OnlyTarget OnlyNo Hand/Target* * * 127   Figure 4.5. Constant error in movement amplitude (mm) with standard error bars as a function of vision condition.   The R2 analysis comparing the position of the limb at different deciles of the movement trajectory relative to movement endpoint (Figure 4.6) yielded a decile by vision by target amplitude interaction F(8, 120) = 2.05, p<.05. Breakdown of the interaction using simple main effects revealed main effects for vision of the hand across deciles 30-80% when reaching to the close target (Fs>8.41, ps < .001). That is, the R2 values were significantly higher for reaches without vision of the hand (i.e., Target only, No hand/target) as compared to reaches with vision of the hand (i.e., hand and target, hand only). However, when reaching for the far target, for deciles 40-80% only reaches with vision of the hand and target, yielded lower R2 values as compared the rest of the vision conditions (Fs>10.13, ps < .001). -30 -25 -20 -15 -10 -5 0Constant error (mm) Hand/TargetHand onlyTarget onlyNo Hand/Target* 128   Figure 4.6. Mean proportion of variance (R2) in movement endpoints explained by the position of the limb at different proportions of the movement. 4.3.2 Electroencephalographic Results 4.3.2.1 Target Encoding Visual inspection of the ERP traces associated with the encoding of the target during the preview period revealed the component of interest (i.e., P3) at 250-300 ms following the onset of the target (see Figure 4.7). Maximal differences were found at electrode PO3 (Figure 4.8), with subsequent analysis yielding a main effect of preview condition, F(4, 60) = 6.5.1, p < .01, as conditions with vision of the target (i.e., hand & target: 1.06±.29 μV; target only: 1.14±.31μV) yielded the smallest amplitude while conditions without vision of the target (i.e., hand only: 1.72±.38, no hand/no target: 1.57±.45 μV) yielded a larger magnitude of the P3. Lastly, for the control condition, where a subsequent movement was not required, the P3 amplitude attained the greatest magnitude (2.15±.54 μV).  * * * * * * * 129   Figure 4.7. Grand average ERP waveforms locked to the onset of the target preview for all vision conditions at electrode PO3. The P3 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. The dashed lines indicate the control condition where participants were not required to perform a subsequent movement to the target.  -1.5-1-0.500.511.522.53-200 -100 0 100 200 300 400 500 600Voltage (μV) Time (ms) Hand/TargetHand onlyTarget onlyNo hand/targControl130   Figure 4.8. Bar graph depicting the grand average P3 amplitudes as a function of preview condition with standard error bars.   4.3.2.2 Movement Execution Visual inspection of the waveform averaged with respect to movement initiation revealed two negative peaks, the first peak MP occurring at 40 ms and the second peak N4 occurring 180 ms following movement initiation (Figure 4.9). Statistical analysis of the MP component, yielded no differences across reaching conditions, while the analysis of the N4 component (Figure 4.10), yielded a main effect for vision, F(3, 45) = 3.48, p < .02. Trials without vision of the hand yielded larger amplitudes (No hand/Target: -3.19 ±.39 μV; Target only: -2.96 ±.38 μV) than trials with vision of the hand, (Hand/Target: -1.91 ±.50 μV; Hand only: -2.07 ± .38 μV respectively. -0.20.81.82.8hand/target hand only target only no hand/no targetcontrolVoltage (μV) Condition * * 131   Figure 4.9. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition. The N4 peak, indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. The dotted line indicates the control condition where participants were not required to perform a subsequent movement to the target, while the vertical dashed line indicates the time where the average movement end occurred.  -4-3-2-101-500 -300 -100 100 300 500 700 900Voltage (μV) Time (ms) Hand & TargetHand onlyTarget onlyNo hand/Notarget132   Figure 4.10. Bar graph depicting grand averages of the N4 peak, as a function of reaching condition, with standard error bars.   Further, when examining activity over parietal-occipital regions of the scalp, relating to activity in the PPC and visual feedback use during the reaching movement, differences were observed across conditions at electrode PO7. The analysis of the activity observed at PO7 revealed a peak at approximately 550 ms following movement initiation, which was also after movement end (Figure 4.11). Analysis of that peak yielded a main effect of vision of the hand, F(3, 45) = 10.13, p < .01 (vision of the hand: -.88 vs. no vision of the hand: .11 μV). -5-4.5-4-3.5-3-2.5-2-1.5-1-0.50hand/target hand only target only no hand/no targetVoltage (μV) * 133   Figure 4.11. Grand average ERP waveforms locked to the onset of the reaching movement as a function of vision condition. The peak, as indicated by the arrow shows the location of the peak where the 50 ms window was taken for averaging of each condition each condition. The dotted line indicates the control condition where participants were not required to perform a subsequent movement to the target, while the vertical dashed line indicates the time where movement end occurred.  4.4 Discussion In the current study, the neural correlates associated with performing reaches under varying conditions of visual feedback of the hand and target were examined. While previous studies examining these neural correlates have conducted reaching movements with continuous vision of the hand (Krigolson et al., 2012), no study to date has examined the various permutations of visual feedback availability of the limb and/or target and the neural activity associated with aiming under these conditions. Importantly, there was reason to believe that neural activity -1.5-1-0.500.511.522.5-500 -300 -100 100 300 500 700 900Voltage (μV) Time (ms) PO7 Hand & TargetHand onlyTarget onlyNo hand/No targetNo movement134  would vary depending on visibility of the limb and/or target, as demonstrated by the behavioural work from Heath (2005). Specifically, when individuals performed reaching movements with vision of the aiming limb, regardless of whether the target is present or not, online corrections to the limb trajectory were expected to occur. Heath (2005) suggested that real-time vision of the limb during the aiming movement facilitates online trajectory amendments via processes involving the dorsal visual stream (Goodale & Milner, 1992; Milner & Goodale, 1995), which would have not otherwise been available had vision of the limb not been provided.  4.4.1 Replication of Previous Behavioural Findings The behavioural findings of the current study were similar to those of Heath (2005) as individuals were most precise when they had vision of both the hand and target, which was followed by the hand only condition and lastly by the target only, and the no hand/no target conditions. As for endpoint accuracy, conditions with vision of the hand (i.e., hand/target and hand only) led to better performance than trials without vision of the hand (i.e., target only and no hand/no target). The analysis of the temporal characteristics of the reaching movements also complemented these findings, as movement times were longer when vision of the hand was available - most of which was time spent after reaching peak velocity. Furthermore, the R2 analysis also yielded differences based on vision of the hand, as the R2 value was lower for the hand/target and hand only conditions which is indicative of greater online trajectory amendments to the limb when vision of the hand was available.  135  4.4.2 Target Encoding For target encoding, the P3 evoked by the onset of the target during target preview was examined. The results revealed that the P3 component was significantly larger when individuals were encoding the target in preparation for a reach without vision of the target (i.e., No target, No hand/No target). This finding was similar to those of Krigolson et al. (2012), who found greater P3 amplitudes for memory-guided (target occluded) vs. visually-guided (target visible) reaching movement. In contrast, contrary to the predictions of the current study, the removal of vision of the hand did not yield an increase in P3 amplitude, even though it had the greatest impact on the outcome of the movement. This was as evidenced by worse CE and VE values when vision of the hand was removed (i.e., Target only, No hand/No target). If indeed the increase in P3 activity was an attempt by the motor system to generate a motor plan in advance as suggested by Krigolson et al., (2012), logic would dictate that it would extend to the limb occluded reach trials. By occurring exclusively for target occluded reaches, it appears that the increase in P3 was an attempt to remember the location of the target. Even more surprising was the control condition, where participants encoded the target but did not have to move to the target subsequently, yielded the greatest P3 amplitude. Instead, an alternate and simpler explanation can be put used to explain the current results using the Locus coeruleus norepinephrine (LC-NE) system (Ashton & Cohen, 2005; Nieuwenhuis et al., 2005). Namely, generation of the P3 could stem from activation of the LC-NE system (Nieuwenhuis et al., 2005), which releases norepinephrine to the brain when greater attention and arousal is required. In releasing NE, structures involved in remembering the spatial location of the target location would be upregulated. 136  4.4.3 Movement Execution  For the movement-related cortical potentials, it was predicted that removing of vision of the target would cause individuals to adopt a more conservative reaching strategy, as found by Krigolson et al. (2012). The conservative strategy would be reflected in smaller MP and N4 amplitudes when executing reaches without vision of the target (i.e., hand only, no hand/no target). The results of the current experiment, however, did not yield any differences in MP and N4 amplitudes with respect to vision of the target (Kirsch, Hennighausen & Rosler, 2010; cf. Krigolson et al., 2012).  With respect to vision of the hand (i.e., target only, no hand/no target), the removal of vision of the hand was expected to impair the visuomotor system’s ability to perform online corrections. As a result, a fERN was expected to occur towards the later part of the reaching movement, which would yield larger N4 amplitudes as compared to reaches where vision of the hand was available. Indeed, the N4 component was significantly larger when vision of the hand was not available (i.e., Target only, No hand/no target). On the other hand, when vision was available during the reaching movement, the amplitude of N4 was smaller. This was likely because when real-time visual feedback of the hand was available, the visuomotor system could detect and correct errors in the movement trajectory by way of the dorsal visual stream (Milner & Goodale, 1993; Milner & Goodale, 1995; Heath, 2005; Westwood, Heath & Roy, 2003). Indeed, this was corroborated by the findings of the current study that showed that activity over visual areas contralateral to the hand and visual field was greater when vision of the hand was available. Also, as this differentiation in activity continued past movement end, it appears as though vision of the hand was used to assess the outcome of the movement. When vision of the hand was occluded, however, the activity mentioned drastically decreased, hence requiring the 137  recruitment of the fronto-central system to assume error detection duties. This would trigger a fERN and result in larger N4 component amplitudes.  .  4.5 Conclusions In the current study, differences in target encoding were observed across target visible conditions but not vision of the hand, which was contrary to the expectations set forth by the motor hypothesis (Krigolson et al., 2012). Instead, an alternate explanation using the LC-NE adaptive gain system provided a better explanation for why a larger P3 occurs for target occluded trials only. Notably, when vision of the target was unavailable, a generalized release of NE from the LC to the brain may have heightened the level of arousal and was aided by remembering the spatial location of the target (see Nieuwenhuis et al., 2005). This memory representation of the target could have in turn been used to help guide the subsequent reaching movement (e.g., Elliott & Madalena, 1987). Furthermore, results for motor related cortical potentials MP and N4 support the idea that these components were related to the two phases of cortical motor output used to initiate and brake the limb, respectively (Sergio & Kalaska, 1997). The current study added to this knowledge by showing that the N4 component associated with the second phase of cortical motor output likely increases as a consequence of fERN, when vision of the hand was unavailable during the reaching movement. In contrast, when vision of the hand was available during the reaching movement, a greater amount of activity was observed over parietal and occipital structures contralateral to the reaching hand and visual field. Therefore, the current study also provides novel evidence—via manipulation of the vision of the hand—that further supports the literature attributing visuomotor processes to the dorsal parietal areas of the brain (e.g., Milner & Goodale, 1995). In all, these findings support the idea of the hierarchical error 138  detection systems (i.e., fERN vs. dorsal stream processing) that are selectively used depending on the vision of the hand during the reaching movement (Holroyd & Krigolson, 2007).  5 General Discussion 5.1 Summary of Findings 5.1.1 Study 1: The Neural Correlates of Visually and Memory-Guided Reaches Under Varying Task Difficulty In study 1, the hypothesis that the brain encodes task difficulty upon seeing a target (presented for 250 ms) was tested. It was previously shown by Kourtis et al., (2012) that when individuals were presented with an informative pre-cue that specified the amplitude of an upcoming reaching movement, the magnitude of the P3 component associated with encoding this pre-cue scaled with the amplitude specified. That is, when a more difficult movement was indicated (i.e., long arrow = long movement amplitude), the P3 was larger than when a movement of lesser difficulty was specified (i.e., short arrow = short movement amplitude). The explanation for these observed effects was based on the context updating hypothesis (Donchin & Coles, 1988), whereby the inverse and forward model of the upcoming movement was updated in response to the pre-cue. The inverse model would be modified so that the motor plan would suit the requirements of the new movement, while the forward model simulated and predicted whether the motor system was able to achieve a successful movement to the target. With more demanding tasks, however, the simulation and prediction would yield less certainty that a successful movement would occur. Such uncertainty would lead to further modification of the motor plan, hence yielding larger P3 amplitudes. Critically, in the experiments by Kourtis et al., (2012), an endogenous cue (i.e., 139  different arrow lengths) was used to elicit the differences in the P3 component. In study 1 of this thesis, however, the endogenous cue was replaced with an exogenous cue. That is, instead of an informative arrow, the actual target was presented. To favour the target encoding processes, a visual target of a specific target size and amplitude was presented to participants prior to generating a reaching movement to the target on each trial. Furthermore, a delay of 2 s was consistently employed between the target presentation and the cue to initiate the movement. Reaching movements to the target were also executed under two vision conditions to exacerbate the effects of the task difficulty: with vision of the target and without vision of the target. If indeed the P3 component scaled with the difficulty elicited by the target itself, it would further support the notion that the P3 component reflects the encoding and context updating of task difficulty, as suggested by Kourtis et al., (2012). The results of study 1 failed to yield any P3 differences with respect to task difficulty (cf. Kourtis et al., 2012). Instead, the results showed that the P3 was modulated by whether participants had vision of the target during the reaching movements. That is, the P3 was significantly larger when encoding the target prior a reach without vision of the target vs. with vision. Therefore, the results of study 1 do not support the idea that participants encode the target for task difficulty when the actual physical target is presented (cf. Kourtis et al. 2012). Instead an alternate explanation of the differences in P3 can be put forth using the locus coeruleus –norepinephrine (LC-NE) adaptive gain system (Nieuwenhuis et al., 2005). The LC-NE theory proposed by (Ashton & Cohen, 2005; Nieuwenhuis et al., 2005) posits that the P3 is a manifestation of activity from the locus coeruleus, located in the brainstem. It was found through various studies that the LC is responsible for a generalized release of NE to the brain, as it has extensive connections to various structures in the cortex including 140  hippocampus, thalamus and even subcortical structures such as the cerebellum (Ashton & Cohen, 2005; Bouret & Sara, 2005; Nieuwenhuis et al., 2005; Samuels & Szabadi, 2008). The release of NE to these structures increases the responsivity of these structures, so that they can be used if recruited. In the case of study 1, participants were informed of the upcoming task demands when they were presented with the target 2 seconds prior to their reach towards it, in either FV or NV. As a result, when participants were encoding the target prior to a reach in NV, it is possible that the LC was recruited to release NE to the projected sites in order for brain to assist with undertaking the more demanding NV reach. In this case, the release of NE could have led to the recruitment of areas associated with iconic memory, which assisted with remembering the location of the target before reaches without vision of the target. In addition, study 1 included hypotheses pertaining to ERP components relating to the reaching movement. Specifically, it has been suggested that the motor related cortical potentials observed during a reaching movements (i.e., MP and N4) are reflective of two phases of cortical motor output (Kirsch & Hennighausen, 2010; Sergio & Kalaska, 2007). The first phase of output relates to motor commands being relayed to the agonist to propel the limb towards the target and the second phase of activity relates to the commands sent to the antagonists to slow down and brake the limb as it nears the target. To assess this hypothesis, the motor related cortical potentials associated with reaching movements to targets of varying task difficulty were examined. Task difficulty was manipulated through variations in movement amplitude and target size. The manipulation of movement amplitude was used to replicate previous findings by Kirsch and Hennighausen (2010), who demonstrated a scaling of the amplitudes of both MP and N4 components. According to Kirsch and Hennighausen (2010), the scaling of these amplitudes 141  were based on the different force requirements of the agonists and antagonists muscles, respectively, when reaching to different movement amplitudes. Furthermore, target size was the other component of task difficulty that was manipulated in study 1 in order to see if it would also yield changes to the motor related cortical potentials (i.e., MP and N4). That is, for reaches with greater task demand, such as when reaching for a small vs. a large target, a greater amount of online control amendments are required to refine the position of the limb so that it lands on the target (Fitts, 1954; Woodworth, 1899). As such, the greater online limb corrections should manifest larger amplitudes of N4 relating to the motor output towards the end of the movement.  Additionally, vision was manipulated in an attempt to see if availability of vision would alter the way these components are expressed. Notably, it was previously found that reaches made without vision were executed more conservatively (i.e., with lower force) than reaches with vision, which also manifested as smaller magnitudes of MP and N4 cortical potentials (see also Krigolson et al., 2012).  The results of the motor related cortical potentials were similar to findings from Kirsch, Hennighausen and Rosler (2010); Kirsch and Hennighausen (2010), in that larger MP and N4 components were observed in longer movement amplitudes vs. shorter ones. Furthermore, a correlational analysis revealed significant correlations between the amplitude of component MP with PA and component N4 with PD. As well, larger N4 components were observed with the smallest target size as compared to the medium and large target sizes, when vision was available. In contrast, the N4 component did not significantly differ between target sizes when vision was not available. This modulation of the second motor related cortical potential when vision was available might relate to subsequent motor impulses that are related to online control amendments to the limb, when vision of the hand and target are available.  142  In sum, the results of study 1 thus indicated that P3 might be a more generalized adaptive gain feature of the CNS (Nieuwenhuis et al; 2005), whereas the idea that MP and N4 reflect cortical-motor output was further supported (Kirsch & Hennighausen, 2010). Critically, the results of study 1 emphasize that visual feedback availability has a strong influence on the target encoding processes, likely reflected by P3. As well, the online control impulses likely reflected by an increase in the magnitude of the N4 component. 5.1.2 Study 2: The Neural Correlates of Reaching Under Varying Delay and Visual feedback The goal of study 2 was to examine how the visuomotor system plans and executes reaches under varied conditions of visual feedback of the target and the delay period between target presentation and movement initiation. To test this idea, the target was previewed for 250 ms, followed by a delay period of 2 or 5 seconds, before movement onset. Following the delay period, the imperative cue to initiate the movement informed participants to move to the previewed target location. On full vision trials, the target re-appeared, while for no vision trials the target did not. Furthermore, in experiment 1, the vision and delay conditions were presented in a blocked schedule, while in experiment 2, conditions were presented in a randomized schedule. Again, the P3 component associated with target encoding during the preview period was of interest. Based on the results from study 1, it was hypothesized that the P3 associated with encoding of the target would be larger on trials without vision of the target. Also, a longer delay period was hypothesized to yield larger P3 components. Finally, it was also hypothesized that randomizing the vision conditions and delay periods would yield larger P3 components than when blocking them. The reasoning behind these predictions is that when individuals are 143  preparing for a movement with the least ideal reaching conditions (i.e., without vision of the target, and with a long delay period: which would likely to cause memory decay) the LC may be used to involve other brain areas in assisting with planning of the movement (i.e., use of iconic memory). However, when reaching conditions are more ideal (i.e., with vision of the target and short delay period), faculties such as memory may not need to be recruited. Hence the P3 component observed during target encoding for these easier conditions were not expected to be as large. The analysis of the P3, however, failed to yielded differences across delay and target re-appearance conditions. The discrepancy between the results of this study and those of study 1 could have been due to the different methodologies used. That is, three target sizes and two movement amplitudes were deployed in study 1, while in study 2, one target size and movement amplitude was used. The only aspect about the target that was changed from trial-to-trial was the direction. As such, the accuracy demands as well as the amplitude of the reaching movement were known prior to each trial. It appears then, that the task was rather easy for the participants, such that a simple modification to the movement plan was enough to account for variations in movement direction and vision from trial-to-trial. Hence, there were no variations in P3 observed, as assistance from the LC-NE system was not required. In contrast to target encoding, the motor related cortical potential yielded significant differences across vision and delay conditions. In experiment 1 of study 2, vision and delay conditions were presented in a blocked schedule. For the MP component, no differences were observed across vision and delay conditions. While for experiment 2, where vision and delay conditions were randomized, there was a main effect of delay on component MP. That is, reaches following a long delay yielded greater MP amplitudes than for short delays. The differences 144  observed at MP for random but not blocked schedules indicate differences in readiness when individuals were aware of the upcoming delay and vision condition on each trial. Typically, prior to a motor response or reaching movement, there was a gradual increase in negative cortical over frontal areas. This activity has been observed when individuals anticipate the imperative tone to move (i.e., CNV). As the imperative tone occurs, the negativity associated with the anticipatory process summates with the first motor related cortical component MP. Thus, when individuals were able to predict the onset of the imperative tone, the first component MP was larger than when individuals were unable to predict the go tone. Indeed, under the blocked schedule, individuals were able to easily anticipate the imperative tone, such that no differences in MP were observed across vision and delay conditions.  However, under the randomized schedule, the ability of individuals to anticipate the imperative tone was compromised by the trial-to-trial variability. As a result, the MP following a 2 s delay was smaller than the 5 s delay. Behaviourally, there was a significant but small difference observed across reaction time with respect to delay condition. Nonetheless, component MP, which typically was associated with cortical motor output associated with movement initiation, was altered by the anticipatory activity associated with the scheduling of delay period from trial-to-trial. For the N4 component, which is the second component of the motor related cortical activity, different results were seen in experiment 1 (blocked) vs. experiment 2 (randomized). In experiment 1, there was a main effect of delay period, as reaches following a long delay period yielded larger N4 amplitudes as compared to short delays. These results were complementary to peak velocity of the limb movement, as PV was greater for reaches following a long 5 s vs. a short 2 s delay. From these results, it can be suggested that following a 5 s delay period, the memory representation of the target would have accrued a significant amount of decay (Elliott & 145  Madalena, 1987). As a result of this decay, individuals may have no longer relied on the memory representation to guide the limb movement, but resorted to simply dumping the limb into the vicinity of the target in a stereotyped manner. As a result, reaches in both FV and NV following a long delay both exhibited greater magnitudes of peak velocity than short delay reaches, which in turn required equally larger magnitudes of N4 and antagonist activity to stop the limb as it neared the target.   In the randomized protocol, the analysis of N4 yielded a main effect of vision because reaches with vision of the target yielded larger amplitudes than trials without vision of the target. To complement this finding, time spent after peak velocity and endpoint precision were longer for reaches with vision of the target vs. without vision of the target. In combining these findings, it appears that the increase in N4 may be related to online visual feedback use toward the latter part of the reaching movement. Furthermore, these results highlights that when individuals are less aware of the condition on the given trial (i.e., randomized protocol), they may rely more on real-time online control to execute their limb movements, which may be counter to the worst case scenario strategy suggested by other authors (cf. Elliott, Hansen, Mendoza & Tremblay, 2004). Whereas when conditions are blocked individuals may preplan their movements more. Therefore, these results show that the presentation schedule of the conditions affects the way the MP and N4 components are manifested. Thus far, it still appears that these components reflect cortical motor output as they reflect behavioural outcomes of the reaching movement (i.e., relation to PV and TaPV). In addition, the components also appear to be affected by other cortical processes taking place, notably anticipatory processes that predict when the imperative tone will occur, as evidenced by the differences in MP observed in the randomized protocol.  146  In all, the results of study 2 with regards to the P3, failed to replicate the findings from study 1. It appears that the removal of target vision may not be the only factor that contributes to the increase in the P3 amplitude (i.e., LC-NE activation). Instead the removal of vision in combination with the uncertainty of target location due to the varying of reaching amplitudes in study 1, may have both contributed the modulation of component P3. On the other hand, the idea that the components MP and N4 reflects two phases of cortical motor output are further supported in this study (Kirsch & Hennighausen, 2010; Krigolson et al, 2012; Sergio & Kalaska, 1998). Additionally, it appears that anticipatory mechanisms (i.e., CNV) used to anticipate the imperative tone may also have a bearing on the manifestation of component MP. 5.1.3 Study 3: The Neural Correlates of Real-Time Vision of the Hand and Target During a Goal-Directed Reaching Movement Study 3 tested the hypothesis that the brain detects trajectory errors differently based on the availability of vision of the target and limb during goal-directed reaches. In the previous two studies, only vision of the target was manipulated while the reaching limb was always visible. For study 3, the experiments manipulated both vision of the target and the hand during the reaching movement. The individual’s ability to amend errors using online visual feedback of the hand and/or target was used as a proxy for determining whether individuals detect errors differently across conditions. Similar to the previous two studies, the neural correlates associated with target encoding and movement execution were examined. In study 1, the larger P3 component occurred when feedback availability of the target during the upcoming movement was not available. In study 2, however, when movement amplitude and target size were held 147  consistent across trials while manipulating visual feedback of the target and delay period, these variations in P3 failed to manifest themselves. Study 3 aimed to elucidate how the P3 manifests when individuals are encoding a target in preparation for a reach without vision of the hand and target. It was hypothesized that the results of study 1 would be replicated and that removing vision of the target would yield larger P3 components. In addition, it was hypothesized that removing vision of the hand, the greater movement outcome uncertainty would yield even larger increases in P3 than removing vision of the target. The results for target encoding revealed that the P3 was largest when participants were encoding the target in preparation for movements without vision of the target (i.e., hand only, no target and hand). This was different from hypothesized. These predictions were based on the hypothesis, that the P3 was associated with forward modelling (Kourtis et al., 2012), when predicting the outcome of the impending movement based on the hand and target vision conditions. Nevertheless, larger P3 amplitudes were observed when preparing for reaches without vision of the target, which was consistent with study 1. Likewise, study 3 reinforced the idea that the P3 component can reflect the activation of the LC-NE system, which aids with remembering the location of the target, through a generalized release of NE to the brain. For the movement-related cortical potentials, the same MP and N4 components as observed in study 1 and 2 were examined. The findings in these earlier two studies supported the idea that MP and N4 reflect two phases of cortical excitability. During these two phases, neural impulses are sent from the primary motor cortex to the agonist and antagonists of the limb, which act to propel the limb to the target (MP) and control the limb as it nears its target (N4), respectively. Recent studies, however, have also implicated error detection processes in the manifestation of the N4 (Krigolson et al., 2007; Torrecillos et al., 2014). Notably, it was 148  observed that when individuals were performing a goal-directed aiming task and encountered a significant amount of error due to target or limb perturbations, the N4 observed scaled with the amount of error. It was suggested by Torrecillos et al. (2012) that the manifestation of this component had to do with feedback error related negativity (FRN), which falls in-line with the hierarchical model of error processing. That is, under normal goal-directed reaching conditions, individuals rely on real-time vision to provide trajectory amendments to rectify low level target errors. As a consequence of these online amendments, the movement endpoint is more accurate than in conditions where the vision of the limb and online corrections were not present. However, under circumstances when the error encountered is so large that the online amendments are unable to rectify the error such as limb perturbation or target jump, these errors become high-level outcome errors. Such high-level errors require a different error detection system, which presumably involves the fronto-central error detection system, including the anterior cingulate cortex (Krigolson & Holroyd, 2007). When the error is detected by this system, it manifests as a negative component over fronto-central regions of the scalp. Such component is often referred to as an error related negativity (ERN) or feedback related negativity (FRN). As a result of this FRN, the motor system attempts to correct the motor plan, so that when the errors encountered on this reaching movement will be minimized on subsequent reaches. Therefore, there is reason to believe that the N4 may not only represent the second phase of cortical motor output to brake and control the limb, but it may also reflect feedback related negativity associated with error detection processes. Therefore, the one of the goals of study 3 was to examine how vision of the hand and target would affect the manifestation of the N4. It was hypothesized that the N4 component would be larger when vision of the hand was removed during the reaching movement. This is 149  because, when vision of the hand is removed, the lack of real-time vision of the hand causes the visuomotor system to be ineffective in correct low-level limb trajectory errors (Krigolson et al., 2007; Westwood, Heath & Roy, 2003). As a result, these lingering low-level errors would precipitate into high-level errors, which would cause the fronto-central error detection systems to be recruited, ultimately manifesting as a FRN. When vision of the target was removed but vision of the hand remained, the N4 was expected to be smaller than when reaches we made without vision of the hand. This was because online corrections to the hand trajectory can still be made, even without vision of the target (see Heath, 2005). Lastly, it was expected that when vision of the hand and target was available during the reaching movement, the N4 would be the smallest. This was because when vision of the hand and target were available, low-level errors that were encountered during the reaching movement could be easily mitigated by the visuomotor system. Hence, the correction of these low-level errors would not require the subsequent recruitment of the fronto-central error observed towards the latter part of the movement. Results of the motor related cortical potential component N4, yielded larger amplitudes for reaches without vision of the hand (i.e., Target only, No hand/No target). While for reaches with vision of the hand (Target & Hand, Hand only), the N4 was significantly smaller. However, what was also observed was that when individuals were reaching with vision of the hand, there was a slow shift in negative activity over occipital regions of the scalp, which occurred on sites contralateral to the direction that the hand. Furthermore, this negative shift in parietal-occipital activity peaked following movement end, suggesting that this activity may have had to do with visual feedback of the hand during the movement up until movement end for knowledge of result. Taken together, these findings lend support to the idea of a hierarchical error detection system (Krigolson et al., 2008). That is, low-level target and trajectory errors were detected and 150  amended by way of visual feedback. These findings are corroborated greater negativity activity was observed over parietal occipital regions, when vision of the hand was available. On the contrary, when vision of the hand was not available during the reaching movement, there was less activity observed over these regions. Instead, for these conditions, the presence of the larger N4 would indicate that the fronto-central error detection system was recruited to detect movement error and make improvements on subsequent movements. Overall, study 3 highlights that the P3 relating to target encoding was reflective of processes relating to the availability of vision the target and not limb vision during the reaching movement, which was similar to that of study 1. In addition, the second motor related cortical potential, namely component N4, was found to be influenced by different error detection systems (Krigolson & Holroyd, 2007) which are active based upon visual feedback availability of the hand (Heath, 2005).  5.2 Study Integration 5.2.1 Implications on Target Encoding Collectively, the manipulations in the studies above were designed to examine whether upcoming knowledge of vision of the target and hand alters the way individuals encode targets. In study 1 and 3, differences in P3 were observed when individuals were encoding the target in preparation for a movement without vision of the target. These findings were comparable to those from Krigolson et al., (2012) who used a similar task comparing visually guided and memory guided reaching. However, their explanation as to why the P3 was greater for NV vs. the FV condition was due to the generation of a motor plan during target encoding. This motor plan is stored in memory until the reach is initiated (Ghafouri, McIllroy & Maki, 2004; 151  Krigolson et al., 2012). Based on the logic that the P3 reflects motor planning processes, it was expected that P3 component prior to a reach without vision of the hand and target, would yield just as large, if not larger of a P3. This is because vision of hand is the most important source of information during a reaching movement, which allows for limb trajectory amendments (e.g., Heath, 2005). The results in study 3, however, showed that the P3 observed before reaches without vision of the target (i.e., no target, no hand/no target) remained the largest, while the P3 observed before reaches without vision of the hand (i.e., hand and target, hand only) were smaller. Therefore, it is thus possible that the modulation of P3 had to do with memory of the target location and not necessarily generating a motor plan when vision of the target or hand was not available (cf. Krigolson et al., 2012). Altogether, from the results of all three studies, the best explanation of the P3 can be put forth using the LC-NE adaptive gain theory. This theory posits that the P3 reflects a phasic increase in activity of the locus-coeruleus, a structure in the pons which releases NE to the cortex based on task demands (Ashton & Cohen, 2005; Nieuwenhuis, 2011). This phasic increase occurs in response to task demands, which increases the gain/responsivity of brain structures, which can be called upon to achieve the desired task outcome. In inferring the LC-NE theory to the results of this study, when individuals are encoding the target prior to a suboptimal upcoming reaching conditions (i.e., no vision of the target), the P3 indicates that, NE is released from the LC to projected brain structures, including areas for focused attention and memory (Aston-Jones, Meijas-Aponte & Waterhouse, 2010; Benarroch, 2009). In sum, the recruitment of these structures during the encoding of the target would help establish a memory representation of the target location. Therefore, when the individuals were informed to move by the imperative tone, they were able to rely on the memory representation of the target to execute their movements to. 152  Future studies could better elucidate the modulation of this P3 by not informing participants, or provide erroneous information about the vision condition on the upcoming reaching movement. 5.2.2 Implications on Movement Related Cortical Potentials The second event of interest was the execution of the reaching movement by the central nervous system. It has been suggested that the brain activity observed during movement execution are a result of two phases of neural excitability that stem from motor output from the primary motor cortex to the agonist and antagonist muscles of the reaching limb (Sergio & Kalaska, 1998). Indeed, recent EEG observations during movement execution revealed a scaling of the cortical potentials that coincide with the force demands of the reaching to different amplitudes (Kirsch & Hennighausen, 2010; Kirsch, Hennighausen & Rösler, 2010). Results of study 1 in the current dissertation also yielded larger MP and N4 components with greater movement amplitude, essentially replicating some of these previous findings. This was also corroborated by a significant correlation between peak amplitude of MP and N4 with the magnitude of peak acceleration and peak deceleration (Kirsch & Hennighausen, 2010; Kirsch, Hennighausen & Rosler; Krigolson et al., 2012). In addition, the manipulation of target size showed that reaching in FV to the small target yielded larger N4 amplitudes followed by the medium and large targets. The larger N4 amplitude was thought to reflect a greater amount of motor output required to undertake online corrections when reaching to the smaller and more difficult target, vs. the larger and easier target. Therefore, more support for the two phases of cortical excitability has been provided through study 1. The subsequent studies in this dissertation, however, showed that MP and N4 were also influenced by other factors, including visual and delay feedback schedule. 153   Study 2 showed that the component MP is affected by certainty of movement start. This was evidenced by the differences observed in component MP between the 2 and 5 delay period in the randomized, but not the blocked protocol. These differences were attributed to the differences in the negative activity (i.e., CNV) associated with anticipating the imperative tone to move. The CNV is a negative shift in brain activity that peaks shortly before the imperative tone (Walter et al., 1964) and can influence the motor related cortical potential, MP, associated with initiating the limb movement to the target. In the random protocol, it was suggested that the MP observed was lower for the 2 second vs. the 5 second delay period, which was attributable to the phenomenon known as ageing (Trillenberg et al., 2000). That is, when the delay period was just beginning, the participant had to consider that the imperative tone could occur at either 2 or 5 seconds. However, as the fore-period ‘aged’ by surpassing 2 seconds, the participant became certain of when the imperative tone would occur, as the only remaining option was the 5 seconds. As a result, it appeared individuals did not anticipate the 2 second delay period, as compared to the 5 second delay period. This was evidenced by a smaller MP in the 2 vs. 5 second delay. On the other hand, in the blocked protocol, participants were certain of the delay period on each trial and as a result, no differences in MP were observed.  Therefore, it appears as though component MP is susceptible to the effects of movement onset certainty.  With respect to component N4, the experiment with random protocol yielded differences between reaches in FV and NV. Specifically, the reaches in FV yielded greater N4 than reaches in NV, which is similar to the findings in study 1 and those of Krigolson et al., (2012). To complement this EEG finding, there was also a greater TaPV for trials in FV than in NV, hinting that the larger N4 may have to do with online feedback amendments.  While in the blocked protocol, differences in N4 were observed across delay condition. To complement this EEG 154  finding, higher PV was also attained for movements following a long delay period, thereby necessitating greater cortical motor output to control the limb as it neared the target. From a visual memory decay and limb control standpoint, it appears that in the long delay period, individuals relied on a stereotyped plan to initiate their limb movements under both visually guided and memory guided conditions. However, the control of the limb differed between the reaches in FV vs. NV, as the reaches in FV appeared to be effective minimizing error, whereas in NV it was not the case. In the final study, the effect of vision of the hand and target on the motor related cortical potential was examined.  In studies 1 and 2 of this thesis, visual feedback was manipulated by way of vision of the target. However, previous research has found that when vision of the hand is maintained but vision of the target is removed, the behavioural characteristics are largely similar to when vision is available (Heath, 2005). This is because when vision of the hand is available, there is still real-time vision of the hand, which permits visual access to the dorsal stream structures. As a result, amendments can be made to the limb trajectory leading to greater movement precision. Indeed when examining the motor related cortical potentials when reaching under varying conditions with vision of the hand and target, main effects for vision of the hand for the N4 component for fronto-central regions (i.e., electrode Fz) were found. Also, greater activity was observed over parietal-occipital regions that were contralateral to the visual field and reaching limb (i.e., electrode PO7). This provided further support for the idea of the dorsal stream’s involvement in integrating somatosensory and visual information when, and only when, real-time vision of the limb is available. Furthermore, under conditions where vision of the reaching limb was not available (i.e., target only, no hand/target) the N4 component was significantly larger. It is possible that when visual feedback of the limb is unavailable, 155  individuals rely instead on the fronto-central error detection system to gather information on how the movement was executed and to improve subsequent reaching movements (Krigolson & Holroyd, 2007).  In sum, with all studies taken into account, MP and N4 were shown to generally reflect the two phases of cortical motor output that occurs in the motor areas of the brain. However, given the variations in these potentials as a consequence of the conditions other than task difficulty, it shows that other processes such anticipation and visual feedback may influence how these potentials are manifested.  Figure 5.1. A schematic depicting the components of interest (P3, MP and N4) and the factors from each study that influenced the amplitude of these components.  156  5.2.3 Implications on Movement Planning and Control Inferring the findings from the studies above with processes underlying reaching movements, more information can be added onto the different steps that are involved in human limb control. Traditionally looking at the model (Figure 5.2) below by Elliott et al. (2010), the programming of a movement is based on the information made available prior to movement initiation (step 1). Based on our findings, this programming can vary with the involvement of the LC-NE system to involve other brain structures that may aid with the motor planning process (Ashton & Cohen, 2005; Nieuwenhaus, 2011). Notably, for reaches where vision of the target is not available (i.e., Krigolson et al., 2012) or reaches of greater difficulty (Kourtis et al., 2012), activation of the locus-coeruleus, and the release of NE to the brain which upregulates brain function so that other faculties such as memory can be used to aid in the preparation of the upcoming reaching movement. 157   Figure 5.2. Taken with permission from Elliott et al., (2010). This figure depicts the different neural processes that occur prior to and during the execution of a limb movement. Further along in Elliott’s model, step 2, there is a specification of the magnitude and timing of the muscular forces, formation of the efference and predicting the sensory consequences of the movement. With respect to the EEG components observed, this process coincides the increasing shift in negativity activity is observed over the fronto-central regions of the brain, which is when the beginnings of the motor plan are generated, as well as the anticipation of when the movement start occurs.   Finally as the imperative go signal occurs (step 3) there is a comparison between the internal model with the efference copy (i.e., less than 100 ms post movement initiation). This occurs with the first phase of cortical output (MP) the agonist of the limb muscles to contract and propel the limb towards the target. As the movement begins, forward modelling uses the efference copy to predict the outcome of the movement. From studies involving the ERN Target encoding (P3) Anticipation of the stimulus (CNV) 1st phase of cortical motor output (MP) 2nd phase of cortical motor output (N4) and high level detection processes when limb vision is not available  7) KR: when limb vision is available 158  (Krigolson & Holroyd, 2007), anterior cingulate cortex (ACC) is involved in comparing the two. If the comparison yields a predicted error, a corrective response would be made through impulse control of the limb. If this error cannot be corrected through impulse regulation, an ERN will occur. Further along, if visual feedback is available, step 5 takes place, as visual and proprioceptive feedback is gathered and processed (i.e., greater than 250 ms post movement initiation). Following the gathering and processing of visual and proprioceptive information, step 6 occurs, where amendments to the limb trajectory can be made in order to correct any low level target and/or trajectory errors. In addition a step 7 can be added, as knowledge of result, is obtained at the termination of the movement. As observed in study 3, vision of the hand resulted in activity in the parietal occipital areas, which was not present when vision of the hand is occluded. While for movements without vision of the hand or movements where the error could not be mitigated, a fronto-central FRN would manifest, reflecting activity from the anterior cingulate cortex and the processing of high level outcome errors. Regardless of the different modalities that these error detection systems use, both provide feedback to the motor system which allow for modifications to the motor plan to be made, so that subsequent movements can be improved upon.   159  Bibilography Aglioti S., DeSousa, J. F. X & Goodale, M. A. (1995). Size-contrast illusions deceive the eye but not the hand. Current Biology 5(6), 679–685. Aston-Jones, G., Meijas-Aponte, C. A., Waterhouse, B. (2010). Norepinephrine: CNS pathways and neurophysiology. In Fink, G. (Ed.). (2010). Stress Science: Neuroendocrinology (pp. 388-399). San Diego, Academic Press, Elsevier Ltd. Ball, T., Schreiber, A., Feige, B., Wagner, M., Lücking, C. H., & Kristeva-Feige, R. (1999). The role of higher-order motor areas in voluntary movement as revealed by high-resolution EEG and fMRI. Neuroimage, 10(6), 682-694. Bouret, S., & Sara, S. J. (2005). Network reset: a simplified overarching theory of locus coeruleus noradrenaline function. Trends in Neurosciences, 28(11), 574-582. Brunia, C. H. M. (1987) Brain potentials related to preparation and action. In H. Heuer & A. F. Sanders (Eds.), Perspectives on Perception and Action (pp. 105-130). Hillsdale, NJ: Erlbaum. Brunia, C. H., van Boxtel, G. J., & Bocker, K. B. (2011). Negative slow waves as indices of anticipation: the Bereitschaftspotential, the contingent negative variation, and the stimulus-preceding negativity. The Oxford handbook of event-related potential peaks. Beggs W. D., Howarth, C. I. (1972). The movement of the hand towards a target. Quarterly Journal of Experimental Psychology, 24, 448 453. 160  Binsted, G, Rolheiser, T. M., & Chua, R. (2006). Decay in visuomotor representations during manual aiming. Journal of Motor Behavior, 38(2), 82-87. Cant, J. S., Arnott, S. R., & Goodale, M. A. (2009). fMR-adaptation reveals separate processing regions for the perception of form and texture in the human ventral stream. Experimental Brain Research, 192(3), 391-405.  Carlton, L. G. (1992). Visual processing time and the control of movement. In L. Proteau and D. Elliott (eds.), Vision and Motor Control (pp. 3-31). Amsterdam: North-Holland. Chua, R. & Elliott, D. (1993). Visual regulation of manual aiming. Human Movement Science, 12(4), 365-401. Connolly, J. D., Andersen, R. A., & Goodale, M. A. (2003). FMRI evidence for a parietal reach region in the human brain. Experimental Brain Research, 153(2), 140-145. Cruikshank, L. C., Caplan, J. B., & Singhal, A. (2014). A perception-based ERP reveals that the magnitude of delay matters for memory-guided reaching. Experimental Brain Research, 232(7), 2087-2094. Dehaene, S., Posner, M. I., & Tucker, D. M. (1994). Localization of a neural system for error detection and compensation. Psychological Science, 5(5), 303-305. Desmurget, M., & Grafton, S. (2000). Forward modeling allows feedback control for fast reaching movements. Trends in Cognitive Sciences, 4(11), 423-431. Desmurget, M., & Grafton, S. (2003). Feedback or feedforward control: End of a dichotomy. Taking Action: Cognitive Neuroscience Perspectives on Intentional Acts, pp. 289-334. Di Russo, F., Antigona, M., Martin, S. I., Petzalis, S. & Hillyard, S.A. (2001). Cortical sources of the early peaks of the visual evoke potential. Human Brain Mapping, 15, 95-111. 161  Donchin, E., & Coles, M. G. (1988). Is the P300 component a manifestation of context updating. Behavioral and Brain Sciences, 11(3), 357-427. Elliott, D., & Allard, F. (1985). The utilization of visual feedback information during rapid pointing movements. The Quarterly Journal of Experimental Psychology, 37(3), 407-425. Elliott, D., Chua, R., Pollock, B. J., & Lyons, J. (1995). Optimizing the use of vision in manual aiming: The role of practice. The Quarterly Journal of Experimental Psychology, 48(1), 72-83. Elliott, D., Hansen, S., Grierson, L. E. M., Lyons, J., Bennett, S. J., & Hayes, S. J. (2010). Goal-directed aiming: Two peaks but multiple processes. Psychological Bulletin, 136, 1023-44. Elliott, D., Hansen, S., Mendoza, J., & Tremblay, L. (2004). Learning to optimize speed, accuracy, and energy expenditure: a framework for understanding speed-accuracy relations in goal-directed aiming. Journal of Motor Behavior, 36(3), 339-351. Elliott D., & Madalena J. (1987), The influence of premovement visual information on manual aiming. Quarterly Journal of Experimental Psychology A 39, 541-559. Falkenstein, M., Hohnsbein, J. H. J., & Blanke, L. (1991). Effects of crossmodal divided attention on late ERP peaks. II. Error processing in choice reaction tasks. Electroencephalography and Clinical Neurophysiology, 78, 447 - 455.  Fitts, P. (1954). The information capacity of the human motor system in controlling the amplitude of movement. Journal of Experimental Psychology 47, 381-391. Ghafouri, M., McIllroy, W.E. & Maki, B.E. (2004). Initiation of rapid reach-and-grasp balance reactions: is a pre-formed visuospatial map used in controlling the initial arm trajectory? Experimental Brain Research 155, 532-536. 162  Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends Neuroscience, 15(1), 20-25. Heath, M. (2005). Role of limb and target vision in the online control of memory-guided reaches. Motor Control 9, 281-309. Held, R., & Gottlieb, N. (1958). Technique for studying adaptation to disarranged hand-eye coordination. Perceptual and Motor Skills, 8(3), 83-86. Henry, F. M., & Rogers, D. E. (1960). Increased response latency for complicated movements and a “memory drum” theory of neuromotor reaction. Research Quarterly. American Association for Health, Physical Education and Recreation, 31(3), 448-458. Holroyd, C. B. (2004). In: A note on the N200 and the feedback ERN. In: Errors, conflicts, and the brain: Current opinions on performance monitoring (Ullsperger M, Falkenstein M, eds), pp 211–218.Leipzig: MPI of Cognitive Neuroscience. Holroyd, C. B. & Coles, M. G. H. (2002). The neural basis of human error processing: reinforcement learning, dopamine, and the error-related negativity. Psychological Review, 109, 679–709. Jeon, Y. W., & Polich, J. (2001). P3a from a passive visual stimulus task. Clinical Neurophysiology, 112(12), 2202-2208. Juckel, G., Karch, S., Kawohl, W., Kirsch, V., Jäger, L., Leicht, G., Lutz, J., Stammel, A. Pogarell, O., Ertl, M., Hegerl, U., Möller, H. J.,  & Mulert, C. (2012). Age effects on the P300 potential and the corresponding fMRI BOLD-signal. Neuroimage, 60(4), 2027-2034. 163  Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (Eds.). (2000). Principles of neural science (Vol. 4, pp. 1227-1246). New York: McGraw-Hill. Kida, T., Nishihira, Y., Hatta, A., Wasaka, T., Tazoe, T., Sakajiri, Y. & Sakamoto, M. (2004). Resource allocation and somatosensory P300 amplitude during dual task: effects of tracking speed and predictability of tracking direction. Clinical Neurophysiology, 115(11), 2616-2628. Kirsch W., & Hennighausen E. (2010). ERP correlates of linear hand movements: Distance dependent changes. Clinical Neurophysiology, 121(8), 1285-1292. Kirsch, W., Hennighausen, E., & Rösler, F. (2010). ERP correlates of linear hand movements in a motor reproduction task. Psychophysiology, 47(3), 486-500. Klapp S.T. (1975) Feedback vs. motor programming in the control of aimed movements. Journal of Experimental Psychology 104,147-153. Klem, G. H., Lüders, H. O., Jasper, H. H., & Elger, C. (1999). The ten-twenty electrode system of the International Federation. Electroencephalography Clinical Neurophysiology, 52, 3-6. Kourtis D., Sebanz N., & Knoblich G. (2012). EEG correlates of Fitts’s law during preparation for action. Psychological Research 76, 514-524. Kornhuber, H. H., & Deecke, L. (1965). Hirnpotentialänderungen bei Willkürbewegungen und passiven Bewegungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pflüger's Archiv für die gesamte Physiologie des Menschen und der Tiere, 284(1), 1-17. 164  Krigolson O., Bell J., Kent C. M., Heath M., & Holroyd C. B. (2012). Reduced cortical motor potentials underlie reductions in memory-guided reaching performance. Motor Control, 16, 353-370. Krigolson, O. E., Holroyd, C. B., Van Gyn, G., & Heath, M. (2008). Electroencephalographic correlates of target and outcome errors. Experimental Brain Research, 190(4), 401-411.  Krigolson, O. E., & Holroyd, C. B. (2006). Evidence for hierarchical error processing in the human brain. Neuroscience, 137(1), 13-17. Krigolson O., & Holroyd C.B. (2007). Hierarchical error processing: different errors, different  systems. Brain Research, 1155, 70-80. Lotze, M., Montoya, P., Erb, M., Hülsmann, E., Flor, H., Klose, U., Birbaumer, N. & Grodd, W.  (1999). Activation of cortical and cerebellar motor areas during executed and imagined hand movements: an fMRI study. Journal of cognitive neuroscience, 11(5), 491-501. Luck, S.J. (2005). An introduction to the event-related potential technique. Cambridge. MIT Press. Luck, S. J., & Hillyard, S. A. (1994). Electrophysiological correlates of feature analysis during visual search. Psychophysiology, 31(3), 291-308. Massion, J. (1992). Movement, posture and equilibrium: interaction and coordination. Progress in Neurobiology, 38(1), 35-56. Messier J. & Kalaska J.F. (1999). Comparison of variability of initial kinematics and endpoints of reaching movements. Experimental Brain Research 125,139-152. 165  Meyer, D. E., Abrams, R. A., Kornblum, S., Wright, C. E., & Keith Smith, J. E. (1988). Optimality in human motor performance: ideal control of rapid aimed movements. Psychological Review, 95(3), 340. Milner, A. D., Goodale, M. A., 1993. Visual pathways to perception and action. Progress in Brain Research 95, 317–337. Milner A. D., & Goodale, M. A., (1995). The Visual Brain in Action. Oxford University Press, Oxford. Naranjo, J. R., Brovelli, A., Longo, R., Budai, R., Kristeva, R., & Battaglini, P. P. (2007). EEG dynamics of the frontoparietal network during reaching preparation in humans. Neuroimage, 34(4), 1673-1682. Nieuwenhuis, S. (2011). Learning, the P3, and the locus coeruleus-norepinephrine system. In: R. Mars, J. Sallet, M. Rushworth, and N. Yeung (Eds.) Neural basis of motivational and cognitive control. Oxford University Press, p. 209-222. Nieuwenhuis, S., Aston-Jones, G., & Cohen, J. D. (2005). Decision making, the P3, and the locus coeruleus--norepinephrine system. Psychological Bulletin, 131(4), 510. Patel, S. H., & Azzam, P. N. (2005). Characterization of N200 and P300: selected studies of the event-related potential. International Journal Medical Sciences, 2(4), 147-154. Paus, T. S. (2001). Primate anterior cingulate cortex: Where motor control, drive and cognition interface. Nature Reviews Neuroscience, 2(6), 417-424. Plamondon, R. (1995). A kinematic theory of rapid human movements. Biological Cybernetics, 72(4), 295-307. 166  Plamondon R, Alimi AI (1997) Speed/accuracy trade-offs in target-directed movements. Behavioral and Brain Sciences 20, 279-303. Polich, J. (2007). Updating P300: an integrative theory of P3a and P3b. Clinical Neurophysiology, 118(10), 2128-2148. Praamstra, P., Kourtis, D., & Nazarpour, K. (2009). Simultaneous preparation of multiple potential movements: opposing effects of spatial proximity mediated by premotor and parietal cortex. Journal of Neurophysiology, 102(4), 2084-2095. Rizzolatti, G., & Gentilucci, M. (1988). Motor and visual-motor functions of the premotor cortex. Neurobiology of Neocortex, 42, 269-284. Samuels, E. R., & Szabadi, E. (2008). Functional neuroanatomy of the noradrenergic locus coeruleus: its roles in the regulation of arousal and autonomic function part II: physiological and pharmacological manipulations and pathological alterations of locus coeruleus activity in humans. Current Neuropharmacology, 6(3), 254-285. Schmidt, R. A, Zelaznik, H. N., Hawkins, B., Frank, J. S., and Quinn, J. T. (1979). Motor output variability: A theory for the accuracy of rapid motor acts. Psychological Review, 86, 415-451. Sergio L.E., & Kalaska J.F. (1998). Changes in the temporal pattern of primary motor cortex activity in a directional isometric force vs. limb movement task. Journal of Neurophysiology, 80, 1577-1583. Sperling, G. (1960). The information available in brief visual presentations. Psychological Monographs 74, 1-29. 167  Spehlmann, R. (1965). The average electrical responses to diffuse and to patterned light in the human. Electroencephology and Clinical Neuropsychology, 19, 560-569. Squires, N. K.; Squires, K. C.; Hillyard, S. A. (1975). Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. Electroencephalography and Clinical Neurophysiology 38(4): 387–401. Sutton, S., Braren, M., Zubin, J., & John, E. R. (1965). Evoked-potential correlates of stimulus uncertainty. Science, 150(3700), 1187-1188. Torrecillos, F., Albouy, P., Brochier, T., & Malfait, N. (2014). Does the processing of sensory and reward-prediction errors involve common neural resources? Evidence from a frontocentral negative potential modulated by movement execution errors. The Journal of Neuroscience, 34(14), 4845-4856. Trillenberg, P., Verleger, R., Wascher, E., Wauschkuhn, B., & Wessel, K. (2000). CNV and temporal uncertainty with ‘ageing’and ‘non-ageing’S1–S2 intervals. Clinical Neurophysiology, 111(7), 1216-1226. Ungerleider, L.G., Courtney, S.M. & Haxby, J.V. (1998). A neural system for human visual working memory. Proceedings from the National Academy of Science, 95, 883-890. Verleger, R., Jaśkowski, P., & Wascher, E. (2005). Evidence for an integrative role of P3b in linking reaction to perception. Journal of Psychophysiology, 19(3), 165-181. Viallet, F., Massion, J., Massarino, R., & Khalil, R. (1992). Coordination between posture and movement in a bimanual load lifting task: putative role of a medial frontal region including the supplementary motor area. Experimental Brain Research, 88(3), 674-684. 168  Wadman, W. J., Van Der Gon, J. D., Geuze, R. H., & Mol, C. R. (1979). Control of fast goal-directed arm movements. Journal of Human Movement Studies, 5(1), 3-17. Westwood D.A., Heath M. & Roy E.A. (2003). No evidence for accurate visual-motor memory: systematic and variable error in memory-guided reaching. Journal of Motor Behavior, 35, 127-133. Whiting, H. T. A., Gill, E. B., & Stephenson, J. M. (1970). Critical time intervals for taking in flight information in a ball-catching task. Ergonomics, 13(2), 265-272. Wolpert D. A. & Ghahramani Z. (2000). Computational principles of movement neuroscience. Nature Neuroscience 3, 1212-1217. Woodworth, R. S. (1899). The accuracy of voluntary movement. Psychological Review 3(3 suppl 13), 1-119.  Vaughan Jr, H. G., Costa, L. D., & Ritter, W. (1968). Topography of the human motor potential. Electroencephalography and Clinical Neurophysiology, 25(1), 1-10. Verleger, R., Jaśkowski, P., & Wascher, E. (2005). Evidence for an integrative role of P3b in linking reaction to perception. Journal of Psychophysiology, 19(3), 165-181. Yazawa, S., Ikeda, A., Kunieda, T., Ohara, S., Mima, T., Nagamine, T. & Shibasaki, H. (2000). Human presupplementary motor area is active before voluntary movement: subdural recording of Bereitschaftspotential from medial frontal cortex. Experimental Brain Research 131(2), 165-177. Zelaznik, H. N., Hawkins, B., & Kisselburgh, L. (1983). Rapid visual feedback processing in single-aiming movements. Journal of Motor Behavior, 15, 217–236 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0355397/manifest

Comment

Related Items