Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Auditory perceived continuity in cochlear implant listeners Panchyk, Halen 2012

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

Item Metadata

Download

Media
24-ubc_2012_fall_panchyk_halen.pdf [ 1.73MB ]
Metadata
JSON: 24-1.0073270.json
JSON-LD: 24-1.0073270-ld.json
RDF/XML (Pretty): 24-1.0073270-rdf.xml
RDF/JSON: 24-1.0073270-rdf.json
Turtle: 24-1.0073270-turtle.txt
N-Triples: 24-1.0073270-rdf-ntriples.txt
Original Record: 24-1.0073270-source.json
Full Text
24-1.0073270-fulltext.txt
Citation
24-1.0073270.ris

Full Text

Auditory Perceived Continuity in Cochlear Implant Listeners by Halen Panchyk  B.Ed., The University of Regina, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in  The Faculty of Graduate Studies (Audiology and Speech Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2012  © Halen Panchyk, 2012  Abstract Cochlear implants are generally considered the most successful of all sensory neural prostheses currently in use (Wilson and Dorman, 2008). Investigation of auditory perception with cochlear implants is important for developing effective and evidence-based approaches for intervention and management of profound hearing loss. Various phenomena of auditory perception have begun to be explored with cochlear implant users. However, perception of a phenomenon that allows listeners to perceptually restore the continuity of sounds that are partially masked or interrupted by other sounds (“auditory induction” or “auditory continuity”) has not yet been investigated in a group of listeners with cochlear implants. In the current study a group of 10 listeners with cochlear implants and 10 control listeners with normalhearing provided judgments on the continuity of a pure tone signal in the presence of four levels of a narrow-band noise masker. The group of listeners with cochlear implants reported perception of auditory continuity, but did so for lower levels of the masker when compared to the normal-hearing control group. A secondary experiment investigated simultaneous masking in listeners with cochlear implants using the same masker levels used in the continuity experiment. The cochlear implant group displayed effective masking at a lower level than the normal-hearing control group, the same level at which auditory continuity was perceived in the first experiment. The differences observed between the two groups may be attributable to the greater effects of masking resulting from poorer frequency resolution, lack of temporal fine structure information and reduced dynamic range for users of cochlear implants compared with listeners with normal-hearing.  ii  Preface This study was reviewed and approved by the Clinical Research Ethics Board of the University of British Columbia and the Providence Health Care Research Institute. The certificate number of the ethics certificate obtained is PCH REB H10-02849.  iii  Table of Contents Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iii Table of Contents ................................................................................................................... iv List of Tables ......................................................................................................................... vii List of Figures ....................................................................................................................... viii List of Abbreviations ............................................................................................................. ix Acknowledgements ................................................................................................................ xi Chapter 1: Introduction ........................................................................................................ 1 1.1 Overview and Purpose of the Study……………………..…………………………...1 1.2 Literature Review: Cochlear Implants and Auditory Continuity……….…………….2 1.2.1  Electrical Auditory Stimulation………………………………....……………..2  1.2.2  Perceiving Sounds With Cochlear Implants…………………………………...7  1.2.3  Masking in Listeners With Normal-hearing..…………………………………12  1.2.4  Masking in Listeners With Cochlear Implants.………………………………15  1.2.5  Auditory Continuity………………………………………………………….17  1.2.6  Requirements for the Perception of Auditory Continuity……………………21  1.3 Objectives of the Study……………………………………………………………...25 1.4 Predicted Results and Outcomes…………………………………………………….25 1.4.1 Predicted Outcomes for Experiment 1: Auditory Continuity….....……………26  iv  1.4.2 Predicted Outcomes for Experiment 2: Simultaneous Masking………….…….27 1.4.3  Correlation Analysis……………………………………………………………29  Chapter 2: Methods ............................................................................................................. 31 2.1 Participants ................................................................................................................. 31 2.2 Stimuli ........................................................................................................................ 34 2.2.1 Experiement 1: Auditory Continuity……………………………………………35 2.2.2 Experiement 2: Simultaneous Masking…………………………………………39 2.3 Equipment .................................................................................................................. 40 2.4 Procedure ................................................................................................................... 41 2.4.1  Experiement 1: Auditory Continuity…………………………………………41  2.4.2  Experiement 2: Simultaneous Masking………………………………………43  2.5 Data Analysis ............................................................................................................. 44 2.5.1  Experiement 1: Auditory Continuity…………………………………………44  2.5.2  Experiement 2: Simultaneous Masking………………………………………45  Chapter 3: Results................................................................................................................ 46 3.1 Continuity Analysis ................................................................................................... 46 3.2 Simultaneous Masking Analyis ................................................................................. 48 3.3 Correlational Analysis and Individual Performance .................................................. 50  v  Chapter 4: Discussion .......................................................................................................... 55 4.1 Summary .................................................................................................................... 55 4.2 Perceived Continuity by Listeners With Cochlear Implants...................................... 56 4.3 Simultaneous Masking Performance by Listeners With Cochlear Implants ............. 60 4.4 The Relationship Between Auditory Continuity & Simultaneous Masking for Listeners with CIs ............................................................................................................................... 63 4.5 Clinical Implications .................................................................................................. 65 4.6 Limitations ................................................................................................................. 66 4.7 Future Research ......................................................................................................... 70 4.8 Conclusions ................................................................................................................ 75 References .............................................................................................................................. 77 Appendices ............................................................................................................................. 92 Appendix A: Collected data by listener Exp. 1: Continuity ............................................... 92 Appendix B: Collected data Exp. 2: Simultaneous Masking .............................................. 95  vi  List of Tables  Table 1  Individual gap detection thresholds and group averages ....................................... 33  Table 2 Collected participant information ........................................................................... 34 Table 3  Rated continuity for the cochlear implant group and control group ....................... 46  Table 4 Pearson’s r correlations for auditory continuity and simultaneous masking ......... 51 Table 5 Pearson’s r correlations for random gap detection and auditory continuity .......... 54 Table 6 Performance means for sub-groups of listeners with cochlear implants with no ADRO and with ADRO ...................................................................................................... 67  vii  List of Figures  Figure 1  Signal resulting in perception of illusory auditory continuity ............................... 18  Figure 2  Single A1 nerve fiber responses to a continuous tone, a discontinuous tone and a  discontinuous tone with a masker imposed during the interruption ........................... 24 Figure 3  Four stimulus conditions (continuous signal, continuous with masker,  discontinuous signal, discontinuous with masker)...................................................... 35 Figure 4  Curtailed stimulus condition.................................................................................. 36  Figure 5 Discontinuous signal with 49 dBA masker and onset/ offset amplitude ramps ..... 37 Figure 6  Simultaneous masking stimulus trial with warning tone ....................................... 40  Figure 7  Proportions of rated continuity to physically continuous and discontinuous stimuli  at 40 dB, 49 dB and 55 dB masker levels. .................................................................. 48 Figure 8  d’ scores for the simultaneous masking task at 40 dB, 49 dB and 55 dB masker  levels ........................................................................................................................... 49 Figure 9  Continuity ratings vs sensitivity d’ scores for the CI and control groups with the  49 dBA masker .......................................................................................................... 52 Figure 10  Continuity ratings vs sensitivity d’ scores for the CI and control groups with 40, 49  & 55 dBA masker levels. ............................................................................................ 53  viii  List of Symbols and Abbreviations (A1) Primary auditory cortex (AB) Advanced Bionics (ABR) Auditory brainstem response (ADRO) Adaptive Dynamic Range Optimization (ANF) Auditory nerve fibres (ANOVA) Analysis of variance (BM) Basilar membrane (CF) Centre frequency (CI) Cochlear implants (DB) Decibels (DBA) Decibels–A (A-weighted scale) (DX) Diagnosis (E) Electrode (E-CAP) Electrically evoked compound action potential (HSD) Honestly significant difference (HZ) Hertz (IPhD) Interaural phase difference (ITD) Interaural time difference (IX) Implantation (KHZ) Kilohertz (MS) Millisecond (MANOVA) Multivariate analysis of variance (NH) Normal-hearing (OHC) Outer hair cells ix  (PPS) Pulses per second (RGD) Random gap detection (RGDT) Random gap detection test (SD) Standard deviation (SM) Simultaneous masking (TFS) Temporal Fine Structure (ULC) Upper level of comfort (USM) Upward spread of masking (1I-2AFC) Single interval two alternative forced choice (2I-2AFC) Two-interval, two-alternative forced choice   Threshold  x  Acknowledgements I would like to express my sincere gratitude to my supervisor, Dr. Valter Ciocca for his patience and guidance and for encouraging me to think in a more critical and detailed manner. I would also like to thank my other committee members, Dr. Lorienne Jenstad and Dr. Sipke Pijl for their time, thoughtful input and help with participant recruiting. Thank you to Grace Shyng, Cindy Gustin and Dr. Brian Westerberg for their help in recruiting participants. A special thanks to Michael Currie for his relentless assistance in recruiting. Thank you to Dr. Nick Haywood and Andrew Vandali for their collaboration during experimental design and feedback throughout the project. I would also like to thank my family and especially my husband, Travis, for their unconditional support and encouragement throughout my years of education. This research was supported in part by a post-graduate research award from the Canadian Institutes of Health Research.  xi  Chapter 1: Introduction 1.1  Overview & Purpose of the Study In recent years, cochlear implantation has become a routine management strategy for a  broad range of individuals with both congenital and acquired hearing loss. Despite welldocumented benefits and favourable outcomes for many cochlear implant (CI) recipients, current literature and clinical anecdotal evidence suggests that both pre-lingually and post-lingually deafened CI recipients experience a variety of challenges in auditory and speech perception when compared with normal-hearing listeners (Zeng, 2004; Bergeson, Pisoni & Davis, 2005; Wilson & Dorman, 2008; Peterson, Pisoni & Miyamoto, 2010). The goal of this research project is to improve our understanding of what these specific challenges are with regards to auditory perception skills. This study will focus on the ability to perceptually restore partially-masked sounds, usually referred to as “auditory continuity”, or the “illusory auditory continuity” phenomenon (Warren, Obusek & Ackroff, 1972; Bregman & Dannenbring, 1977; Bregman, 1990). Auditory continuity is thought to act as perceptual compensation for the effects of masking, effectively helping people to listen to important sounds in everyday noisy environments (Warren et al, 1972). Among the auditory perceptual difficulties experienced by many listeners with CIs is the difficulty listening to important sounds, particularly speech, in these noisy environments. Difficulties with perception of speech in noise may result in increased equipment costs due to the need for devices which improve signal to noise ratios, refusal of implantation or non-use following implantation due to lack of perceived benefit (Ray et al., 2006). These difficulties may also have the potential to lead to increased social isolation or anxiety due to communication difficulties (Knapp, 1948; Mulrow et al., 1990; Chisolm, Abrams & McArdle, 2004). 1  Improved understanding of the perceptual processes that contribute to or inhibit auditory perception in noise for listeners with CIs is paramount in improving the ability of cochlear implants to provide more access to perception of sounds in every day environments. The study described in the following chapters investigated the perception of auditory continuity in listeners with CIs to determine if they, as a group, perceive auditory continuity, and how their perception compares with a normal-hearing control group. The relationship between the auditory continuity percept and auditory masking was also investigated in a secondary experiment. Chapter 1 will review the existing literature about auditory perception in listeners with CIs and about the perception of auditory continuity in normal-hearing listeners. These two areas of research can provide the background to inform the study of auditory continuity in listeners with CIs. Chapter 1 will conclude with specific objectives and predictions of the study. A description of the methods used to conduct the study will comprise chapter 2. The results will be presented in chapter 3. Chapter 4 will include a discussion of the results and of their implications, as well as proposed directions for future research. This chapter will also provide concluding remarks based on the study. 1.2  Literature Review: Auditory Continuity and Cochlear Implants 1.2.1  Electrical Auditory Stimulation  Cochlear implants transmit sound electrically to the auditory nerve through an array of electrodes inserted into the cochlea. The incoming auditory signal is picked up by the microphone at ear level, then analysed and converted to an electrical signal by the external speech processor. The electrical signal is then transmitted by transcutaneous radio frequency to the implanted receiver/ stimulator which then activates the electrodes within the cochlea  2  accordingly (Wilson & Dorman, 2008). Modern implants include an array of electrodes (the exact number of electrodes varies depending on manufacturer) and a digital speech processor with memory for multiple programs, settings and features for different listening situations. Some cochlear implant systems, such as those produced by Cochlear Ltd. employ speech processing strategies that limit the number of active electrodes per stimulation cycle to a subset of the available electrodes. For example, the Advanced Combinational Encoder strategy, ACE (Vandali, Whitford, Plant & Clark, 2000; Skinner et al., 2002) limits the active maxima to 8 to 12 of 20 or 22  possible electrodes. There are many differences between acoustic hearing and hearing with a cochlear implant. With a cochlear implant the incoming acoustic signal is converted to a digital signal by the speech processor, and then it is transmitted to the auditory nerve as a sequence of electrical pulses through the electrodes. By dividing the incoming signals into a series of band-pass filtered channels and then allocating these channels to the multiple electrodes inserted into the cochlea, multi-channel cochlear implants attempt mimic the tonotopic organization of the hair cells present in a normally hearing cochlea. With normal acoustic hearing, sound energy travelling along the basilar membrane (BM), ranging from approximately 20 Hz to 20,000 Hz, is transduced by a single row of approximately 3,500 inner hair cells, and amplified by three rows of approximately 12,000 outer hair cells (Engstrom & Sjostrand, 1954; Duval, Flock & Wersall, 1966; Yost, 2007). By contrast, the frequency “selectivity” of auditory channels for listeners with CIs is guided by the neural populations excited by the relatively few electrodes placed along the cochlea (Shannon, 1983a; Kwon & Van Den Honert, 2009). With a multi-electrode cochlear implant, sound is transduced by an electrode array typically consisting of 12 to 22 electrodes (depending on manufacturer and model) and these electrodes receive signals from the speech  3  processor which analyses the range of frequencies considered important for speech perception, 200 Hz to 8000 Hz. (Shannon, 1983a; Wilson & Dorman, 2008; Kwon & Van Den Honert, 2009). In addition to differences in transduction of sound within the cochlea, the cochlear mechanics which result from the activity of outer hair cells (OHC) that are normally present in acoustic hearing are also absent in electrical auditory stimulation. In normal-hearing, OHCs provide a sharpened tuning to the response of the basilar membrane. The “tuning” of the frequency response in electrical auditory stimulation is dependant only on the electrode activity and spacing, and on the placement of the electrodes relative to their target nerve fibres (Shannon, 1983a & 1983b). The frequency response of a common 22-electrode array operates effectively as a series of symmetrical band-pass filters (Glass, 1983). Due to overlap between frequencies of adjacent band-pass filters in the speech coding strategy, some controlled stimulation of electrodes adjacent to the stimulating electrode arises. For example, with the 22-electrode Nucleus (Cochlear Ltd) electrode array, when an incoming signal stimulates one electrode at its centre frequency (CF), directly adjacent electrodes will also be stimulated at a level approximately 6 dB lower (Vandali et al., 2000; Clark, 2003; Vandali, 2011). The range and resolution of detectable frequencies for a cochlear implant listener is dependent on many factors such as the number of active electrodes and insertion depth, but also on auditory nerve fibre survival and cochlear ossification (Chatterjee & Shannon, 1998; Nelson, Donaldson and Kreft, 2008; Kwon & Van Den Honert, 2009). Frequency range and selectivity are further influenced by the spread of electrical stimulation across the cochlea. Interaction between adjacent electrodes within the cochlea is a common artefact of CI electrode stimulation, and can effectively reduce the number of “separate” frequency channels through which auditory input is conveyed to the auditory nerve (Shannon, 1983b; Shannon, 1985; Wilson et al., 1991; 4  Cohen, Richardson, Saunders & Cowan, 2003). Due to the lack of current channel independence resulting from such factors as current interaction and summation, and spread of excitation of auditory nerve fibres, it is estimated that only four to eight independent frequency sites are effectively available for stimulation at any one time, even with implants that have as many as 22 electrodes (Fishman, Shannon & Slattery, 1997; Friesen, Shannon, Baskent & Wang, 2001; Wilson & Dorman, 2008). Dynamic range in terms of loudness (between threshold and maximum comfortable loudness level) has also been shown to be smaller and more variable than that of normal-hearing listeners. Furthermore, the number of discriminable loudness steps available to listeners with cochlear implants has been shown to vary independently of the size of their individual dynamic range (Nelson, Schmitz, Donaldson, Veimeister & Javel, 1996). The frequency selectivity and dynamic range (of frequency and loudness) available to listeners with cochlear implants is therefore limited (relative to acoustic hearing), highly variable across individuals, and dependant on several factors related to both the device itself and the individual implant recipient. Many techniques are employed to attempt to deal with the limitations of electric hearing. These methods include the use of multiple electrodes, different stimulation modes, variations in pulse widths and stimulation rates. With electrical stimulation of the cochlea, current is generated between two electrodes, an active electrode and a reference electrode. When the two electrodes are spaced widely apart (i.e. “monopolar stimulation” where the reference electrode is located outside of cochlea), greater current spread and channel interaction may occur, but the current level required to produce auditory percepts decreases (Chatterjee, 1999). “Bipolar” stimulation uses another intra-cochlear electrode as the ground for the active electrode, thereby narrowing the current focus and limiting channel interaction but increasing activation thresholds 5  (Pfingst, Zwolan & Holloway, 1997; Chatterjee 1999). Currently, it is difficult to estimate the amount of current spread within the cochlea for an individual listener. The narrower current focus afforded by bipolar stimulation may be offset by the increase in current required to produce auditory percepts, which can itself cause widening of the current field. Current interactions are further reduced by positioning the electrode array closer to the modiolus (the central axis within the cochlea containing the spiral ganglion of the auditory nerve). Electrophysiological evidence including electrically evoked auditory brainstem responses (ABR), responses from single fibres from the primary auditory cortex (A1), and electrically evoked compound action potentials (E-CAP) has shown that electrode arrays placed closer to the medial wall of the cochlea, and therefore closer to their neural targets, result in lower activation thresholds and less current interaction than arrays placed in the centre of the scala tympani or towards the lateral wall (van den Honert & Stypulkowski, 1987; Shepherd, Hatsushika & Clark, 1993; Cohen et al., 2003). Another commonly-used method for alleviating the current interaction that occurs between simultaneously activated electrodes is non-simultaneous (or “temporally interleaved”) stimulation, where electrodes are activated in rapid succession instead of simultaneously. While temporally interleaved stimulation of electrodes has been shown to result in improved speech recognition over simultaneous stimulation, refractory interaction (resulting from supra-threshold stimulation of a nerve fibre which elevates the threshold for another stimulus presented in very short succession) and facilitative effects resulting from residual polarization of the nerve membrane can still occur in nonsimultaneous stimulation (Wilson et al, 1991; Stickney et al., 2006; Bonnet, Boermans, Avenarius, Briaire & Frijns, 2012).  6  It has been proposed that higher stimulation rates can improve the temporal information available to listeners with CIs, resulting in better speech recognition (Wilson, Lawson, Zerbi, Finley & Wolford, 1995; Frijns et al., 2003). However, there is contrary evidence that higher rates of stimulation do not necessarily result in improved speech recognition for all listeners (Vandali et al, 2000; Holden et al, 2002). For example, Holden et al. (2002) found that increasing the pulse rate from 720 pulses per second (pps) to 1800 pps under non-simultaneous stimulation resulted in increased benefit for some listeners and in decreased speech recognition for others. Evidence from animal models also suggests that increased pulse rates can result in increased channel interaction (Middlebrooks, 2004). Although the previously described methods are in use to decrease current channel interaction in electrical cochlear stimulation, channel independence continues to present a significant concern in the design and use of cochlear implants due to the technological limitations of electrical auditory stimulation (Wilson & Dorman, 2008; Bonnet et al., 2012; Fredelake & Hohman, 2012). 1.2.2 Perceiving Sounds with Cochlear Implants It has been widely accepted that cochlear implantation can result in significant, and often invaluable, gains in auditory perception, and can contribute to the acquisition or maintenance of spoken language for profoundly hearing impaired individuals (for a review, see Zeng, 2004; Wilson & Dorman, 2008). Most listeners with CIs undergo considerable improvement following implantation, when their post-implant speech recognition abilities are measured against their preimplant performance. For example, although general implantation criteria vary according to service program, location and individual case factors, in most cases an adult CI candidate must meet a clinical criterion of 60% or less correct for open-set sentence recognition with optimal acoustic amplification, and may perform as low as 0% correct prior to implantation (Clark, 2003; 7  Zeng, 2004). After implantation, the majority of post-lingually deafened recipients exhibit performance above 80% correct in open-set sentence recognition in a controlled, quiet environment (National Institutes of Health, 1995; Clark, 2003; Waltzman & Thomas, 2006; Wilson & Dorman, 2008). Thanks to the input provided through cochlear implants, recipients are often able to form or maintain connections with the hearing world in cultural, educational and vocational domains, opportunities that may be otherwise largely unavailable to individuals with profound hearing loss. However, it has also been well-documented that outcomes of implantation are highly variable, and that recipients can still experience a range of significant challenges in speech and auditory perception, when compared with their normal-hearing peers (Tyler et al., 1984; Chatterjee & Shannon, 1998; Clark, 2003; Waltzman & Thomas, 2006; Wilson, 2006; Wilson & Dorman, 2008; Peterson et al., 2010). Both similarities and differences in the way listeners with CIs and normal-hearing listeners respond to temporal characteristics of auditory stimuli have been reported. Shannon (1989) investigated temporal resolution in multi-electrode listeners with CIs, and found that gap detection thresholds were comparable to those of normal-hearing listeners, independent of electrode placement within the cochlea or distance between active electrodes. As in normalhearing, temporal resolution has been shown to be an important aspect of perceiving speech with a cochlear implant. Muchnik (1993) investigated temporal gap detection thresholds and open set speech recognition ability in a group of 14 multiple electrode listeners with CIs. He found a considerable amount of variability between listeners with CIs on both tasks. Most crucially, his results demonstrated a significant correlation between performance in the random gap detection test and in the speech recognition test, with listeners displaying better open set speech recognition also having lower temporal gap detection thresholds. Unlike temporal resolution  8  (measured as gap detection), important differences between normal-hearing listeners and listeners with cochlear implants have been found in tasks measuring the use of temporal fine structure cues (the fast variations in the amplitude of the signal). Temporal fine structure is not perceivable with current cochlear implants beyond approximately 300 Hz. (Shannon 1983a & 1986; Zeng, 2002). As a consequence, temporal fine structure is discarded by modern CI strategies and only temporal envelope (the larger pattern of amplitude fluctuations in a signal over time) is coded by the speech processor. These temporal processing differences between listeners with cochlear implants and normal-hearing listeners may have important effects on how the stimuli in the current study is perceived and will be discussed further in following sections. Differences in perception of temporal integration have also been found (Shannon, 1983a, 1986, 1990; Pfingst, De Haan & Holloway, 1991; Donaldson, Viemeister & Nelson, 1997). For example, Donaldson et al. (1997) found a much shallower and highly variable integration slope (the function that describes the increase in loudness as a function of increased duration of a signal) in a group of listeners with multiple electrode CIs when compared with a control group of normal-hearing listeners. The mean integration slope for the CI group was 0.42 dB/doubling of duration with a standard deviation of .38 dB. This slope is very shallow compared with the 2.5 dB/doubling consistently found in normal-hearing listeners (Gerken, 1990, Donaldson et al., 1997). While the differences in temporal integration between normal-hearing listeners and listeners with cochlear implants have been consistently observed, it is unclear how these differences may affect the perception of the stimuli in the current study. For this reason this difference between listeners with cochlear implants and those with normal hearing will not be discussed further.  9  One of the most critical and generally cited difficulties in listening with cochlear implants is the perception of speech in noisy environments (Tyler et al., 1984; Fu, Shannon & Wang; 1998; Friesen et al., 2001; Stickney, Zeng, Litovsky & Assman, 2004; Nelson & Jin, 2004). In order to reach the same level of speech recognition performance as normal-hearing listeners, listeners with CIs have been shown to require considerably higher signal to noise ratios (Dorman, Loizou & Tu, 1998; Nelson, Jin, Carney &Nelson, 2003; Nelson & Jin, 2004; Stickney et al 2004). This difficulty is likely due in part to a lack of spectral detail in the signal processed through a CI, caused by the signal processing filters used for each frequency channel, the small number of frequency channels, reduced dynamic range and spread of excitation within the cochlea (Chatterjee & Shannon, 1998; Fu et al., 1998; Friesen et al., 2001; Nelson et al., 2003; Nelson & Jin 2004; Stickney et al., 2004; Ricketts et al., 2006; Stickney et al., 2006; Cooper & Roberts, 2009; Chatterjee, Peredo, Nelson & Baskent, 2010). For example, Fu et al. (1998) reported a significant decline in the ability to recognise phonemes in the presence of noise for listeners with CIs using a continuous interleaved sampling (CIS) processing strategy (Wilson et al, 1991). The performance of listeners with cochlear implants was significantly worse than the performance of normal-hearing listeners who heard the same, spectrally unaltered speech stimuli in noise; the performance of the listeners with CIs was also significantly worse for speech in noise than in quiet. Furthermore, when the normal-hearing controls listened to CIS-like processed speech stimuli under the same noise conditions, they demonstrated a similar decline in recognition. The similarity in performance between the listeners with CIs and normal-hearing listeners in the processed speech condition suggests that the lack of spectral resolution in speech processed through cochlear implants is a significant contributor to the difficulty in noise demonstrated by listeners with CIs. Another factor that is likely to affect the perception of  10  speech in noise with cochlear implants is the lack of temporal fine structure information, as evidenced by studies investigating perception of speech in noise by listeners with hearing impairment, who also lack use of temporal fine structure cues (Lorenzi, Gilbert, Carn, Garnier & Moore, 2006). Auditory scene analysis (ASA) is a term used to denote the ability to perceive sounds in the presence of other sounds through grouping and segregation of concurrent auditory stimuli within the environment (Bregman, 1990). While ASA has been studied in detail in normal hearing listeners, some aspects of auditory scene analysis have begun to be investigated in listeners with cochlear implants as well (Cooper & Roberts, 2007; 2009; Oxenham, 2008; Carylon, Long, Deeks & McKay, 2007). These studies, focusing on the segregation of concurrent streams of sounds (stream segregation), have generally concluded that listeners with CIs lack the ability to segregate streams of concurrent sounds as observed in normal hearing listeners. For example, Cooper and Roberts (2007) investigated stream segregation in eight cochlear implant listeners using two tones designed to stimulate differing electrodes at their centre frequencies. The separation between the stimulated electrode CFs and the timing of the tones was varied. The percepts reported by the listeners varied predictably with increasing electrode separation and, unlike normal hearing listeners, timing of the tones had no effect. These results lead the investigators to conclude that the percepts reported by the listeners with cochlear implants may reflect channel discrimination rather than auditory stream segregation. In order to lessen some of the difficulties in auditory perception with cochlear implants, speech processing features are continuously being developed and improved upon to provide increased speech intelligibility, access to low-level ambient sounds and listening comfort in the presence of loud sounds for users of cochlear implants. Features that adjust the gain of incoming  11  sounds in relation to the listener’s threshold of audibility or comfort level are commonly added to their everyday listening programs. They are also used to create programs for use in special situations, such as particularly noisy environments or listening to music. A relatively new feature which is being implemented as a default in many speech processor models is Adaptive Dynamic Range Optimization (ADRO) (James et al., 2002; Blamey, 2005). The purpose of ADRO is to provide increased intelligibility of speech, particularly at low input levels, by continuously adapting the dynamic range of the incoming acoustic signal to the effective dynamic range of the electric output signal. ADRO uses channel specific adaptive gain to gradually adjust output from the speech processor within individual frequency bands in relation to the dynamic range of the listener. In other words, if the level of a sound falls below the listener’s audibility threshold, gain is applied and if a sound rises above the listener’s comfort level, it is reduced. Sounds that fall within the listener’s dynamic range are also subject to gain rules designed to prevent background noise from maintaining too high a high level in relation to the listener’s dynamic range (James, 2002). ADRO has been shown to improve speech perception for Listeners with CIs in both quiet and noise (James, 2002; Blamey, 2005). The effort to develop methods of improving perception noise with cochlear implants is ongoing, as is the attempt to better understand the underlying causes of these difficulties. As the masking of important sounds such as speech is a particularly common occurrence in everyday listening environments, the effects of masking on the perception of sounds have been extensively investigated in both acoustic and electrical hearing. 1.2.3 Masking in Normal-hearing Listeners In natural environments, it is uncommon for sounds to occur in isolation. Usually sounds occur in conjunction with other sounds, and can interfere with one another. When humans are 12  exposed to multiple sounds under these conditions, the presence of one sound may impair the perception of another sound, resulting in masking. Auditory masking is the phenomenon by which the threshold of audibility for one sound is raised because of the interfering presence of another sound (American Standards Association, 1960; for early studies of masking, see Wegel & Lane, 1924; Egan & Hake, 1950). Masking is a significant component of everyday listening environments and it can result from either steady-state sounds (such as air conditioner noise) and/or fluctuating sounds (such as competing speech). Masking can be separated into two basic types: simultaneous and non-simultaneous (or “temporal”) masking. Simultaneous masking refers to the masking effect of a sound (the masker) that occurs at the same time as the sound of interest (the signal). Non-simultaneous masking occurs when the masker is presented either before the signal (referred to as forward masking) or after the signal (referred to as backward masking). For normal-hearing listeners, detecting sounds in the presence of maskers follows largely predictable masking patterns for both types of masking. These patterns are affected by the frequency and the level of both the signal and masker. For example, maskers have been shown to be more effective when their level is higher than the level of the signal, all other conditions being equal. As the masker level increases relative to the level of the signal, the “masking threshold” (the level of the signal at which it is just audible in the presence of a masker) becomes higher (Egan & Hake, 1950; Wegel & Lane, 1924). The amount of masking has also been shown to be larger when the frequency of the masker is similar to the frequency of the signal, all other conditions being equal. For example, speech noise represents an especially common and effective masker in everyday listening environments, because it is similar in spectral content to the speech signal that listeners are trying to focus on (Darwin, 2008). Conversely, as the difference between the frequency of the signal  13  and that of the masker increases, the masker level required to reach the masking threshold also increases (Egan & Hake, 1950; Wegel & Lane, 1924; see also, Moore, 2004, for review). When listening to sounds in the presence of maskers, normal-hearing listeners demonstrate a non-linear pattern of response as a function of frequency. Experiments have demonstrated that when the masker frequency is held constant and the signal frequency is varied, the masker is more effective at masking higher frequency than lower frequency signals (Wegel & Lane, 1924; Egan & Hake, 1950). This phenomenon is referred to as the “upward spread of masking” (Moore, 2004; Yost, 2007). An upward spread of masking (USM) can be observed under both simultaneous and non-simultaneous masking conditions. However, the USM effect is not as strong when observed under non-simultaneous masking conditions, due to the different underlying mechanisms in simultaneous and non-simultaneous masking. Suppression, the reduction in response of auditory nerve fibres (ANF) to one stimulus due to the simultaneous presentation of another stimulus, is thought to be the primary mechanism underlying the effects of simultaneous masking (Houtgast, 1972; Moore, 1978; Rodriguez et al, 2009). Forward masking is thought to be a result of ANF adaptation, the reduction in ANF response to a stimulus presented immediately after a previous stimulus (Houtgast, 1972; Moore 1977; Rodriguez et al, 2009). Both suppression and adaptation result in a reduction of ANF response compared with the response elicited by a stimulus presented in isolation. However, there is a larger amount of USM under simultaneous masking conditions because suppression produces a markedly larger reduction in ANF response than adaptation (Delgutte, 1990; Rodriguez et al, 2009; Yasin & Plack, 2005). As described previously, for both types of masking the masking threshold is high (i.e., greater masking occurs) when the frequency of the masker is similar to the frequency of the  14  target. Due to this frequency-dependent response, masking thresholds are often used as a measure of the frequency selectivity of the auditory system. Aspects of frequency selectivity can be categorised by psychophysical tuning curves. These curves plot the level of masker necessary to mask a fixed-frequency signal as a function of masker frequency (Fletcher, 1940; Moore, 1978; Glasberg & Moore, 1990; Moore 2004). Psychophysical tuning curves can be obtained using either simultaneous or non-simultaneous masking, but have been found to be sharper for forward masking than for simultaneous masking, due to the lack of suppression in nonsimultaneous masking (Houtgast, 1972, Moore 1977). Another important, and perhaps more obvious, aspect of masking that differs between simultaneous and non-simultaneous masking is the effect of timing. In order for simultaneous masking to take place, the masker and signal must occur (by definition) at the same point in time. For non-simultaneous masking, the effect of the masker on the signal differs as a function of the temporal gap between them and as a function of the duration of the masker (Moore & Glasberg, 1983). The amount of masking for a forward masker has been shown to increase with increasing masker duration up to 200 ms (Kidd & Feth, 1982; Zwicker, 1984). A non-simultaneous masker is also known to be more effective in masking the signal as the temporal gap between them becomes smaller. This is true for both forward and backward masking. However, forward masking is more effective at greater temporal gaps than backward masking. Forward maskers are effective at masking a signal with a temporal difference up to about 100 ms, whereas backward maskers are effective only up to 20 ms or so after the signal has occurred (Moore & Glasberg, 1983; Yasin & Plack, 2005; see also Yost, 2000; Moore, 2004 for reviews). As discussed briefly in this section, the characteristics of auditory masking in normalhearing listeners are complex but do follow predictable and understandable patterns. The  15  conditions under which masking takes place and the patterns of masking that result in normalhearing listeners helps to provide a baseline against which to compare masking phenomena in Listeners with CIs. 1.2.4  Masking in Listeners with Cochlear Implants  While masking also occurs in hearing with cochlear implants, there are a number of differences in the mechanisms of masking between acoustic and electric hearing. As discussed previously, the physiological tuning properties of the basilar membrane are absent in electronic hearing, resulting in important differences in frequency selectivity and dynamic range between electric and acoustic hearing. The frequency selectivity and dynamic range available to individual Listeners with CIs are dependent on many factors and are highly variable. For non-simultaneous masking, both listeners with cochlear implants and normal-hearing listeners generally demonstrate a 10 to 20 millisecond (ms) backward masking effect and a 100 to 200-ms forward masking effect (Shannon, 1990). However, it is difficult to directly compare simultaneous masking in electric hearing and acoustic hearing because of the differences in frequency response and selectivity between acoustic and electric hearing and because of the effects of current interaction and spread of excitation within the cochlea. Current summation interactions can be caused by the interaction of electrode current fields during simultaneous or closely temporally interleaved activation (Shannon, 1983b; Shannon, 1985). These interactions can have compounding refractory or faciltatory effects, stimulating the auditory nerve fibres at an intensity and across a frequency range, that differ from the stimulation that would be provided by fully independent electrode current fields (Shannon, 1983b; Shannon, 1985; Stickney et al. 2006). Such interactions can result in a masker that may have a greater effect both in level and across frequencies in comparison with normal-hearing listeners. These effects may contribute to 16  variability or difficulty in auditory processing and in speech recognition (Chatterjee & Shannon, 1998; Fu & Nogaki, 2004; Nelson & Jin, 2004; Stickney et al., 2006; Kwon & Van Den Honert, 2009). Channel interactions in a cochlear implant are influenced by many factors, including the distance between the electrode array and auditory nerve fibres, the arrangement of the electrodes on the array, the amount and condition of the surviving neurons, and the span and consistency of the current field produced by the electrodes (Shannon, 1983b; Shannon, 1985; Frijns, Briairre & Grote, 2001; Cohen et al, 2003; Cohen, Saunders & Knight, 2006; Stickney et al., 2006). Complex sounds that may act as maskers in the environment, such as speech noise, are broadband sounds that would typically excite multiple electrodes, possibly compounding the effects of masking through the effect of current field interactions. Due to these current interactions and reduced spectral resolution within the cochlea, the extent to which a simultaneous masker of a given level or frequency can be effective may be much larger in Listeners with CIs than in normal-hearing listeners, creating more frequent and larger amounts of masking (perceived as loss of clarity or distortion) in everyday situations (Fu & Nogaki, 2004; Stickney et al, 2006). Perceptual processes that occur along the auditory pathway and that help to restore or decipher less than optimal auditory stimuli, such as auditory continuity, are known to contribute greatly to auditory perception for normal-hearing listeners, and may be equally or even more valuable for Listeners with CIs. 1.2.5  Auditory Continuity  When a segment of a sound of interest is temporarily masked, the auditory system is often able to restore the interrupted information by generating the percept of an uninterrupted sound underneath a masker (Miller and Licklider, 1950; Thurlow, 1957; Vicario, 1960). This 17  ability is especially useful for the perception of speech in noisy environments. This phenomenon has been referred to as auditory continuity, auditory induction, or auditory perceptual restoration, and it is thought to be a perceptual compensation for the effects of masking (Warren et al., 1972; Bregman, 1990; Plomp, 2002; Petkov, O’Connor and Sutter, 2007). The term auditory continuity will be used hereafter to refer to this phenomenon. Perception of auditory continuity can occur when a truly continuous sound is temporarily masked by a louder sound, but it can also be exploited to create a continuity illusion (“illusory continuity” or “perceptual restoration”; Warren et al., 1972) when a segment of a foreground sound is deleted and a masker fills the interval, as illustrated in figure 1. The signal (A) consists of two flanking tones with an interposed gap. When the gap is filled with a masker (B), normalhearing listeners incorrectly report hearing the signal as uninterrupted, continuing through the noise (Miller and Licklider, 1950; Vicario, 1960; Thurlow, 1957; Warren et al., 1972).  Figure 1. Signal resulting in perception of illusory auditory continuity. This figure is a schematic representation of a pure-tone signal (A) that is interrupted and a noise masker (B) is interposed over the interruption, resulting in perception of illusory auditory continuity of the signal. (Reproduced from Bregman, 1990).  18  Within the literature describing the perception of auditory continuity, distinctions have been made between different types of continuity perception and their relation to perception of simple and complex sounds. Auditory induction, as described previously, is divided into two distinct types; temporal and contralateral (Warren, 1984). Contralateral induction is a product of binaural auditory processing (thought to prevent mislocalization due to masking) which occurs when a signal is masked in one ear, and is perceptually restored based on the signal heard in the other ear. Temporal induction involves the perceptual restoration of masked portions of signals based on the properties of the signal proceeding and following the interruption (Warren et al., 1972; Warren, 1984). Temporal induction comprises illusory auditory continuity, as described previously, and can itself be divided in to three types: homophonic continuity, heterophonic continuity and contextual catenation (Warren, 1984). Homophonic continuity occurs when an interrupted steady-state sound is perceived as continuous in the presence of a louder sound of the same spectral properties, resulting in a simultaneous perception of two levels of the same sound (Warren et al., 1972). Heterophonic continuity involves the illusory perception of continuity of an interrupted steady-state sound in the presence of another louder sound, which also differs in terms of its spectral characteristics, such as that which occurs when a tone is interrupted by a noise masker (Miller & Licklider, 1950, Vicario, 1960; Thurlow, 1957; Houtgast, 1972; Warren et al., 1972). The most complex form of temporal induction is contextual catenation, in which a spectrally complex and variable sound is interrupted and restored based on contextual information preceding and following the interruption (Warren, 1970; Warren & Obusek, 1971; Dannenbring, 1976; Warren, 1984; Ciocca & Bregman, 1987). Phonemic restoration, in which interrupted portions of speech are restored in the presence of noise based on linguistic contextual information, is an example of contextual catenation (Warren, 1970; Warren & Obusek, 1971;  19  Samuel, 1981, Warren, 1984). A recent study has measured phonemic restoration of interrupted sentences in normal-hearing listeners under conditions designed to simulate listening with a cochlear implant (Baskent, 2012). In the study, the number of spectral channels available to the listeners were systematically reduced (from 32 to 4) and fine temporal structure information was removed. The author found that with 8 channels (approximating that which is available to most Listeners with CIs), normal-hearing listeners perceived virtually no phonemic restoration effect. This finding suggests that the lack of spectral resolution and temporal fine structure information provided by current cochlear implants may prevent higher level contextual perceptual restoration of speech signals (Baskent, 2012). The current study focuses on heterophonic continuity, investigating the ability of different levels of a narrow band noise to restore perception of an interrupted pure tone. However, within the discussion section, references will be made to possible future studies focusing on contextual catenation by investigating phonemic restoration in listeners with cochlear implants. Auditory continuity can be explained as a result of perceptual grouping, also referred to as “auditory scene analysis” (Bregman, 1990). In auditory scene analysis there are two distinct, but not necessarily independent, perceptual processes: sequential and simultaneous grouping. In sequential grouping, auditory objects are formed by grouping frequency components across time (such as connecting notes of the same melody together). Simultaneous grouping denotes the ability to group frequency components that overlap in time according to their temporal and/or spectral properties, such as grouping frequency components produced by one talker apart from those produced by another sound source (Bregman, 1990). The perception of continuity employs principles of both sequential and simultaneous grouping. First, the separate segments of an interrupted sound must be grouped as a single event. That is, the section of the signal preceding  20  the masker and the one following it are sequentially grouped together to be perceived as a single, continuous event (Bregman, 1990; Bregman, Colantonio & Ahad, 1999; Darwin, 2005). Sound segments that are not aligned along the same frequency trajectory, and are unlikely to be grouped sequentially, do not produce a strong continuity percept (Ciocca & Bregman, 1987). In making continuity judgements, listeners also group sounds simultaneously: they perceptually synthesize the physically-absent signal, and at the same time they perceive the physically present masking sound, so in this way they are simultaneously perceiving two sounds (Bregman et al., 1999). Evidence from the study of auditory continuity in normal-hearing listeners shows that various requirements must be fulfilled to facilitate the perception of auditory continuity. 1.2.6  Requirements for the Perception of Auditory Continuity  In order for auditory continuity to be perceived, there are a number of requirements that must be met. These requirements include both aspects of auditory processing and physical properties of the sounds themselves. For example, if a signal alternates with a masker, such as illustrated in figure 1, continuity of the signal is heard by normal-hearing listeners if the frequency of the signal is within the frequency range of the masker, and if the masker has a high enough level to be able to mask the signal (Elfner & Caskey, 1965; Houtgast, 1972; Warren, Obusek & Ackroff, 1972). Houtgast (1972) proposed that the capability of the masker to mask the signal prevents any perceptible change in neural activity in the frequency region of the signal during the transition from the signal to the masker. This has been referred to as the energetic masking requirement. Another requirement is that there must be no evidence that there have been changes in the signal immediately preceding or following the masker. For example, silences of 50 ms that precede and follow a masker have been shown to prevent the perception of auditory continuity (Warren et al., 1972). Bregman and Dannenbring (1977) also provided 21  evidence that the strength of continuity is decreased if “signal edges” are perceived preceding and following the masker. They have shown that adding rising or falling amplitude ramps, 25 to 50-ms long, to a pure tone signal at the onset or offset around the masker, deteriorates perception of continuity compared to a condition with an abrupt signal onset and offset (Bregman & Dannenbring, 1977). The requirement that there be no perceptible signal edges is referred to as the edge masking requirement (Haywood, Chang & Ciocca, 2011). Another requirement concerns the duration of the interruption. Auditory continuity of pure tones is best perceived when the noise is neither too brief, nor too long. For normal-hearing listeners, this percept has been found to be most robust when the duration of the masker is between 10 and 300-ms long (Warren, 1999). The requirements for auditory continuity help to inform proposed explanations of how perceptual restoration takes place. Both spectral and temporal properties of sounds and how these properties are processed play a significant role in auditory continuity. As mentioned previously, in order for illusory continuity to be perceived, there should be evidence that the neural activity during the gap in the signal is similar to the activity that would be produced by a continuous signal (Thurlow & Elfner, 1959, Houtgast, 1972; Thurlow & Erchul, 1978). Data collected from both human and animal studies suggest that multiple neural mechanisms work together to facilitate perceptual restoration by mimicking the response that would have been given to the uninterrupted signal (Petkov et al., 2003; 2007; Petkov and Sutter 2011). The hypothesized mechanisms are represented in figure 2 (reproduced from Petkov et al., 2007). This figure illustrates a heuristic model of the responses from single primary auditory cortex (A1) neurons. The figure shows how these responses correlate with the perception of a continuous tone, a discontinuous tone with a silent gap, and a discontinuous tone with an interposed masker.  22  Each row in figure 2 shows schematic peristimulus time histograms for A1 neuron responses with sustained (A–C), offset (D–F), and onset (G–I) responses to the three stimuli. The interrupted tone with a noise masker results in the perception of auditory continuity and therefore, responses in the third column (C, F, and I) are like those to a continuous tone (first column; A, D and G). For example, the sustained neural response in C is comparable to that for the truly continuous tone (A). However, when the response to the continuous tone (A) is compared to that for the discontinuous tone with no masker (B), the responses are markedly dissimilar. Responses C, F and I demonstrate that the masker has adequately filled the gap between the flanking tones and the neural responses to the stimulus are comparable to that of the truly continuous tone. In this case, the perceptual system would consider the stimulus as continuous because the masker has served to maintain the neural activity required to perceive a continuous tone (Petkov et al., 2007). In light of the role of auditory neurons in representing the interrupted signal through the masker, as described by Petkov et al. (2003; 2007) and Petkov and Sutter (2011), temporal auditory processing (the encoding of auditory cues based on timing) may play an important role in the formation of auditory continuity. Temporal processing, specifically temporal resolution, occurs when the auditory system resolves changes in a stimulus over time by encoding timedependent changes in responding neural populations (Moore, 2004). Experiments have shown that neural populations responding to stimuli designed to elicit auditory continuity exhibit temporally-precise inhibitory and excitatory components (Petkov et al., 2007). The perceived continuity of a signal may be largely dependent on the ability of the auditory system to resolve the temporal characteristics required for perceptual restoration (Vinnik, Itskov & Balaban, 2010). It is important to note that listeners with cochlear implants have been shown to demonstrate 23  temporal resolution acuity that is comparable to that of normal-hearing listeners on tasks such as temporal gap detection (Shannon, 1983a; Shannon, 1989). For example, both normal-hearing listeners and Listeners with CIs have demonstrated minimal gap detection thresholds as small as 2 ms under similar stimulus conditions (Shannon, 1989; Ronken, 1970). Therefore, there is in principle no expectation that listeners with cochlear implants would be unable to perceive auditory continuity because of differences in temporal resolution abilities.  Figure 2. Schematic representation of single A1 nerve fibre responses to a continuous tone (first column), a discontinuous tone (second column) and a discontinuous tone with a masker imposed during the interruption (third column). (Reproduced from Petkov et al., 2007)  24  1.3  Objectives of the Study This study aims to contribute to the growing body of knowledge surrounding auditory  perception with cochlear implants. As auditory continuity is a basic process for the perception of sounds in noise for normal-hearing listeners, it is important that continuity be investigated in listeners with cochlear implants. Currently, there are no known studies that have investigated the perception of auditory continuity in listeners with cochlear implants. The initial investigation of auditory continuity in Listeners with CIs proposed in this study will contribute to the research in the area of auditory perception with cochlear implants, helping to work towards a comprehensive understanding of how Listeners with CIs perceive speech and other sounds in natural listening environments. This research has the potential to contribute to the improvement of implant devices and speech processing strategies in cochlear implants. 1.4  Predicted Results and Outcomes This study was comprised of two experiments: a continuity experiment in which  participants were asked to judge the continuity of a signal in the presence of a masker (experiment 1), and a simultaneous masking study (experiment 2). Experiment 1 required participants to provide a judgement of continuity or discontinuity for each single-interval trial which was comprised of a continuous, discontinuous or curtailed signal, and a masker of varying level (absent, less intense than the signal [40 dBA], equal to the signal [49 dBA] and greater intensity than the signal [55 dBA]). Experiment 2 presented two stimulus intervals with each trial, both containing a masker (of the same levels used in experiment 1); in each trial, one of the two intervals also contained a signal with the masker. Participants were asked to identify which interval they believed contained the signal. In both experiments, the signal was comprised of a  25  49 dBA signal and a masker of either 40, 49 or 55 dBA. There was also a control condition in both experiments in which the masker was absent. The stimuli and conditions of both experiments will be described in greater detail in the methods section. 1.4.1  Predicted Outcomes for Experiment 1: Auditory Continuity  Given their ability to perceive level differences between sounds (Rogers, Healy & Montgomery, 2006) and the findings about temporal processing by Listeners with CIs, it was predicted that listeners with cochlear implants would demonstrate the perception of auditory continuity. However, due to known differences between acoustic and electrical hearing, the listeners with CIs were not expected to experience perceptual restoration under the same conditions as normal-hearing listeners. Both the CI group and the normal-hearing control group were expected to perceive continuity for all truly continuous signals. Illusory continuity was also expected to be perceived in trials containing an interrupted signal when the masker was of sufficient level to effectively mask the signal were it present, and when no signal edge was detectable (for example, in the absence of silent gaps immediately preceding and following the maskers). Based on evidence from Warren et al. (1972), normal-hearing listeners were not expected to hear a continuous signal when the level of masking noise was less intense than the signal (that is, absent or 40 dBA); they were expected to perceive weak continuity with the 49dBA masker and strong continuity with the 55-dBA masker. It was expected that the poorer frequency resolution and the spread of excitation of nerve fibres (compared with acoustical stimulation) in cochlear implants (Shannon, 1983a&b; Wilson & Dorman, 2008; Kwon & Van Den Honert, 2009) would result in greater effects of masking for listeners with CIs at the masker level equivalent to the signal. Therefore, listeners with CIs were expected to perceive illusory continuity not only with the more intense masker level (i.e., 55 dBA) but also when the masker 26  was equal (i.e. 49 dBA) and possibly experience some weaker continuity when the masker was less intense than the signal (i.e., 40 dBA). It was also expected that listeners with cochlear implants may have trouble discriminating between the signal and the masker due to differences in spectral resolution and lack of temporal fine structure cues. Therefore, it is possible that the listeners from the CI group would perceive homophonic continuity (where the signal and masker have the same spectral properties) instead of perceiving heterophonic continuity (where the signal and masker have differing spectral properties; see Warren, 1984 for review). Both participant groups were expected to report perception of discontinuity with the curtailed stimuli at all masker levels due to the comparable (to normal-hearing listeners) temporal resolution abilities demonstrated by listeners with cochlear implants described previously. Accurate temporal resolution should allow both the listeners with cochlear implants and those with normal-hearing to resolve the 50-ms silent gaps between the signal-flankers and masker in these conditions, therefore violating the edge requirement for perception of auditory continuity. Evidence from previous studies demonstrates considerable individual variability on perceptual tasks in listeners with CIs, due to factors such as the distance between electrodes and nerve fibres, stimulation mode and nerve fibre survival rates (Shannon, 1983a & b; Nelson et al., 1996; Cohen, 2003). Therefore, listeners with CIs were also expected to show larger variability in their judgments of continuity both within and across subjects. 1.4.2  Predicted outcomes for Experiment 2: Simultaneous Masking  In the simultaneous masking experiment, the normal-hearing control group was expected to demonstrate effective masking when the masker level was greater than that of the signal, based on evidence from early masking studies (Wegel & Lane, 1924; Egan & Hake, 1950; 27  Moore, 1978). Specifically, they were expected to correctly identify the interval containing the signal in all trials in which the masker level was lower than that of the signal (that is, absent or 40 dBA). They were also expected to be accurate with the 49-dBA masker level, although performance was expected to be poorer than for the 40-dBA maskers. Finally, normal-hearing listeners were expected to show the lowest accuracy with the highest masker level (55 dBA). listeners with CIs were expected to employ different perceptual cues than the normalhearing listeners in the simultaneous masking experiment. Both experiments used a narrow-band masker that contained energy mostly within the frequency channel of the signal. The upper and lower frequency boundaries of the frequency band allocated to the target electrode for the listeners with CIs who use the Nucleus system by Cochlear Ltd. has a bandwidth of approximately 180 Hz (but this may vary depending on device manufacturer and model). With a masker bandwidth of 200 Hz, most of the energy of the masker would fall within the frequency band of the target electrode for the listeners with CIs. For this reason, it was expected that the signal and masker would sound very similar to the listeners with CIs. Whereas the normalhearing listeners were expected to identify the interval with the signal by selecting that which they perceived to contain a tone or have a tone-like quality, listeners with CIs may be unable to make use of this cue due to the lack of temporal fine structure (TFS) information provided by cochlear implants. Instead, they may make their judgements based on intensity cues, selecting the louder interval as the one that contained the signal plus the masker. Rogers et al. (2006) conducted a study on the perception of changes in the level of a signal under free field presentation. They found that intensity difference limens for listeners with CIs varied from 1 to 5 dB and averaged about 3 dB; these limens were larger on average than those for normal-hearing listeners (about1 dB; Veimeister, 1974). If the listeners with CIs were to rely largely on intensity  28  cues for the detection of the signal in the masking experiment, high accuracy should be observed for the 40-dBA masker condition, where the combination of the masker and the 49-dBA signal will result in a 9.5-dB increase in intensity over the 40 dBA masker alone. This difference is well over the 3-dB average intensity difference limen found by Rogers et al. (2006). When the masker intensity is equal to that of the signal (49 dBA), this will result in a 3 dB increase in intensity over the masker-only interval. For this condition it is expected that some decline in the performance of listeners with CIs’ will be observed, as the difference between the two intervals will be equal to the average difference that listeners with CIs are able to detect, and may be lower than the difference limen for some listeners. In the conditions containing the loudest masker level (55-dBA) the difference in intensity between the two intervals (1 dB) is below the average difference limen for listeners with CIs, and is not expected to be sufficient to allow consistent identification of the interval containing the signal. Given the range of intensity difference limens found in the Rogers et al. study (2006), and based on previous evidence suggesting greater variability in performance in auditory perception experiments (Shannon, 1983; Nelson et al., 1996; Cohen, 2003), it was also expected that listeners with CIs would demonstrate greater intra-subject variability than the normal-hearing group in the masking experiment. In summary, listeners with CIs were expected to correctly identify the interval containing the signal in the conditions with an absent masker and with the 40-dBA masker, but to be unable to consistently identify the signal in the conditions with the 49- and the 55-dBA maskers. 1.4.3  Correlation Analysis  Because the seminal study by Warren et al. (1972) found a correlation between the results of continuity and simultaneous masking experiments using the same signal and masker 29  levels, it was expected that a relationship would be observed between the performances of each group on the two experimental tasks. Specifically, masker conditions that resulted in an increase in continuity perception in the first experiment were expected to cause a decrease in ability to detect the signal in the second experiment. A second prediction was that the strength of the correlations between proportions of continuous responses (experiment 1) and d’ scores (experiment 2) were expected to vary within masker levels. No correlations, or weak correlations, were expected within the 40- and the 55-dBA masker levels, because masking and continuity were predicted to be very weak or very strong at those levels, respectively. For example, both participant groups were expected to show consistent judgements of continuity (ceiling effect) in experiment 1, and strong masking (floor effect) in experiment 2 with the 55dBA maskers. Conversely, a stronger correlation between the results of the two experiments was expected with the 49-dBA maskers, where more variability among participants was expected both in terms of continuity perception and signal detection.  30  Chapter 2: Methods 2.1  Participants Twenty participants, including ten listeners with cochlear implants (CI group) and ten  participants with normal-hearing (control group) were recruited for this study. The age of the participants was between 20 and 77 years old. The mean age of the CI group was 48.9 years (SD = 21.08); the mean age of the control group was 49.6 years (SD= 20.49). The subject selection criteria are reported below. Participants for the cochlear implant group were patients of the St. Paul’s Hospital Adult Cochlear Implant program; they were recruited primarily through poster advertisements in local clinics and by e-mail. Although the listeners with CIs reported a range of ages at implantation and varied hearing histories, all participants were implanted after developing spoken language; they either reported having normal-hearing prior to implantation or used acoustic amplification prior to implantation. At the time of this study, all listeners with CIs reported that they had been making full-time use of their cochlear implant for at least twelve months prior to testing. This criterion was adopted because a 12-month period of CI use has been shown to be an adequate amount of time for post-lingually deafened adult implantees to reach a plateau in speech perception ability (Helms et al, 1997; Wilson, 2006; see also Wilson and Dorman, 2008, for a review of post-implant progress in speech performance). This criterion ensured that all CI participants were able to fully understand the experimental instructions. Because many adult listeners with CIs acquire profound hearing loss due to age-related factors, it was anticipated that some of the listeners with CIs would be older adults (that is, 65 years of age or older). The process of aging has been shown to affect not only peripheral hearing sensitivity, but also the neural processing of sound stimuli, and primarily the processing of temporal cues (Gordon-  31  Salant, 2006; Walton, 2010). In order to control for the effects of factors related to aging on auditory processing, each participant in the control group was age-matched (within 5 years) to the age of each participant in the CI group. A gap detection test was also administered to all participants to screen for temporal resolution ability that was outside of the normal range (Ciocca, ©2004, custom software). This test was based on the random gap detection test by Auditec of St. Louis (Keith, 2000), and it consisted of stimuli containing either one or two 17ms bursts of white noise separated by a brief silent gap of varying duration. Gap durations were presented randomly in a single-interval, two-alternative forced choice task at either 0, 2, 5, 10, 15, 20 and 30 ms, with 10 trials of each duration. Following each trial, participants were asked to state if they perceived one or two noise bursts. Gap detection threshold was determined as the lowest gap duration which a listener could detect in 5 out of 10 trials. All participants were required to detect a silent gap of, at most, 20 ms in at least 5 out of 10 trials, in order to participate in the experimental sessions. The average gap detection threshold was 9.2 ms for the CI group and 7.2 ms for the control group (8.2 ms for both groups combined); individual thresholds ranged from 2 to 15 ms for the CI group, and from 2 to 20 ms for the control group. The individual results of the gap detection test, as well as the average thresholds for both groups are presented in table 1. All values were within the expected, typical range of normal temporal processing. In a hearing sensitivity screening, CI participants were also required to have the ability to detect a pure tone of 30 dBA or lower at 500 Hz, 1 KHz and 2 KHz under sound field presentation, using the same speech processor settings that were used during the experimental sessions. Participants for the control group were required to demonstrate hearing sensitivity in sound field equal to or better than 30 dBA at standard audiometric frequencies from 500 Hz to  32  2,000 Hz, without the assistance of any personal amplification device. This pre-test had the purpose of ensuring adequate audibility of all stimuli. All participants from both groups passed the hearing sensitivity screening. Table 1 Individual gap detection thresholds (in msand average for CI group and control group (NH) Participant  RGD  Participant  RGD  CI_S1  5  NH_S1  5  CI_S2  5  NH_S2  5  CI_S3  15  NH_S3  5  CI_S4  2  NH_S4  5  CI_S5  15  NH_S5  2  CI_S6  10  NH_S6  20  CI_S7  5  NH_S7  5  CI_S8  10  NH_S8  5  CI_S9  15  NH_S9  5  CI_S10  10  NH_S10  15  Average  9.2  Average  7.2  Note: the Abbreviation “RGD” denotes “random gap detection”.  Control participants were recruited primarily from poster advertisements placed on the Point Grey campus of the University of British Columbia, and at community centres and senior centres in Vancouver. The use of human participants for the current study was approved by the Behavioural Research Ethics Board of the University of British Columbia; all participants provided written, informed consent prior to participating in the study. The ethics certificate issued for this study can be found in Appendix A. Clinical data were obtained for all listeners with CIs from St. Paul’s Hospital clinical records. These data included details of each CI user’s implanted device and speech processor, current processing strategy, MAP, available programs and features. All participants answered a  33  brief questionnaire prior to beginning the study. Participants from the CI group answered questions about the history and etiology of their hearing loss and implantation. Participants from the control group provided their age and answered questions about prior participation in auditory experiments. The information that was obtained from all participants is summarised in Table 2. Table 2 Collected participant information. CI  Age of Dx  S1 S2  Preschool 6 yrs  S3  Birth  7 yrs  S4  3 yrs  7 yrs  S5  3 yrs  10 yrs  S6  40 yrs  16 yrs  S7  14 yrs  S8  Grade School 6 yrs  S9  36 yrs  2 yrs  S10  46 yrs  5 yrs  Time Since Ix 11 yrs 3 yrs  2 yrs  Model & Features  Strategy  Stim. Mode  Active Max.  Pulse Rate  Age  NH  Age  Previous studies  Harmony* None Freedom ADRO Freedom ADRO Freedom Autosens. Freedom None CP810 Autosens. ADRO Freedom ADRO CP810 ADRO Freedom ADRO CP810 ADRO  HIRes  Paired  15/15  36  S1  35  None  ACE  MP1+2  8/22  56  S2  59  None  ACE  MP1+2  12/22  20  S3  20  None  ACE  MP1+2  8/22  26  S4  25  None  ACE  MP1+2  8 /19  26  S5  30  SPEAK  BP+3  6 /20  3712 pps 900 pps 500 pps 900 pps 720 pps 250 pps  58  S6  59  Electrophysiology None  SPEAK  MP1+2  8 /20  77  S7  76  Amplification  ACE  MP1+2  8 /22  48  S8  53  None  ACE  MP1+2  10/22  76  S9  76  Amplification  ACE  MP1+2  8/20  250 pps 900 pps 500 pps 900 pps  66  S10  66  Auditory memory & sensitivity  Note: The abbreviation “Dx” denotes “diagnosis of hearing loss”, “Ix” denotes “implantation”, “Strategy” denotes “speech processing strategy”, “Stim. Mode” denotes “electrode stimulation mode”, “Active Max.” denotes the maximum number of electrodes activated per stimulus cycle of the total available for stimulation”, “pps” denotes “pulses per second”, “Autosens.” denotes “use of autosensitivity features”. Note: CI participant S1 uses an Advanced BionicsTM device. All other CI participants are implanted with devices from CochlearTM.  2.2  Stimuli The procedure consisted of two experimental sessions, with two separate sets of stimuli.  The main experiment was designed to investigate auditory continuity in listeners with CIs. The second experiment investigated simultaneous masking in listeners with CIs.  34  2.2.1  Experiment 1: Auditory Continuity  The main experiment contained stimuli comprised of a pure tone signal that was either physically continuous or discontinuous. The discontinuous stimuli consisted of two pure tone segments of the same frequency, level and duration, referred to as “flankers”, separated by a silent gap or by a noise burst (masker). The flankers and the continuous signal were always presented at a level of 49 dBA. Figure 3 is a schematic representation of four conditions that were included in the experiment: a continuous signal without a masker (A); a discontinuous signal without a masker (B); a continuous signal with a masker (C); a discontinuous signal with a masker (D).  Figure 3. Schematic representations of four stimulus conditions; A: continuous signal (pure tone), B: discontinuous signal with silent gap, C: continuous tone with imposed masker (narrow-band noise), D: discontinuous signal with masker filled gap. (Reproduced from Petkov, et al., 2007)  35  A fifth condition (curtailed condition) consisted of a discontinuous tonal signal that contained a masker; in this condition, a 50-ms silent gap was present preceding the onset, and following the offset of the masker (see Figure 4).  Figure 4. Curtailed stimulus condition. A masker is placed within the deleted portion of the signal, with a 50-ms. gap separating the onsets and offsets of the signal and masker.  The masker was comprised of a 300-ms sample of white noise that was band-pass filtered with a 200-Hz -3 dB bandwidth centred on the frequency of the signal, and had a slope of approximately 70 dB of attenuation at ½ octave above and below the cut-off frequencies. The bandwidth of the masker was chosen to ensure that it was wide enough to be perceptually distinct from the signal, yet narrow enough to remain frequency-specific. Raised cosine ramps of 5 ms were added to the onset and offset of the signal and masker to prevent the perception of clicklike transients due to spectral splatter that could have resulted in the perception of signal edges, thus weakening the perception of continuity. For the discontinuous stimuli containing a masker, these onset and offset ramps were positioned to overlap so as not to add to the overall duration of the stimulus. For example, at masker onset, the noise was ramped up to maximum amplitude during the first 5 ms of overlap and the first flanker was ramped down to zero during the final 5 36  ms of overlap (see Figure 5, which shows a schematic representation of the stimulus components including the onset and offset ramps). In the continuous conditions, the signal was a 1300-ms pure tone presented at 49 dBA. The discontinuous signal was comprised of two 510-ms puretone flankers, separated by a 280-ms gap. For the trials that contained a discontinuous signal and a masker, the duration of the interruption of the signal coincided with the onset and offset of the masker. In the curtailed condition, the flankers were shortened to 450 ms to include an additional 50-ms gap before the onset and after the offset of the masker. The curtailed condition was included in order to investigate the effect of purposely inducing perception of signal edges by listeners with CIs; for the control group, such edges disrupt the perception of continuity (Bregman and Dannenbring, 1977). The total sequence duration was 1300 ms for all stimuli.  Figure 5. Discontinuous signal with 49 dBA masker and onset and offset amplitude ramps.  A set of stimuli were generated for each CI user prior to testing, using the subject’s clinical data obtained from St. Paul’s Hospital. The centre frequency of the electrode most 37  closely corresponding to 1,000 Hz in each subject’s individual frequency MAP was used to set the frequency of the signal and the centre frequency of the masker. For nine participants who were implanted with a standard 22-cochlear electrode Nucleus device (© Cochlear Americas Inc.), a frequency of 1 KHz most closely corresponded to the centre frequency of electrode (E) 15 (ie. the 7th electrode from the apical end of the array). The upper and lower -3 dB bandwidth frequencies of E15 are 1090 Hz and 910 Hz, respectively, with an arithmetical CF of 1,000 Hz. The frequency of the signal and the CF of the masker were, therefore, set to 1 KHz for these participants. One CI participant was implanted with a device from Advanced Bionics (©Advanced Bionics, LLC), which consists of a 16-electrode array. The electrode with a CF most closely corresponding to 1 KHz for this subject was E6 (the 6th electrode from the apical end of the array), with upper and lower frequencies of 1053 Hz and 877 Hz respectively, and a centre frequency of 965 Hz. For this subject, the frequency of the signal and the CF of the masker were set to 965 Hz. All normal-hearing control participants were tested with stimuli comprised of a 1000-Hz signal, and a masker with a 1000-Hz centre frequency. The ability of the masker to generate a continuity percept was varied by manipulating its level in relation to that of the signal. The masker level was set to three different levels relative to the intensity of the signal: 40, 49, and 55 dBA. The masker levels were chosen for their relationship to the estimated average dynamic range of the listeners from the CI group and their expected ability or inability to produce masking percepts in the listeners. The 55-dBA masker level was selected to produce clear perception of continuity, while avoiding activation of automatic gain control for loud sounds in the speech processors – this feature typically has a knee-point of approximately 65 dBA in response to narrow-band noise. The 40-dBA masker was chosen to be clearly above threshold for both the CI and control groups, but was expected to  38  generate a weak or no continuity percept. The conditions in which the masker was absent will be referred to as the “masker absent” or “A” conditions. All signal and masker levels were set using dBA weighting to ensure adequate emphasis of the test frequency (1 KHz) in the sound field, and to ensure consistency with previous work investigating auditory continuity in normalhearing listeners under similar conditions (Bashford & Warren, 1987; Ciocca & Bregman, 1987). In summary, a total of 12 conditions were used in the main experiment, including three signal-plus-masker conditions (continuous signal, discontinuous signal with no gap, and discontinuous signal with a gap preceding and following the masker [curtailed]) and four masker levels (masker absent, 40, 49, and 55 dBA). Each subject was tested on 20 repetitions of each condition, resulting in a total of 240 trials in the experimental session. 2.2.2  Experiment 2: Simultaneous Masking  The spectral characteristics and the levels of the maskers in this experiment were identical to those of the maskers in experiment 1. The signal frequency was, again, set to the centre frequency of the electrode most closely corresponding to 1000 Hz for each CI subject, and set to 1000 Hz for each normal-hearing control subject. The main difference is that maskers and signal (when present) were simultaneous and had a duration of 300 ms. In intervals where both the signal and masker were simultaneously present, the onsets and offsets of the stimuli coincided. Figure 6 illustrates a sample stimulus sequence, containing the warning tone followed by two stimulus intervals, with the signal present in the first interval. Participants were presented with 20 trials for each condition, following a block randomization design. In summary, the experiment was comprised of four masker level conditions and 20 repetitions of each, for a total of 80 trials.  39  Figure 6. Schematic representation of a stimulus trial in the simultaneous masking experiment with a low-frequency warning tone preceding the stimulus intervals. Interval 1 contains a masker and the signal. Interval 2 contains only a masker.  2.3  Equipment The stimuli for both experiments were synthesized using a Capybara 320 sound  computation engine that was controlled by the Kyma X sound synthesis software (© 1990-2006, Symbolic Sounds Corporation) on an Apple iMac desktop computer (© Apple Computers Inc.). The presentation of the stimuli and the data collection were controlled by custom software written with the Revolution 3.0 software package (Runtime Revolution Ltd., © 2000-2009) on an Apple iMac desktop computer (© Apple Computers Inc.). The testing of all participants took place in a double-walled sound-treated booth under sound-field conditions. The stimuli were presented through an external audio interface (MOTU UltraLite) that was used to set the overall output level of the stimuli. The stimuli were presented through a Maico-ME60 sound-field amplifier (Maico Hearing Instruments Inc.) connected to a loudspeaker (Grason-Stadler Inc.) positioned at one meter directly in front of the subject. The level of the stimuli at the listening position (1 meter distance from the loudspeaker, at ear level) was calibrated using a Larson Davis System 824 sound level meter (©2006, Larson Davis Inc.). 40  2.4  Procedure After obtaining each CI subject’s consent, clinical data were obtained from St. Paul’s  Hospital to generate individual stimuli prior to the experimental sessions. Participants from both groups were tested individually in a sound treated booth in a single session lasting approximately one and a half hours, with an experimenter (the author) present at all times. First, participants completed a brief questionnaire about their hearing history, after which they were seated directly in front of the loudspeaker at a 1 meter distance in the sound-attenuated booth. A computer monitor was placed beside the loudspeaker to assist with visual prompting and reinforcing the participants during the experimental sessions. Participants from both the normal-hearing group and the CI group took part in basic threshold testing at standard audiometric frequencies from 500 Hz to 2 KHz, and in a brief gap detection test modeled after the Random Gap Detection Test (AUDiTEC of St. Louis; Keith, 2000). The upper level of comfort (ULC) was also obtained at the test frequency for all participants to ensure audibility of stimuli without causing discomfort. The ULC reported by all participants exceeded the levels of the stimuli and did not fall below 66 dBA. 2.4.1  Experiment 1: Auditory Continuity  Following a brief explanation of the continuous and discontinuous stimuli, participants were instructed to provide a judgement of whether the signal in each trial was continuous or discontinuous in a single-interval, two alternative forced choice task (1I-2AFC). Judgements were provided verbally or by pointing to the “continuous” or “discontinuous” on-screen prompt. Instruction was followed by a training session to ensure understanding of the task, as well as to ensure that participants were able to discriminate continuous and discontinuous stimuli before  41  moving on to the experimental session. The training session began with ten trials including either a continuous or a discontinuous signal with no masker, to familiarize the subject with the continuous and discontinuous signals. The criterion for continuing with the main experiment was a score of 8/10 correct for the trials in which the masker was not present. All participants from both groups met or exceeded the criterion. Participants were then asked to make continuity judgements on ten trials containing either a continuous or discontinuous signal with a 38-dBA masker (which was audible for all listeners based on the hearing sensitivity threshold screening). This masker level was chosen in order to familiarize the subject with making continuity judgements in the presence of low-level masker but was not expected to elicit an illusory continuity percept. The participants received feedback following each trial, and were able to repeat the instruction and training session if needed. There was no set criterion for performing the second familiarization task, but all listeners from both participant groups performed well (obtaining a score of 8/10 correct or higher). Following training, participants underwent 20 repetitions of each experimental condition following a randomized block design. Within each trial, participants heard a single stimulus containing either a continuous or discontinuous signal with a masker of varying level, and were required to give a judgement of “continuous” or “discontinuous” -either verbally or by pointing to one of the two prompts displayed on the computer monitor. Responses for each trial were recorded as “continuous” or “discontinuous” by the experimenter. These scores were then converted to a proportion of continuous responses for each continuity condition at each masker level. Both the presentation of the stimuli and the data collection were controlled by the experimenter, who operated the custom experimental software. Participants were provided the  42  opportunity to take a break at any time and offered the opportunity to take a break outside of the sound booth following the completion of the first experiment. 2.4.2  Experiment 2: Simultaneous Masking  In the second experiment, participants were asked to listen to a sequence of two intervals: an interval that contained a masker without the signal, and another interval that contained a masker plus the signal (two-interval, two alternative forced choice task; 2I-2AFC). Prior to beginning the second experiment, participants received thorough instruction and underwent a two-part training session similar to that in the first experiment. The first block in the training session was composed of ten test trials of two intervals each, one interval that contained the signal, and another interval that contained no signal. The two intervals were separated by a 500ms silent interval. The criterion for continuing with the main experiment was a score of 8/10 correct. The criterion was met or exceeded by all participants in both groups. The participants were then presented with a block of ten trials of two intervals: a soft masker (38-dBA level) was present in both intervals; a signal present in one of the intervals. A 500-ms amplitude-modulated warning tone was presented at the beginning of each trial to prompt the subject to prepare to attend to the stimulus. The first stimulus interval began 500 ms after the presentation of the warning tone. Following each trial participants received feedback on the computer monitor as to whether their answer was “correct” or “incorrect”. The participants had the opportunity to repeat the instructions and the training sessions if required. No criterion had to be met by participants in this second familiarization task in order to proceed to the second experimental session. Following training, participants were presented with 20 repetitions of each test stimulus following a randomised block design. The trials in the experimental session began, like the training session, with a warning tone. Listeners responded after the presentation of the stimuli in 43  the second interval by selecting the interval that contained the signal. They were provided with on-screen visual feedback indicating whether their response was “correct” or “incorrect”. Responses were scored as “1” for “correct”, or “0” for “incorrect” for further data analysis. As for the continuity task, the experimenter operated the custom software to control the presentation of the stimuli and the data collection; participants had the option to take a break at any time. 2.5  Data Analysis 2.5.1  Experiment 1: Auditory Continuity  Raw data was collected from each subject in the form of a “continuous” or “discontinuous” response for each trial. For each subject, a proportion of continuous responses was calculated for the 20 trials in each condition. Results from all listeners were averaged for each condition, to obtain a representation of group performance. A high proportion of “continuous” responses was expected in all truly continuous conditions; a high percentage of “continuous” responses was also expected in the discontinuous signal conditions at masker levels of 49 and 55 dBA. A three-way, repeated-measures multivariate analysis of variance (MANOVA) was conducted on the proportion of continuous responses to assess the effect of the independent variables on group performance. The three factors analyzed were; 1) group (CI or normalhearing), 2) Continuity (continuous, discontinuous or curtailed), 3) masker level. All three independent variables were analysed as within-subject factors. Where significant results were found, Tukey HSD post-hoc comparisons were used to further differentiate between conditions.  44  2.5.2  Experiment 2: Simultaneous Masking  Raw data was collected in the form of a judgement of the presence of the signal in the “first” or “second” interval for each trial sequence. Responses were scored for each trial as a “correct” or “incorrect” response. In a preliminary analysis, the percentage of correct responses was then calculated for each subject across the 20 trials for each condition, and then averaged across listeners to determine overall group performance. Because correct percent responses can be a biased measure of sensitivity, a d’ analysis was performed using the entire data set for each participant in both listener groups. Trials in which the listener responded that the signal was present in the first interval, and the signal was, in fact, present were classified as a “hit” (H). Trials in which the listener responded that the signal was present in the first interval when, in fact, it was present in the second interval, were termed a “false alarm” (F). The d’ scores were calculated by first transforming the proportions of hits and false alarms into their corresponding z scores [z(H) and z(F), respectively], and then by subtracting the transformed false alarm rates from the transformed hit rate using the following formula for two-interval, forced choice tasks: d’ = 1√2 [z (H) – z(F)] (Macmillan and Creelman, 2005). A high d’ value was expected for those conditions in which the masker could not effectively mask the signal (i.e., absent masker or 40 dBA masker), allowing the listeners to accurately detect the signal. The d’ scores were then analyzed in a two-way MANOVA. The factors were 1) group and 2) level. Both factors were analyzed as within-subject factors. Tukey HSD post-hoc comparisons were then applied to further investigate differences between the groups.  45  Chapter 3: Results The results from the main continuity experiment and from the secondary simultaneous masking experiment are presented separately in the following sections. 3.1  Continuity Analysis For each condition, the proportion of continuous responses from each listener, ranging  from “0” (no “continuous” response) to “1” (20 “continuous” responses out of 20), were combined to give an average proportion of “continuous” responses for each group. The mean proportions of rated continuity for the absent masker for the continuous and discontinuous conditions, as well as for all curtailed conditions are presented in Table 3. Due to observed floor/ ceiling effects, none of these conditions were included in the statistical analysis. Table 3 Mean proportions of rated continuity and standard deviations (SD) for the cochlear implant group (CI) and control group (NH) to continuous and discontinuous stimuli with absent masker (A) and curtailed stimuli at all masker levels. CI NH Condition Mean (SD) Mean (SD) A-Cont  1 (0)  1 (0)  A-Disc  0 (0)  0.006 (0.015)  A-Curt  0 (0)  0 (0)  40-Curt  0.125 (0.190)  0.039 (0.071)  49-Curt  0.1 (0.083)  0.056 (0.054)  55-Curt  0.175 (0.184)  0.078 (0.093)  Note: the abbreviation “Cont” represents the physically continuous stimulus condition, “Disc” represents the physically discontinuous stimulus condition, “Curt” represents the curtailed condition. . A, 40, 49, or 55 represents the level of the masker in dBA.  The mean proportions of continuous responses (with standard error bars) for all continuous and discontinuous conditions at the three masker levels are shown in Figure 7. Due to 46  a violation of the sphericity assumption, a multivariate procedure was used to analyse the data, as recommended by Howell (2003, pp.476-477). The data were analyzed using a three-way repeated-  measures MANOVA with “group” (CI and control), “continuity” (physically continuous or discontinuous signal) and “masker level” (40, 49, and 55 dBA) as the repeated factors. While perceived continuity for the physically continuous stimuli was equally strong for the two groups, an important difference was observed for the physically discontinuous stimuli. For discontinuous stimuli, only the cochlear implant group reported a high proportion of continuous responses with the 49-dB level masker; stronger continuity perception was also reported by the CI group at the 40-dB masker level. The difference between CI and NH groups as a function of masker level and stimulus type was statistically significant -group x continuity x masker level interaction, (F(2, 8)=17.7, p=0.001. All of the two-way interactions were statistically significant: group x continuity interaction, F(1,9)=12.6, p<0.01; group x level interaction, F(2,8)=7.28, p<0.05; continuity x level interaction, F(2,8)=106.6, p<0.001. The physically continuous stimuli received a higher proportion of continuous responses than the discontinuous stimuli overall, as shown by a significant main effect of continuity, F(1,9)=102.5, p<0.001. As predicted, there was also significant main effects of group, F(1,9)=24.2, p<0.001, and of level, F(2,8)=55.7, p<0.001. The group and level effects show that the CI group perceived the stimuli as more continuous overall, and that the perception of continuity was dependent on the level of the masker, respectively. Tukey post-hoc tests showed that perceived continuity was stronger for the CI group than for the control group for the discontinuous stimuli at the 49-dBA masker level (p<0.01). The two groups did not differ for the discontinuous condition at the 40-dBA and 55-dBA masker levels (p>0.05).  47  CONT  DISC  CONT  DISC  CONT  DISC  Physical Continuity of Signal  Figure 7. Unweighted mean proportions of rated continuity by the cochlear implant group (CI) and control group (NH) to physically continuous (“CONT”) and discontinuous (“DISC”) stimuli at 40 dB, 49 dB and 55 dB masker levels. Error bars indicate +/- standard errors.  3.2  Simultaneous Masking Analysis Responses to the simultaneous masking stimuli were collected from the ten participants  in each group for all level conditions and coded as “1” (correct identification of interval containing signal) or “0” (incorrect identification of interval containing signal). These values were averaged for each condition and each listener, and then transformed into H and F scores in order to calculate a d’ score for each subject and all level conditions. The un-weighted means for both groups are presented in Figure 8. As predicted, both groups showed better performance with the absent masker and the 40 dBA maskers; performance declined as the masker level increased to 49 and 55 dBA. The mean scores showed slightly lower performance for the CI group compared with the control group for the 40 and 55 dBA masker levels. As was the case in  48  the continuity experiment, a larger difference was seen between the two groups at the 49-dBA masker level. Again, a larger amount of variability was observed for the group of listeners with cochlear implants and due to a violation of the sphericity assumption, the d’ scores were analyzed using a two-way MANOVA (“group” and “level” were the two within-subjects factors). The absent masker condition was excluded from the analysis due to observed ceiling effects -the mean d’ scores calculated for both the CI and control groups with the absent masker were 3.29 (SD = 0). As predicted, the analysis revealed significant main effects of group, F(1,9)=5.84, p<0.05, and of level, F(2,8)=147.5, p<0.05. The interaction of group and level was not statistically significant, F(2,8)=1.75, p>0.05  Figure 8. Unweighted mean d’ scores for the simultaneous masking task for the cochlear implant group (CI) and control group (NH) at 40 dB, 49 dB and 55 dB masker levels. Error bars indicate +/- standard errors.  Although the MANOVA found no significant interaction effect between the CI and Control groups, a Tukey-corrected post-hoc analysis was conducted on the data collected in order to compare the two groups across all three level conditions separately. Because the overall 49  F can be weakened by distributing it across all degrees of freedom for all groups in an overall test such as a MANOVA, it is warranted to conduct post-hoc individual comparisons despite the lack of a significant interaction effect (Howell, 2003, pp.372-373). The post-hoc analysis revealed a significant difference between the two groups for the 49-dB masker level (p<0.05), but not for the 40-dB level (p>0.05) or the 55-dB level (p>0.05). These results are consistent with the prediction that masking is stronger at a lower masker level for the CI group than for the control group. 3.3  Correlational Analysis and Individual Performance The first prediction of the relationship between the continuity and the masking results  was that there should be a positive relationship between the strength of masking (e.g., low d’ scores in experiment 2) and the strength of the perception of continuity (e.g., high proportion of continuous responses in experiment 1) (Warren, Obusek & Ackroff, 1972). This prediction was confirmed by the general pattern of responses across masker levels for both the CI and the control groups. For example, perceived continuity was low with the absent masker and the 40 dBA masker level, for which the amount of masking was also small (large d’ scores). The opposite pattern was observed at the 55-dBA masker level. Pearson’s r correlations were then calculated between proportions of continuous responses on the continuity task and simultaneous masking d’ scores at each masker level and for each group. All correlation coefficients for the separate masker levels are reported in Table 4.  50  Table 4: Pearson’s r correlation values for auditory continuity and simultaneous masking tasks Group  40 dBA  49 dBA  55 dBA  CI  r=0.2167 p >0.050  r=0.667 p >0.050  r=0.4227 p >0.050  NH  r=0.217 p >0.050  r=0.620 p=0.056  r=0.229 p >0.050  Combined  r=-0.014 p >0.050  r=-0.414 p=0.070  r=-0.020 p >0.050  Contrary to the initial prediction, no significant correlations were found between performance on the auditory continuity task and the simultaneous masking task for either participant group within each masker level. As the lack of significant correlations between the two tasks within each masker level may be due to the small sample size, the data from the CI and control groups were combined. Again, no significant correlations were found. As the correlation between masking and continuity was most likely to be observed at the 49-dBA masker level, where both continuity and masking data were likely to be most variable, the data for this masker level are displayed in Figure 9. The data in this figure show that although the expected negative correlation just fails to reach statistical significance, the pattern of the data show that there are two clusters of data points, one corresponding to responses for the normal hearing listeners and one corresponding to the listeners with cochlear implants. As expected the data for the CI group is more variable than that of the control group. The pattern for the control group in particular suggests the presence of a positive correlation between sensitivity and continuity perception; this pattern is in fact the opposite of the expected correlation. However, the correlation for the control group just failed to reach statistical significance; it is also the case that continuity was generally  51  weak for the control group at this masker level as most of the proportions of continuity ratings were below 0.5.  Sensitivity (d’ Scores)  Figure 9. Continuity ratings vs. sensitivity d’ scores for the CI and control groups with the 49 dBA masker.  When all the data are combined, the same pattern of responses can be observed (see figure 10). The negative correlation demonstrates the expected relationship between the two tasks, in that when perception of continuity is strong (experiment 1), sensitivity to the signal is low, due to effective masking (experiment 2). The data points in figure 10 are, again separated into two clusters of data. This pattern results from the large difference in performance between the CI group and control group on the 49-dBA masker level conditions and is also demonstrated in figure 9. The data points in figure 10 representing the results with the 40-dBA masker (diamond symbols) show that the data for control listeners (closed symbols) is clustered in the lower-right area of the figure, indicating high sensitivity with the simultaneous masker and the perception of a discontinuous signal in the continuity task. Listeners from the CI group (open symbols) were more variable along the continuity dimension; sensitivity with the simultaneous masker was also 52  more variable than for control listeners, but remained relatively high (d’ > 2 for all but one listener). Finally, the data points for the 55-dBA masker level (square symbols) are clustered in the upper left area of figure 10, suggesting low sensitivity to the signal with the simultaneous masker, and the strong perception of a continuous signal in the continuity task. The correlation between the two tasks at the 55-dBA masker level failed to reach statistical significance for either participant group or the two groups combined. While the correlation coefficient is statistically significant in this case, it is unclear that the strong correlation coefficient, which is statistically significant, reflects the existence of a real correlation in the population.  Figure 10. Continuity ratings vs. sensitivity d’ scores for the CI and control groups at 40, 49 & 55 dBA masker levels.  As a few listeners with CIs reported some continuity perception for the curtailed stimuli, a correlational analysis was carried out between the results of all participants in the curtailed 53  conditions in the first experiment and the results of the random gap detection test (RGDT) taken at the beginning of the experimental session. This analysis investigated whether there was a relationship between the ability of the participants to resolve the gaps between stimuli in the RGDT and the perception of continuity in the curtailed conditions where a 50-ms silent gap separated the masker and flanking signals. No significant correlations were found between the performance of the participants on the RGDT and curtailed conditions in the continuity experiment (see Table 5). Table 5: Pearson’s r correlation values for random gap detection and auditory continuity (curtailed conditions). Group  40 dBA  49 dBA  55 dBA  CI  r= -0.462  r= -0.525  r= -0.245  p >0.050  p >0.050  p >0.050  r=0.325  r= -0.086  r= -0.109  p >0.050  p >0.050  p >0.050  r= -0.123  r= -0.230  r=-0.020  p >0.050  p >0.050  p >0.050  NH  Combined      54  Chapter 4: Discussion This chapter will discuss the findings in relation to the hypotheses and the predictions put forth in chapter 1. Clinical and theoretical implications and limitations of this study, as well as recommended future directions for research, will also be discussed. 4.1  Summary The findings of the present study show that both CI and control listeners perceived the  physically continuous signal as continuous, regardless of masker level. As was expected, both the CI and control groups reported: a perception of discontinuity for the discontinuous signal with an absent masker and with the 40 dBA masker, and for all curtailed stimuli (although three listeners with CIs reported continuous percepts in some instances, as discussed in a later section); a perception of continuity for the discontinuous signal and a 55-dBA masker. The two groups differed in their perception of the discontinuous signal when the masker level was equal to that of the signal (49 dBA). At this level, listeners with CIs mostly reported a continuous percept, while the control listeners predominantly heard the signal as discontinuous. This difference is consistent with the prediction that a pure tone signal and a narrow-band noise of the same level generate a percept of continuity for users of cochlear implants, likely due to their relatively poor frequency selectivity, increased susceptibility to the effects of masking and lack of temporal fine structure cues. Overall, these findings show that listeners with CIs do experience perception of auditory continuity, but they do so under different conditions than normal-hearing listeners. In experiment 2, when the masker and the signal were simultaneous, the listeners with CIs were less accurate than the control listeners overall. This difference was statistically significant at the 49-dBA masker level only. This finding is consistent with the prediction that  55  listeners with CIs would experience more difficulty detecting the signal at an intermediate masker level, due to the effects of spread of excitation, poor frequency resolution and lack of fine temporal coding with a cochlear implant. The CI group also demonstrated greater variability in performance with the 49-dBA masker level condition. At the 40-dBA level, both participant groups were easily able to identify the interval containing the signal. When the masker level was sufficiently high to produce a large amount of masking (55 dBA), both groups were unable to consistently identify the interval containing the signal. By using similar stimuli in both experiments, it was possible to directly explore the relationship between the simultaneous masking and auditory continuity. For both groups, a consistent perception of continuity occurred at the same masker levels at which masking was strong (55 dBA for control listeners; 49 and 55 dBA for listeners with CIs). 4.2  Perceived continuity by listeners with cochlear implants In the auditory continuity experiment, the CI group reported a percept of continuity for  continuous stimuli at all masker levels (absent, 40, 49 and 55 dBA), and for discontinuous stimuli at the 49- and 55-dBA masker levels. The perception of continuity of the discontinuous signals is of particular interest. Because the signals in these stimulus conditions are physically discontinuous, the consistent reports from the CI group of a continuity percept on these conditions supports the prediction that the perceptual restoration of missing sounds is possible with the use of a cochlear implant. As described in previous sections, normal-hearing listeners have been known to have the ability to perceptually restore missing sounds if the missing portion of sound is replaced by a masker sufficient to mask the signal were it actually present (that is, the masker should have as much energy as the signal in the frequency of the signal). In the current study, listeners with CIs 56  demonstrated perceptual restoration similar to that observed previously in normal-hearing listeners, with some important differences. For the stimuli which contained a discontinuous signal and a masker that was less intense than the signal (40 dBA), the listeners with CIs and normal-hearing listeners both perceived the signal as discontinuous. When the masker level was the same (49 dBA) or higher (55 dBA) than that of the signal, the listeners with CIs reported perceiving the signal as continuous, clearly demonstrating their ability to perceptually restore missing portions of sound. The normal-hearing control group only perceived continuity when the masker was louder than the signal (ie., with the 55-dBA masker). The differences seen between the listeners with CIs and normal-hearing listeners are likely to be related to the differences in the effects of auditory masking between the two populations. As previously discussed, listeners with CIs have generally been shown to differ from normal-hearing listeners in their response to auditory maskers, showing stronger masking effects than normal-hearing listeners for the same masker levels. As explained in the introduction section, the higher susceptibility to masking by listeners with CIs is likely due to effects of current interaction and spread within the cochlea, poorer frequency resolution and lack of TFS (Shannon, 1983; Shannon, 1985; Stickney et al. 2006). Although it is difficult to predict the exact patterns of stimulation within the cochlea of each listener with a CI, the narrowband maskers used in this study are likely to have activated multiple electrodes in close proximity. The resulting spread of excitation is likely to cause difficulty discriminating between the signal and masker, for the listeners with CIs than for the control listeners, which could result in a stronger perception of illusory continuity for the listeners with cochlear implants. An additional contributing factor in the listeners with cochlear implants’ perception of strong continuity with the 49-dBA masker, is that the frequency resolution provided by the  57  Cochlear Implants may not have permitted them to distinguish the signal from the masker during stimulus presentation. In spite of their ability to hear differences between the signal and the masker when paying attention to these stimuli presented in a sequence during the familiarization sessions, during debriefing following the experimental sessions two listeners from the cochlear implant group spontaneously remarked that they had difficulty telling the signal from the masker during the experimental trials (or that they were unsure if they were hearing one, the other or both). The differences in performance between the participants from the CI group and normalhearing participants may reflect differences in perception of the signal and masker between the two groups. Although the listeners with CIs were able to differentiate between the signal and a 38-dBA masker in the practice trials, the experimental conditions in which the masker was equal (49 dBA) or higher (55 dBA) in level than the signal may have resulted in more difficulty differentiating the signal and masker. This difficulty may have caused a greater perception of continuity in these trials compared with the normal-hearing listeners. The observed effects of reduced spatial resolution in the CI group are consistent with the recent findings of Baskent (2012), which investigated phonemic restoration in normal-hearing listeners, under simulated cochlear implant listening conditions. As discussed in a previous section, the study concluded that perception of phonemic restoration (a form of auditory continuity) was affected by the lack of spatial resolution, characteristic of cochlear implant processing. The curtailed conditions were not included in the statistical analysis due to observed floor effects for listeners from both the CI and control groups. While the responses to the curtailed conditions were not statistically different between the two groups, three listeners from the CI group reported more frequent continuity perception in some conditions. The participants who reported this continuity perception each did so for stimuli at two different masker levels (CI_S1  58  at 40 dBA, CI_S3 & CI_S7 at 55 dBA). Moreover, all three listeners with CIs who reported continuous judgements on a curtailed condition provided a proportion of continuous responses near 0.5, which may suggest confusion or guessing rather than consistent continuity perception. Thirdly, CI_S1, who provided the highest proportion of continuous responses in a curtailed condition (0.65; at the 40-dBA masker level) also provided a high proportion of responses that were equal or near equal for all discontinuous conditions in which a masker was present (0.85, 0.85 and 0.8 for the 40, 49 and 55 dBA levels, respectively). It is apparent that CI-S1 perceived an overall high proportion of continuity that was unrelated to the level of the masker. While it is difficult to know with certainty the reason(s) for the atypical performance of this listener, it is noteworthy that this listener was the only one who used an Advanced Bionics (AB), LLC system. This listener received auditory input with a different mode of stimulation (paired pulsatile stimulation [AB] vs. sequential [Cochlear]), and at a much higher stimulation rate than all other listeners with CIs, who used systems produced by Cochlear Americas Inc. (3712 pps vs. a range of 250 – 900 pps per active electrode). The number of active maxima and of electrodes also differed between these systems. One or more of these factors might have affected the perception of auditory continuity by this listener (CI_S1). For example, it is possible that a very high stimulation rate can contribute to lower thresholds for auditory stimuli, but may also lead to increased neural adaptation and electrode interaction (Kiang & Moxon, 1972; Shannon, 1983a & b; Middlebrooks, 2004; Bonnet, Boermans, Avenarius, Briaire & Frijns, 2012). Similarly, there is evidence that a paired mode of stimulation can result in greater spread of excitation when compared with a fully sequential stimulation mode (Battmer, Zilberman, Haake & Lenarz, 1999; Stickney et al.; 2000; Buechner et al., 2005). Therefore, CI_S1 might have been particularly susceptible to the effects of masking in comparison with the other listeners with CIs. The two  59  other listeners with CIs who reported continuity percepts for the curtailed stimuli did not have any other clear inconsistencies in their individual results, but also provided a lower proportion of continuous responses for the curtailed stimuli than CI_S1 (0.55 and 0.45 at the 55 dBA masker level). It is also possible that these results may not be representative of any consistent perception of continuity, but of confusion or a lack of confidence in the judgement experienced by these listeners with CIs in the presence of the loudest masker. Finally, it is possible that the continuity percepts in response to the curtailed stimuli from some listeners with CIs could have been due to the temporal integration characteristics of the CI population (Shannon, 1983a, 1989.). Current interaction and summation within the cochlea during electrical stimulation affects not only the auditory signal across frequencies and intensities, but also across time (Shannon, 1983a, 1989; Pfingst et al, 1991; Donaldson et al, 1997). There is a possibility that this spread of excitation over time in addition to refractory and facilitative effects may have contributed to the sporadic reports of continuity perception from some listeners with CIs to the curtailed stimuli because of an impaired ability to clearly perceive gaps between the flanking signal and masker. However, the failure to find a correlation between continuity perception for the curtailed conditions and gap detection thresholds for the RGDT, in addition to the fact that all gap detection thresholds were within the normal range, suggests that an impaired temporal processing ability is not a likely explanation for the continuous percepts in the curtailed conditions. 4.3  Simultaneous Masking Performance by Listeners with CIs Previous studies have found that the effects of auditory masking differ between listeners  with CIs and normal-hearing listeners (Shannon, 1983; Shannon, 1985; Stickney et al. 2006). Since there is evidence that auditory continuity is closely related to the process of auditory 60  masking (Warren et al, 1972; Bregman, 1990; Plomp, 2002; Petkov et al, 2007) the present study investigated simultaneous masking with the same signal and maskers used in the continuity experiment. As previously discussed, auditory continuity involves both the simultaneous and sequential grouping of sounds. During auditory continuity perception, listeners perceive sound components presented sequentially across time, but when perceiving the signal continuing in the presence of the masker, they also perceive two sounds simultaneously (Bregman, 1990; Bregman, Colantonio and Ahad, 1999; Darwin, 2005). The results of experiment 2 showed that the ability of listeners with CIs to identify the interval containing the signal when it was presented with a simultaneous masker varied with the level of the masker. Both the CI and the control groups were consistently able to identify the interval containing the signal when the masker level was less than that of the signal (ie. absent or 40 dBA). Unlike the control group, the CI group showed a significant decline in the ability to detect the signal in the presence of the 49-dBA masker. In addition to a steeper decline in ability to correctly identify the interval containing the signal with increasing masker level, the CI group also demonstrated greater variability on all conditions than the control group. The CI group demonstrated the greatest variability in performance on the 49-dBA masker condition. When the results are viewed in terms of individual performance, five of the listeners with CIs obtained a d’ score over 1, meaning they were able to consistently identify the interval containing the signal at the 49 dBA-level, performing equally well as some of the control listeners. The remaining five listeners with CIs operated only at chance level with this masker. These results support the previous findings of higher susceptibility to masking by the listeners with CIs. The reasons for this difference could be due to either the effects of current  61  interactions on the perception of masking or the inability of the listeners with CIs to consistently perceive the signal as distinct from the masker, as previously discussed. An important difference between the performances of the two listener groups in this task is that listeners with CIs may have based their judgements on intensity cues as opposed to detecting the signal (or a tone-like quality) in the intervals containing both signals and maskers. As mentioned in the previous section, two listeners with CIs commented that they were not always sure when they were hearing the signal or masker during the auditory continuity task. Furthermore, four participants from the CI group spontaneously reported (either during or after completing the simultaneous masking experiment) that they identified the interval that they perceived as louder to be the one containing the signal. Only one participant from the CI group remarked that the interval containing the signal sounded sharper, whereas four of the control participants remarked that they relied on the presence of a tonal quality as a cue to the selection of the interval containing the signal. If it is the case that listeners with CIs used predominantly (or exclusively) intensity cues to perform this task, the following pattern of results would be expected on the basis of previous intensity discrimination findings for this population (Rogers et al., 2006). At the 40-dBA masker level, correct identification of the interval containing the signal would be expected because the intensity difference between the signal-plus-masker interval and the masker only interval would be 9.5 dB. This prediction follows from the finding that listeners with CIs have been shown to be able to discriminate intensity differences within a range of 1 to 5 dB (Rogers et al., 2006). When the masker is equal in level to the signal (i.e., 49 dBA), the difference in intensity between the two intervals would be only 3 dB, the average difference limen for intensity discrimination in listeners with CIs found by Rogers et al. (2006). In this condition, only some individual listeners with CIs should be able to perform consistently above  62  chance, and the rate of correct identification based on intensity would decline accordingly. This prediction was again confirmed by the findings of the present study. Finally, with the 55-dBA masker the difference in level between the two intervals is relatively small (1 dBA); very few listeners with CIs would be expected to consistently perceived such an intensity difference, leading to the observed poor sensitivity at this masker level. In summary, the observed results are consistent with the use of intensity cues by listeners with CIs to perform this task. Greater individual variability in the ability to correctly identify the interval containing the signal was also predicted for the CI group in comparison with the control group. The greater variation in responses seen for the CI group is likely due to the many factors that have been shown to influence the perception of masking and auditory perception in general in listeners with CIs. Greater variability in individual performance on this masking task may have resulted from individual participant factors such as amount of auditory nerve fibre survival, or device characteristics including speech processing strategy, default noise reduction features (eg, ADRO or autosensitivity) or proximity of electrode array to its target nerve fibres (Shannon, 1983; Nelson et al., 1996; Cohen, 2003). One or more of these factors can differ between listeners with CIs and can result in great variability in auditory perception between individual listeners and are likely to have influenced the variability observed in the current experiment. 4.4  The Relationship Between Auditory Continuity and Simultaneous Masking for  Listeners with Cochlear Implants To the best of the author’s knowledge, the perception of auditory continuity has not yet been studied in the cochlear implant population. The current study had the purpose of determining whether listeners with CIs perceive continuity in a similar way to listeners with normal-hearing. Auditory continuity has been shown to be closely related to the perception of 63  auditory masking for normal-hearing listeners (Houtgast, 1972; Warren et al., 1972; Petkov, O’Connor and Sutter, 2007). This relationship is often presented as the “energetic masking” requirement, which states that illusory continuity should occur only if the masker is capable of masking the missing sound, if this sound were present (Warren et al., 1972, Houtgast, 1972). Because listeners with CIs experience the effects of auditory masking at higher signal-to-noise ratios than normal-hearing listeners (Shannon, 1983; Shannon, 1985; Stickney et al. 2006), the energetic masking requirement is expected to follow the patterns of masking seen in this population. That is, listeners with CIs were expected to perceive physically discontinuous signals as continuous for masker levels at which substantial masking would occur in a simultaneous masking task. This prediction was confirmed by the outcomes of the present study when listeners from the CI group reported continuity perception with the 49-dBA masker. The CI group showed a consistent illusory continuity percept at the 49-dBA masking level in the auditory continuity experiment and an inability to consistently identify the signal at this masker level in the simultaneous masking experiment. In this respect, listeners with CIs showed the same relationship between masking and perceived continuity as listeners with normal-hearing. Although the overall results support the connection between masking and auditory continuity, this connection was not demonstrated by the individual data within each masker level. This failure to observe a correlation between masking and perceived continuity data may be due to the small sample size and/or to the lack of sensitivity of the tasks used in the current study. Using the method of adjustment, Warren et al. (1972) obtained a pattern of auditory continuity perception as a function of both intensity and frequency separation of signal and masker. The resulting curves for the six participants closely resembled the auditory masking curves obtained from the same listeners. The current study did not investigate the effect of the 64  frequency separation between the signal and masker, and employed only four levels of intensity. It is possible that if the present study had included a different task, a closer connection between the masking and continuity data, such as that found by Warren et al., might have been observed. This hypothesis will be discussed further in a following section on “Future research”. 4.5  Clinical Implications The results of these experiments add to the growing understanding of the processes that  affect the ability of listeners with CIs to listen under noisy conditions. Listeners with cochlear implants can demonstrate speech perception abilities that approximate or match those of their normal-hearing peers under ideal listening conditions (in quiet). However, their speech recognition performance declines under degraded listening conditions, such as in noisy environments (Carlyon, Long, Deeks and Mckay, 2007; Cooper and Roberts, 2009; Henkin, Tetin-Schneider, Hildesheimer and Kishon-Rabin, 2007; Peterson et al., 2010; Ricketts, Grantham, Ashmead, Haynes and Labadie, 2006). One of the main goals of new developments in cochlear implant technology is the improvement of speech perception in challenging environments. Given the evidence about the relationship between masking and perceived continuity, the investigation of auditory continuity in the CI user population could prove useful for the development and the evaluation of new methods and technology in cochlear implants. Research and development in the field of cochlear implants is taking many different directions, such as the ongoing effort to decrease current interaction among adjacent electrodes and to improve frequency resolution. For example, the use of neurotrophic drugs has been proposed to induce neurocite growth toward the electrode array in order to decrease the space between the array and its target, effectively decreasing current spread and interaction (Roehm 65  and Hansen, 2005; Pettingill et al, 2007; Ramekers, Versnel, Grolman and Klis, 2012). An intramodiolar or intraneural implant, which positions the electrodes directly adjacent to the auditory nerve axons through an opening at the base of the cochlea, has also been proposed to increase channel independence, lower thresholds and increase dynamic range (Arts, Jones & Anderson, 2003; Badi et al, 2003 & 2007; Middlebrooks & Snyder, 2008). If the perceptual processes that contribute to both auditory continuity and the understanding of speech in noise are affected by changes in current interaction and frequency resolution, evidence from auditory continuity studies may provide ongoing motivation, direction and means for evaluation of progress in these areas. 4.6  Limitations The present study is a preliminary investigation of the perception of auditory continuity  by listeners with cochlear implants. Therefore, the study has a number of limitations that require further investigation. A limitation of the current study is the small, heterogeneous participant group. The CI group’s results displayed greater variability when compared with the control group, as is often the case with cochlear implant studies (Shannon, 1983a & b; Nelson et al., 1996; Cohen, 2003). This variability can be attributed to a variety of factors. For listeners with CIs, several individual and device-related factors can affect the overall pattern of stimulation, including the level of current from the stimulus, the amount of neural survival, the placement of the electrode array, the mode of stimulation (ie. monopolar or bipolar, simultaneous or interleaved), current field interactions and the growth of new bone or tissue effecting current flow within the cochlea (Cohen et al, 2003). In the current study it was not possible to deactivate processing features in the cochlear implant speech processor programs. Rather, individual listeners were asked to choose their “everyday” programs, or the program with the least features. 66  In the current study, of the nine listeners from the CI group who used a Cochlear Ltd. device, seven used ADRO as a default feature and two did not. Because ADRO continuously adjusts the gain of incoming sounds in relation to the listener’s dynamic range, it is possible that ADRO could have had an effect on how the stimuli were perceived by the listeners who used it. Table 6 contains the means for the two subgroups on both tasks at all masker levels. The ADRO group reported a slightly higher proportion of continuity and were less sensitive to the signal on the masking task, except with the 55 dBA masker. However, some of the differences were small and given the lack of statistical tests due to the small sample size, it is unclear whether the differences resulting from the use of ADRO in this sample represent actual differences in the population. Table 6 Performance means for groups of listeners with cochlear implants with no ADRO and with ADRO. Group  40 dBA  49 dBA  55 dBA  No ADRO  AC: 0.2  AC: 0.775  AC: 0.85  (2 listeners)  SM: 3.29  SM: 1.645  SM: 0.112  ADRO  AC: 0.243  AC: 0.893  AC: 0.936  (9 listeners)  SM: 2.911  SM: 0.867  SM: 0.84  Note: “AC” denotes the auditory continuity task and “SM” denotes the simultaneous masking task.  Individual factors unrelated to the implanted device can also affect performance for listeners with CIs, including motivation and fatigue. For example, listeners with cochlear implants typically require more listening effort and experience more fatigue in auditory perception tasks than normal-hearing listeners (Zeng, 2004). Possible sources of individual variability in the performance of CI participants in the current study include differences in hearing histories, device models, stimulation modes and MAPing strategies, age and CI 67  experience. For example, within the CI participant group three different speech processor models were used with three different processing strategies; years of experience listening with a CI ranged from two years to 16 years (for a summary of participant information, see Table 1). The possibility of variability in temporal processing abilities among participants presented a special challenge in the current study. The age of participants can affect performance due to auditory perceptual changes, such as a decline in temporal processing abilities that are age-related (Gordon-Salant, 2006; Walton, 2010). Care was taken to control for the effects of age by recruiting participants for the control group who were age-matched with participants from the cochlear implant group within five years to help to control for effects of age related changes in auditory processing. However, two individuals of similar age may show marked differences in performance on auditory perception tasks despite their similarity in age. Effects of age related changes on peripheral, central and cognitive processing are highly variable and dependant on several factors such as experience, environment and genetics (Gordon-Salant, 2006). Therefore, age-matching is an imprecise solution to ensuring a close match between temporal processing abilities in the CI and control groups. Simple temporal gap detection tasks such as those used in the current study have been shown to demonstrate age-related differences (Schneider, Pichora-Fuller, Kowalchuk & Lamb, 1994; Snell, 1997; Strouse, Ashmead, Ohde & Grantham, 1998). However, these differences are inconsistent and when a more complex temporal processing task (such as those used in the experimental sessions) is introduced, age related differences have been shown to be more consistent (Fitzgibbons & Gordon-Salant, 1994; He, Horwitz, Dubno & Mills, 1999; Lister, Besing & Koehnke, 2002; Gordon-Salant, 2006). In addition to the possible effects of age, evidence has been found that suggests listening with a cochlear implant may also result in differences in auditory temporal processing from normal68  hearing listeners. As previously discussed, listeners with CIs have demonstrated important differences on tasks measuring temporal integration of auditory stimuli (Shannon, 1983a, 1986, 1990; Donaldson, Viemeister & Nelson, 1997). In the current study, a test was used to measure temporal resolution (in the form of temporal gap detection) to screen for participants whose temporal processing abilities were outside of the normal range. In order to fully control for the effects of aging on auditory perception, and to investigate the effect of temporal processing of listeners with CIs on auditory continuity perception, a more thorough screening test battery for temporal processing may be required. The components of the stimuli used in the current study may be partly responsible for some of the variability between CI participants as well as the differences observed between the two groups. As mentioned previously, many of the CI participants commented that during the experimental session, the signal and the masker were difficult to tell apart. This presents the possibility that although the continuity experiment was designed to elicit a perception of heterophonic continuity (where the signal and masker differ in terms of their spectral characteristics), some of the CI participants may have perceived homophonic continuity (where the signal and masker are spectrally identical; Warren, 1984). All of the listeners with a Cochlear Ltd. System had a 180 Hz bandwidth at their test electrode; the listener with the Advance Bionics Inc. device had a 176 Hz bandwidth at the test electrode. Due to the bandwidths of the listeners with CIs’ device electrodes, the 200-Hz bandwidth masker was expected to stimulate adjacent electrodes. However, some energy will still spread to the bandwidths of adjacent electrodes (approximately 6 dB below the stimulating level at the target electrode by the centre frequency of the adjacent electrodes) even for pure tones. In other words, both the masker and signal used in the current experiment resulted in stimulation of adjacent 69  electrodes and overlapping populations of neurons and therefore may have sounded similar to the listeners with CIs. In addition, as previously discussed, the distance of electrodes from their stimulation sites and the amount and condition of auditory nerve fibres available for stimulation contribute further to spread of excitation and current summation. The lack of TFS information also contributed to the difficulty in distinguishing the pure tone and the masker. On the other hand, the pure tone had a flat envelope, which the masker had a noisy envelope, which could have provided a cue to the distinction of the signal and masker in the stimuli. With the information obtained from this initial investigation it is difficult to directly draw these conclusions, but each of these factors alone or in combination may have contributed to the difficulty the listeners with CIs described in separating signal and masker. The issues presented in this section can potentially be addressed by future investigations which take into consideration the limitations of the current study. 4.7  Future Research Many of the limitations of the current study could be addressed by future investigations.  Larger scale investigations may help to provide more data with less variability within the CI user group. Also, recruitment of more participants may provide enough data to analyse subgroups of participants based on subject characteristics which may affect continuity percepts, such as age, etiology, differing levels of auditory nerve survival, and age at time of hearing loss or implantation (including children and adults who were pre-lingually implanted). As previously discussed, there are many factors that contribute to clinical outcomes with cochlear implantation and a great deal of individual variability is thought to be attributable to the combination of these factors. In the future, investigation of an aging population of listeners with cochlear implants, specifically, may reveal important information about the effects of aging and age related hearing 70  loss on the perception of auditory continuity and help to determine how these effects may differ from or relate to those of cochlear implantation itself. The auditory perception of individual Listeners with CIs is extremely complex and although the exact role of each of the previously mentioned factors is not well understood, the degree of preservation of spiral ganglion cells in the individual implantee is generally considered to be of great importance in individual outcomes with a cochlear implant (Nadol, 1997; Shepherd, Meltzer, Fallon & Ryugo, 2006). Future studies could investigate individual implantee differences, such as auditory nerve survival, and their effects on perceptual outcomes, such as the ability to perceive auditory continuity, with cochlear implants. The amount of surviving ANFs is estimable through a combination of imaging techniques, evoked responses and behavioural techniques and is related to the etiology of the individual’s hearing loss, the duration of auditory deprivation and surgical trauma (Hinojosa & Marion, 1983; Cohen, et. al, 2003; Cohen, 2009; Earl & Chertoff, 2012). Loss or damage of spiral ganglion cells result in increased thresholds for generating action potentials along the auditory pathway, and in decreased spatial selectivity of the electrodes (Shepherd, et al, 2006). There is evidence that a larger number of functional spiral ganglion neurons is closely related to better auditory perceptual performance with a cochlear implant (Nadol, 1997; Zeng, 2004; Shepherd et al, 2006). By correlating estimated auditory nerve survival and performance on auditory continuity tasks, future studies have the potential to obtain further information about the effects of auditory deprivation on individual clinical outcomes with cochlear implants. In the current study it was not possible to control the activation of the individual listeners’ devices directly, therefore specific electrodes that were activated could vary across individual devices and presentation levels. In future investigations, the use of direct electrical 71  stimulation of electrodes would allow more accurate control of the stimulus signal. Processing features such as ADRO could also be selectively activated or deactivated in order to investigate their effects on perception of the stimuli. In future investigations, subgroups of participants could also be based on device characteristics such as implant manufacturer, processing strategies and features, device models and stimulation modes. ADRO, a default processing feature in modern Cochlear TM devices, intended to increase the audibility of speech by adjusting the gain of incoming signals to optimal levels within the dynamic range of the output, has been shown to improve the perception of speech in noise with cochlear implants (James, 2002; Blamey, 2005). The gain rules used in this feature, in addition to adjusting the gain of sounds near the listener’s threshold and comfort levels, also act upon sounds within the listener’s dynamic range that are classified as background noise to reduce these sounds in relation to a pre-determined target (James, 2002). It is possible that by adjusting the gain of potential environmental masking noise in relation to a perceived signal, ADRO may have an effect on auditory continuity perception in listeners with cochlear implants. The stimuli in the current study were not designed to investigate any potential effects of ADRO on auditory perceptual restoration, and therefore the current study cannot directly draw any conclusions about how ADRO may affect continuity perception for listeners with CIs. Future investigations could systematically study the effects of this adaptive gain adjustment on the perception of more complex auditory continuity percepts, for example by investigating phonemic restoration at quiet conversational levels or in the presence of low level noise. Such further research could help to provide more information about the effects of subject and device characteristics on the perception of auditory continuity.  72  The current study investigated auditory continuity perception under sound field conditions with monaural CI use. Auditory continuity has been shown to be affected by binaural factors, with continuity perception inhibited by interaural time differences (ITDs) and interaural phase differences (IPhDs) between the signal and masker (Thurlow, 1987; Kashino & Warren, 1996; Darwin, Akeroyd & Hukin, 2002). Different modes of device usage could be investigated, including bilateral implantees, bi-modal implantees who use a hearing aid in the un-implanted ear and users of hybrid devices that utilize acoustic amplification for residual low frequency hearing. Investigations of auditory continuity under sound field conditions with signal and masker of differing degrees of azimuth for monaural listeners, bilateral listeners with CIs, hybrid listeners and normal-hearing participants could provide more information about the perception of auditory continuity in realistic conditions under different usage modes. Bilateral implantation is becoming increasingly widespread and investigation into the perceptual advantages to bilateral modes of CI use and cost-benefit analysis of this practice is ongoing. Investigation of auditory continuity in bilateral and bimodal CI use could help to provide further information about the potential benefits of bilateral implantation practices and bimodal cochlear implant usage. More in-depth and specific investigation into the nature of the percept of auditory continuity experienced by listeners with CIs is also warranted. As an initial investigation, the current experiment was designed to test listeners with CIs on a very basic auditory continuity task in order to ensure that the study’s main goal (to demonstrate auditory continuity perception in listeners with CIs) was met. The main requirements for auditory continuity perception were investigated in listeners with CIs, namely the energetic masking requirement (Elfner and Caskey, 1965; Houtgast, 1972; Warren, 1972) and the edge masking requirement (Warren, 1972; Bregman & Dannenbring, 1977). However, only one aspect of the energetic masking 73  requirement, the effect of level on the perception of auditory continuity, was investigated in the present study. Another aspect of this requirement, the effect of frequency of the signal and masker on the perception of auditory continuity, was not investigated. As discussed earlier, in order for a discontinuous signal to be perceived as continuous (in a normally hearing participant group), the flanking signal must be within the frequency range of the masking noise (Elfner and Caskey, 1965; Houtgast, 1972; Warren, 1972). The manipulation of the frequency of the signal in relation to the frequency of the masker may be investigated in the CI population to determine what effect frequency manipulation has on their perception of auditory continuity and how it may differ from that of a normal-hearing control group. Listeners with CIs have limited frequency resolution when compared with normal-hearing listeners, owing to the limited number of electrodes transducing the signal within the cochlea and the broad spread of the electrical stimulation field. This results in wider auditory filter bandwidths and an impaired ability to detect frequency differences (Shannon, 1983a; Clark, 2003; Wilson and Dorman, 2008; Kwon and Van Den Honert, 2009). The effect of frequency can have important implications for listening in noisy environments, as much of the problematic noise that many listeners with CIs encounter is speech noise in multi-talker environments. In these situations, the signal (speaker of interest) and the masking noise (other talkers) typically overlap in frequency (Miller, 1947). However, it is possible that listeners with CIs experience masking even when the masker and signal do not overlap in frequency, because of their poorer frequency resolution. If listeners with CIs demonstrate differences from normally hearing listeners when frequency is manipulated, this may provide important information about their ability to listen in noise. Because of the importance of investigating speech perception in noise, future studies could also build on the work of Baskent (2012) who, as previously mentioned, has begun to 74  investigate phonemic restoration in conditions that simulate the auditory input received by listeners with cochlear implants. Future studies could examine possible correlations between performance on more complex phonemic restoration tasks and speech in noise performance in listeners with CIs. Such investigations may provide further insight into how auditory continuity perception may relate to speech in noise abilities for listeners with CIs. As discussed in previous sections, the perception of auditory continuity is closely related to patterns of auditory masking. The ability to perceive speech in noisy environments is also closely related to masking. It has been suggested that auditory continuity is a perceptual compensation for the effects of masking (Warren, 1984). It can be deduced that there may be a connection between perception of speech in noise and auditory continuity, although this has not yet been directly investigated in previous literature. Future investigation into how listeners with CIs experience auditory perceptual restoration may contribute to better understanding of auditory perceptual processes in listeners with CIs, as well as the relationship between auditory continuity and speech in noise. 4.8  Conclusion The current study investigated the perception of auditory continuity in a group of  listeners with cochlear implants. The findings show that listeners from the cochlear implant group perceived auditory continuity at masker levels at which simultaneous masking was effective. Because of greater masking experienced by listeners with CIs at the 49-dBA masker level, they perceived stronger continuity at this masker level than control listeners. No significant differences were observed at other masker levels or with the curtailed stimuli. This study provided some supporting evidence for the idea that larger amounts of simultaneous masking can be experienced by listeners with CIs compared with normal-hearing listeners at the same physical signal-to-noise ratio. These results provide the basis for future investigations of the 75  perception of auditory continuity, and about the relationship between this phenomenon and the perception of important sounds such as speech in everyday, noisy listening environments by listeners with CIs.  76  References Arts, H., Jones, D., Anderson, D. (2003). Prosthetic stimulation of the auditory system with intraneural electrodes. Annals of otology rhinology and laryngology supplement. 191: 2025. American Standards Association (1960). Acoustical Terminology, S1, 1-1960. New York, NY: American Standards Association. Badi, A., Kertesz, T., Gurgel, R. Shelton, C., Normann, R. (2003). Development of a novel eighth-nerve intraneural auditory neuroprosthesis. Laryngoscope. 113(5): 833-42. Bashford J, Warren R. (1987) Multiple phonemic restorations follow the rules for auditory induction. Perception and psychophysics. 42(2): 114-121. Baskent, D. (2012). Effect of speech degradation on top-down repair: phonemic restoration with simulations of cochlear implants and combined electric–acoustic stimulation. Journal of the association for research in otolaryngology, in press. Epub ahead of print, May 9, 2012. PMID: 22569838. Battmer R.D., Zilberman Y., Haake H., Lenarz, T. (1999). Simultaneous analog stimulation (SAS)–continuous interleaved sampler (CIS) pilot comparison study in Europe. Annals of otology, rhinology and laryngology. Supplement. 177:69–73. Bergeson, T., Pisoni, D., Davis, R. (2005). Development of audiovisual comprehension skills in prelingually deaf children with cochlear implants. Ear hear. 26(2): 149-164. Blamey, PJ. (2005). Adaptive Dynamic Range Optimization (ADRO): A digital amplification strategy for hearing aids and cochlear implants. Trends in amplification. 9(2): 77-98. Bonnet, RM., Boermans, PP., Avenarius, OF., Briaire, JJ., Frijns, JH. (2012). Effects of pulse width, pulse rate and paired electrode stimulation on psychophysical measures of 77  dynamic range and speech recognition in cochlear implants. Ear & hearing. 34(4):48996. Bregman, A. (1990). Auditory scene analysis: The perceptual organization of sound. Cambridge, MA: MIT Press. Bregman, A., Colantonio, C., Ahad, P. (1999). Is a common grouping mechanism involved in the phenomena of illusory continuity and stream segregation? Perception and Psychophysics. 61: 195-205. Buechner, A., Frohne-Buechner, C., Stoever, T., Gaertner, L., Battmer, D., Lenarz, T. (2005). Comparison of a paired or sequential stimulation paradigm with Advanced Bionics’ High-Resolution mode. Otology & neurotology. 26:941–947. Carlyon, R., Long, C., Deeks, J., Mckay, C. (2007). Concurrent sound segregation in electric and acoustic hearing. Journal of the association for research in otolaryngology. 8: 119-133. Chatterjee, M. (1999). Effects of stimulation mode on threshold and loudness growth in multielectrode cochlear implants. Journal of the accoustical society of America. 105: 850–860. Chatterjee, M., Peredo, F., Nelson, D., Baskent, D. (2010). Recognition of interrupted sentences under conditions of spectral degradation. Journal of the acoustical society of America. 127(2): 37-41. Chisolm T., Abrams H., McArdle R. (2004). Short and long term outcomes of adult audiological rehabilitation. Ear and hearing. 25: 464–477. Ciocca, V., Bregman, A. (1987). Perceived continuity of gliding and steady-state tones through interrupting noise. Perception and psychophysics. 42: 476-484.  78  Clark, G. (2003). Cochlear implants: Fundamentals and applications. New York, NY: SpringerVerlag. Cohen, L., Richardson, L., Saunders, E., Cowan, R. (2003). Spatial spread of neural excitation in cochlear implant recipients: Comparison of improved ECAP method and psychophysical forward masking. Hearing research. 179: 72-87. Cohen, L.T, Saunders, E., Richardson, L. (2004). Spatial spread of neural excitation: Comparison of compound action potential and forward masking data in cochlear implant recipients. International journal of audiology. 43(6):346-55. Cohen, LT., Saunders, E., Knight, MR., Cowan, RS. (2006). Psychophysical measures in patients fitted with Contour and straight Nucleus electrode arrays. Hearing research. 212(12):160-175. Cohen, L.T. (2009). Practical model description of peripheral neural excitation in cochlear implant recipients: 1. Growth of loudness and ECAP amplitude with current. Hearing research. 247: 87-99. Cooper, H., Roberts, B. (2007). Auditory stream segregation of tone sequences in cochlear implant listeners. Hearing research.225: 11-24. Cooper H., Roberts, B. (2009). Auditory stream segregation in cochlear implant listeners: Measures based on temporal discrimination and interleaved melody recognition. Journal of the acoustical society of America. 126(4): 1975-87. Darwin, C.J. (2005). Simultaneous grouping and auditory continuity. Perception and psychophysics. 67(8):1384-1390. Darwin, C.J., Akeroyd, M.A. and Hukin, R.W. (2002). Binaural factors in auditory continuity. Proceedings of the 2002 international conference on auditory display. Kyoto, Japan.  79  Delgutte, B. (1990). Physiological mechanisms of psychophysical masking: Observations from auditory-fibres. Journal of the acoustical society of America. 87: 791–809. Donaldson, GS., Viemeister, NF., Nelson, DA. (1997). Psychometric functions and temporal integration in electric hearing. Journal of the acoustical society of America. 101(6):37063721. Dorman, M.F., Loizou, P.C., and Tu, Z. (1998). The recognition of sentences in noise by normalhearing listeners using simulations of cochlear implant signal processor with 6–20 channels. Journal of the acoustical society of America. 104: 3583–3585. Duvall, A, Flock, A, Wersäll, J. (1966). The ultrastructure of the sensory hairs and associated organelles of the cochlear inner hair cell, with reference to directional sensitivity. Journal of cell biology. 29(3): 497-505. Eapen, R. and Buchman, C. (2009). Bilateral cochlear implantation: current concepts. Current opinion in otolaryngology & head and neck surgery. 17: 351-355. Earl, BR., Chertoff, ME. (2012). Mapping auditory nerve firing density using high level compound action potentials and high-pass noise masking. Journal of the acoustical society of America. 131(1): 337-352. Egan, J., Hake, H. (1950). On the masking pattern of a simple auditory stimulus. Journal of the acoustical society of America. 22: 622-630. Elfner, L.F., Caskey, W.E. (1965). Continuity effects with alternately sounded noise and tone signals as a function of manner of presentation. Journal of the Acoustical Society of America. 38: 543-547. Engstrom, H, Sjostrand, FS. (1954). The structure and innervation of the cochlear hair cells. Acta otolaryngologica. 44(5-6):490-501.  80  Fishman, KE., Shannon, RV., Slattery, WH. (1997). Speech recognition as a function of the number of electrodes used in the SPEAK cochlear implant speech processor. Journal of speech, language and hearing research. 40(5): 1201-1215. Fitzgibbons PJ, Gordon-Salant S. (1994). Age effects on measures of auditory duration discrimination. Journal of speech and hearing research. 37:662–670. Fletcher, H. (1940). Auditory patterns. Reviews of modern physics. 12: 47-65. Fredelake, S., Hohmann, V. (2012). Factors affecting predicted speech intelligibility with cochlear implants in an auditory model for electrical stimulation. Hearing research. 287:76-90. Friesen, LM., Shannon, RV., Baskent, D., Wang., X. (2001). Speech recognition in noise as a function of the number of spectral channels: Comparison of acoustic hearing and cochlear implants. Journal of the acoustical society of America. 110(2): 1150-1163. Frijns, JH., Briaire, JJ., Grote, JJ. (2001). The importance of human cochlear anatomy for the result of modiolus hugging multi-channel cochlear implants. Otology and neurotology. 22(3): 340-49. Fu, QJ., Shannon, RV., Wang, X. (1998). Effects of noise and spectral resolution on vowel and consonant recognition: acoustic and electric hearing. Journal of the acoustical society of America. 104(6): 3586-96. Fu, QJ and Nogaki, G. (2004). Noise susceptibility of cochlear implant users: the role of spectral resolution and smearing. Journal of the association for research in otolaryngology. 6: 1927.  81  Gerken, GM., Bhat, VK., Hutchinson-Clutter, M. (1990). Auditory temporal integration and the power function model. Journal of the acoustical society of America. 88(2): 767-778. Glasberg, B., Moore, B., (1990). Derivation of auditory filter shapes from notched-noise data. Hearing Research. 47: 103–138 Glass, I. (1983). Tuning characteristics of cochlear nucleus units in response to electrical stimulation of the cochlea. Hearing Research. 12(2):223-37. Gordon-Salant, S. (2006). Speech perception and auditory temporal processing performance by older listeners: implications for real-world communication. Seminars in hearing. 27(4): 264-68. He, N., Horwitz, A., Dubno, J., Mills, J. (1999) Psychometric functions for gap detection in noise measured from younger and aged subjects. Journal of the acoustical society of America. 106:966–978. Helms, J., Müller, J., Schön, F., Moser, L., Arnold, W., Janssen T... & Hochmair-Desoyer, I. (1997). Evaluation of performance with the COMBI40 cochlear implant in adults: A multicentric clinical study. Journal for oto-rhino-laryngology and it’s related specialties. 59(1): 23-35. Henkin, Y., Tetin-Schneider, S., Hildesheimer, M., Kishon-Rabin, L. (2009). Cortical neural activity underlying speech perception in post lingual adult cochlear implant recipients. Audiology and neurotology. 14:39-53. Hinojosa, R., Marion, M. (1983). Histopathology of profound sensorineural deafness. Annals of the New York academy of science. 405: 459-484. Houtgast, T. (1972). Psychophysical Evidence for Lateral Inhibition in Hearing. Journal of the acoustical society of America. 51: 1885-1894. 82  Howell, D. (2003). Fundamental Statistics for the Behavioural Sciences (5th ed). Belmont, CA: Wadsworth. James, C., Blamey, P., Martin, L., Swanson, B., Just, Y., Macfarlane, D. (2002). Adaptive dynamic range optimization for cochlear implants: A preliminary study. Ear and Hearing. 23: 49– 58. Kashino, M. and Warren, R.M. (1996). Binaural release from temporal induction. Perception & psychophysics. 58: 899-905. Kiang, N.Y. and Moxon, E.C. (1972). Physiological considerations in artificial stimulation of the inner ear. Annals of otololgy, rhinology and laryngology. 81: 714-730. Kidd, G., Feth, L. (1982). Effects of masker duration in pure-tone forward masking. Journal of the acoustical society of America. 72: 1384-1386. Keith, R.W. (2000). Random Gap Detection Test (RGDT). St. Louis, MO: AUDiTEC of St. Louis Ltd. Knapp, PH. (1948). Emotional aspects of hearing loss. Psychosomatic medicine. 10(04): 203202. Kwon, B., Van Den Honert, C. (2009). Spatial and temporal effects of interleaved masking in cochlear implants. Journal of the association for research in otolaryngology. 10: 447457. Lister J, Besing J, Koehnke J. (2002). Effects of age and frequency disparity on gap discrimination. Journal of the acoustical society of America. 111:2793–2800. Lorenzi, C., Gilbert, G., Carn, H., Garnier, S., Moore, B. (2006). Speech perception problems of the hearing impaired reflect inability to use temporal fine structure. Proceedings of the national academy of sciences of the United States of America. 103(49): 18866-18869.  83  Macmillan, N., Creelman, C. (2005). Detection Theory: A User’s Guide. (2ed). New York, NY: Lawrence Earlbaum Associates Inc. Middlebrooks, J. C. (2004). Effects of cochlear-implant pulse rate and inter-channel timing on channel interactions and thresholds. Journal of the acoustical society of america. 116: 452–468. Middlebrooks, JC, Snyder, RL. (2008). Intraneural stimulation for auditory prosthesis: modiolar trunk and intracranial stimulation sites. Hearing research. 242(1-2): 52-63. Miller, G. A. (1947). The masking of speech. Psychological bulletin. 44: 105–129. Miller, G. A., Licklider, J. C. R. (1950). The intelligibility of interrupted speech. Journal of the acoustical society of america. 22(2):167–173. Moore, B. (1978). Psychophysical tuning curves measured in simultaneous and forward masking. Journal of the acoustical society of America. 63(2): 524-532. Moore, B. (2004). An Introduction to the Psychology of Hearing. (5ed). London, UK: Elselvier Academic Press. Moore, B., Glasberg, B. (1983). Growth of forward masking for sinusoidal and noise maskers as a function of signal delay: implications for suppression in noise. Journal of the acoustical society of America. 73: 1249- 1259. Muchnik, .C, Taitelbaum, R., Tene, S., Hildesheimer, M. (1994). Auditory temporal resolution and open speech recognition in cochlear implant recipients. Scandanavian audiology. 23(2):105-9. Mulrow, CD., Aguilar, C., Endicott, JE., Tuley, MR., Velez, R., Charlip, WS... & DeNino, LA. (1990). Quality-of-life changes and hearing impairment: a randomized trial. Annals of internal medicine. 113: 188–194.  84  Nadol, JB. (1997). Patterns of neural degeneration in the human cochlea and auditory nerve: implications for cochlear implantation. Otolarynoglogy head and neck surgery. 117(3,1): 220-228. National Institutes of Health. (1995). NIH consensus conference on cochlear implants in adults and children. Journal of the american medical association. 274(24):1955-61. Nelson DA, Schmitz JL, Donaldson GS, Viemeister NF, Javel E. (1996). Intensity discrimination as a function of stimulus level with electrical stimulation. Journal of the acoustical society of America. 100:2393–2414. Nelson, DA., Donaldson, GS., and Kreft H. (2008). Forward-masked spatial tuning curves in cochlear implant listeners. Journal of the acoustical society of America. 123(3): 1522– 1543. Nelson PB, Jin SH. (2004). Factors affecting speech understanding in gated interference: cochlear implant users and normal-hearing listeners. Journal of the acoustical society of America. 115(5):2286-94. Nelson, P., Jin, S.-H., Carney, A. E., and Nelson, D. A. (2003). Understanding speech in modulated interference: Cochlear implant users and normal-hearing listeners. Journal of the acoustical society of America. 113:961–968. Oxenham, A. (2008). Pitch perception and auditory stream segregation: Implications for hearing loss and cochlear implants. Trends in amplification. 12(4): 316-331. Peterson, N., Pisoni, D., Miyamoto, R. (2010). Cochlear implants and spoken language processing abilities: Review and assessment of the literature. Restorative neurology and neuroscience. 28: 237–250.  85  Petkov, C., O’Connor, K., and Sutter, M. (2003). Illusory sound perception in macaque monkeys. Journal of neuroscience. 23: 9155–9161. Petkov, C., O’Connor, K., Sutter, M. (2007). Encoding of illusory continuity in primary auditory cortex. Neuron.54: 153-65. Petkov, C., Sutter, M. (2011). Evolutionary conservation and neuronal mechanisms of auditory perceptual restoration. Hearing research. 271: 54-65. Pettingill, L., Richardson, R., Wise, A., O’Leary, S., Shepherd, R. (2007). Neurotrophic factors and neural prostheses: potential clinical applications based upon findings in the auditory system. Transactions on biomedical engineering. 54(6,1): 1138-48. Pfingst, B. E., De Haan, D. R., and Holloway, L. A. (1991). Stimulus features affecting psychophysical detection thresholds for electrical stimulation of the cochlea. I: Phase duration and stimulus duration. Journal of the acoustical society of America. 90: 1857– 1866. Pfingst, B.E., Zwolan, T.A., Holloway, L.A. (1997). Effects of stimulus configuration on psychophysical operating levels and on speech recognition with cochlear implants. Hearing research. 112: 247–260. Plomp, R. (2002) The perception of multiple sounds. The Intelligent ear: on the nature of sound perception (pp.34-59). London, UK: Lawrence Erlbaum Associates. Ramekers D, Versnel H, Grolman W, Klis S. (2012). Neurotrophins and their role in the cochlea. Hearing research. 288(1): 19-33. Reike, L., Opstal, A., Formisano, E. (2008). The auditory continuity illusion: A parametric investigation and filter model. Perception and psychophysics. 70(1):1-12.  86  Ricketts, T., Grantham, D., Ashmead, D., Haynes, D., Labadie, R. (2006). Speech recognition for unilateral and bilateral cochlear implant modes in the presence of uncorrelated noise sources. Ear and hearing. 27(6): 763-773. Rodríguez, J., Neely, S., Patra, H., Kopun, J., Jesteadt, W., Tan, H., Gorga, M. (2009). The role of suppression in psychophysical tone on tone masking. Journal of the acoustical society of America. 127: 361-369. Roehm, P., Hansen, M. (2005). Strategies to preserve or regenerate spiral ganglion neurons. Current opinion in otolaryngology & head and neck surgery. 13(5): 294-300. Rogers, CF., Healy, EW., Montgomery, AA. (2006). Sensitivity to isolated and concurrent intensity and fundamental frequency increments by cochlear implant users under natural listening conditions. Journal of the acoustical society of America. 119(4): 2276-2287. Ronken, D. (1970). Monaural detection of a phase difference between clicks. Journal of the acoustical society of America. 47: 1091-1099. Samuel, A. G. (1981). Phonemic restoration: Insights from a new methodology. Journal of Experimental Psychology. 110(4): 474-494. Schneider BA, Pichora-Fuller MK, Kowalchuk D, Lamb M. (1994). Gap detection and the precedence effect in young and old adults. Journal of the acoustical society of America. 95:980–991. Shannon, R. V. (1983a). Multichannel electrical stimulation of the auditory nerve in man I: Basic psychophysics. Hearing Research. 11: 157-189. Shannon, R.V. (1983b). Multichannel electrical stimulation of the auditory nerve in man II. Channel interaction. Hearing research. 12: 1-16.  87  Shannon, R.V. (1985). Loudness summation as a measure of channel interaction in a cochlear prosthesis. In: Schindler, R.A., Merzenich, M.M. (Eds.), Cochlear Implants. (pp. 323334). New York, NY: Raven Press. Shannon, R. V. (1986). Temporal processing in cochlear implants. In: Collins, M.H., Glattke, T.H., Harker, L.A. (Eds.), Sensorineural Hearing Loss: Mechanisms, Diagnosis, Treatment (pp. 349–368). Iowa City, IA: University of Iowa Press. Shannon, R.V. (1989). Detection of gaps in sinusoids and pulse trains by patients with cochlear implants. Journal of the acoustical society of America. 85(6): 2587- 2592. Shannon, R.V. (1990). Forward masking in patients with cochlear implants. Journal of the acoustical society of America. 88(2): 741-744. Shepherd, R.K., Hatsushika, S., Clark, G.M., 1993. Electrical stimulation of the auditory nerve: the effect of electrode position on neural excitation. Hearing research. 66: 108–120. Shepherd, Robert K., N.E. Meltzer, J.B. Fallon, D.K. Ryugo. (2006). Consequences of deafness and electrical stimulation on the peripheral and central auditory system. In: Cochlear Implants. Waltzman, SB., Roland, JT. (eds). pp. 25-39. New York, NY: Thieme Medical Publishers Inc. Skinner, MW., Holden, LK., Whitford, LA., Plant, KL., Psarros, C., Holden, TA. (2002). Speech recognition with the nucleus 24 SPEAK, ACE and CIS speech coding strategies in newly implanted adults. Ear and hearing. 23(3): 207-223. Snell, K. Age-related changes in temporal gap detection. (1997). Journal of the acoustical society of America. 101: 2214–2220.  88  Stickney G., Loizou P., Mishra L., Assmann, P., Shannon, RV., Opie, J. (2000). Channel interaction and speech processing strategies for cochlear implants. Journal of the acoustical society of America. 108(5,2): 2601. Stickney , GS., Loizou, P., Mishra, L., Assmann, P., Shannon, V., Opie, J. (2006). Effects of electrode design and configuration on channel interactions. Hearing research. 211: 33–45 Stickney GS, Zeng FG, Litovsky R, Assmann P. (2004). Cochlear implant speech recognition with speech maskers. Journal of the acoustical society of America. 116:1081–1091. Strouse A., Ashmead D., Ohde R., Grantham D. (1998). Temporal processing in the aging auditory system. Journal of the acoustoustical society of America.104:2385–2399. Thurlow, W.R. (1957). An auditory figure-ground effect. American journal of psychology. 70: 653-654. Thurlow, W.R. (1987). Auditory continuity effects with binaural stimuli. Perception & Psychophysics. 42: 173-9. Thurlow, W.R. and Elfner, L.F. (1959). Continuity effects with alternately sounding tones. Journal of the acoustical society of America. 31: 1337-1339. Thurlow, W.R. and Erchul, W.P. (1978). Understanding continuity effects with complex stimuli. Journal of the American auditory society. 4: 113-116. Tyler, RS, Lowder, MW, Otto, SR, Preece, JP, Gantz, BJ, McCabe, BE. (1984). Initial Iowa results with the multichannel cochlear implant from Melbourne. Journal of speech and hearing research. 27(4): 596-604. Van den Honert, C., Stypulkowski, P.H. (1987). Single fiber mapping of spatial excitation patterns in the electrically stimulated auditory nerve. Hearing research. 29: 195–206. Vandali, A. Personal communication. January 1, 2011.  89  Vandali, A. E., Whitford, L. A., Plant, K. L., and Clark, G. M. (2000). Speech perception as a function of electrical stimulation rate: using the Nucleus 24 cochlear implant system. Ear & hearing. 21: 608-624. Veimeister, NF. (1974). Intensity discrimination of noise in the presence of band-reject noise. Journal of the acoustical society of America. 56(5): 1594-1600. Vicario, G. (1960). L’effetto tunnel acustico (The acoustic tunnel effect). Rivista di Psicologia. 54: 41- 52. Vinnik, E., Itskov, P., Balaban, E. (2010). A proposed neural mechanism underlying auditory continuity illusions. Journal of the acoustical society of America. 128(1):19-25. Walton, JP. (2010). Timing is everything: Temporal processing deficits in the aged auditory brainstem. Hearing research. 264: 62-69. Waltzman, S., Thomas, R. (2006). Cochlear Implants (2nd ed). New York, NY: Thieme Medical Publishing Inc. Warren, RM. (1970). Perceptual restoration of missing speech sounds. Science. 167: 392-393. Warren, RM. Perceptual restoration of obliterated sounds. (1984). Psychological bulletin. 96(2):371-383. Warren, R. M. (1999). Auditory perception: A new analysis and synthesis. Cambridge, UK: Cambridge University Press. Warren, RM., Obusek, CJ. (1971). Speech perception and phonemic restorations. Perception and psychophysics. 9: 358-362. Warren, R.M., Obusek, C.J., Ackroff, J.M. (1972). Auditory induction: Perceptual synthesis of absent sounds. Science 176(4039): 1149-1151.  90  Wegel, R., Lane, C. (1924). The Auditory masking of one sound by another and its probable relation to the dynamics of the inner ear. Physics Review. 23: 266-285. Wilson, B., Finley, C., Lawson, D., Wolford, R., Eddignton, D., Rabinowitz, W. (1991). Better speech recognition with cochlear implants. Nature. 352: 236–238. Wilson, B., Lawson, D., Zerbi, M., Finley, CC., Wolford, R. (1995). New processing strategies in cochlear implantation. American journal of otolaryngology. 16: 669–675. Wilson, BS. (2006). Speech processing Strategies. In: Cooper, H., Craddock, LC. Eds. Cochlear implants: a practical guide (2nd ed). London, UK: Whurr. pp. 21-69. Wilson, B., Dorman, M. (2008). Cochlear implants: Current designs and future possibilities. Journal of rehabilitation research and development. 45:5: 695-730. Yassin, I., Plack, C. (2005). The role of suppression in the upward spread of masking. Journal of the Association for research in otolaryngology. 6: 368-377. Yost, W. (2007) Fundamentals of Hearing: An Introduction (5ed). San Diego, CA: Academic Press. Zeng, F.G. (2004). Trends in cochlear implants. Trends in amplification. 8(1): 1-28. Zwicker, E. (1984). Dependence of post-masking on masker duration and it’s relation to temporal effects in loudness. Journal of the acoustical society of America. 75: 219-223.  91  Appendix A: Data Summary by listener Exp. 1: Continuity  92  93  94  Appendix B: Data Summary by listener Exp. 2: Masking  95  

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.24.1-0073270/manifest

Comment

Related Items