Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A laboratory and clinical study on vitreous fluorophotometry 1986

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


UBC_1986_A6_7 P36.pdf [ 7.78MB ]
JSON: 1.0085293.json
JSON-LD: 1.0085293+ld.json
RDF/XML (Pretty): 1.0085293.xml
RDF/JSON: 1.0085293+rdf.json
Turtle: 1.0085293+rdf-turtle.txt
N-Triples: 1.0085293+rdf-ntriples.txt

Full Text

LABORATORY and CL-INIGAL- STUDY C D X I VITREOUS F*I_.UOROF»HOTOMETRY by j^PANG, KIAH TIONG B.Sc., The University of Toronto, 1983 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF , MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Physics We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 1986 €> Pang, Kian Tiong, 1986 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by t h e head o f my department o r by h i s o r her r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f P H Y S I C S The U n i v e r s i t y o f B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date 3 0 t h A p r i l 1986 DE-6 (3/81) A B S T R A C T The optical, electronic sensoring and data acquisition systems were assembled and the software developed for a vitreous fluorophoto- meter which was then calibrated and used to quantify the integrity of the blood-retinal barrier in a pilot study of diabetic retinopathy and multiple sclerosis compared to normal controls. Breakdown of the blood-retinal barrier was quantified by measu- ring fluorescence in the vitreous at standard time intervals over one hour following intravenous injection of sodium fluorescein. The plasma dye concentration was measured throughout the procedure. Leakage was expressed as a penetration ratio of the average concen- tration at 3mm from the retina to the total plasma dye concentration. The results from diabetic subjects showing well defined stages of retinopathy severity demonstrated the proper functioning of the instrument by showing values in approximate agreement with retino- pathy severity, thus confirming the findings of previous observers. Of 16 multiple sclerosis subjects, results showed no s i g n i f i - cant difference between activity categories or level of current activity. Abnormally high penetration ratio was associated with active periphlebitis. A new finding was the presence of abnormally high leakage in 2 subjects showing no ocular signs of disease. Sub- jects without or with inactive periphlebitis showed breakdown of the blood-retinal barrier comparable in severity to diabetic subjects showing no or mild retinopathy. The vitreous diffusion constant of the dye for normal controls and multiple sclerosis subjects was not significantly different from that in water. T A B L E O F * C O N T E N T S ABSTRACT i i LIST OF TABLES v LIST OF FIGURES v i i ACKNOWLEDGEMENT ix Chapter I INTRODUCTION 1 1.1 Vitreous Fluorophotometry 1 1.2 Blood-Retinal Barrier 3 1.3 Applications 4 1.4 Multiple Sclerosis 4 1.5 Aim 6 Chapter II THEORY AND ALGORITHMS 7 2.1 Systematic Errors 8 2.2 Theory 12 2.3 The C-V Group 18 2.4 The L-A Group 21 2.5 Other Methods 24 Chapter III THE APPARATUS 27 3.1 Sodium Fluorescein 27 3.2 The Hardware 30 3.3 Other Material 39 3.4 The Software 42 Chapter IV CALIBRATIONS AND PROTOCOL 57 4.1 Instrument Preparation 57 4.2 Subject Preparation 63 4.3 Blood-Plasma Preparation 67 Chapter V ANALYSIS AND DISCUSSION 68 5.1 Classifications 68 5.2 F-Numbers 70 5.3 Intra-ocular Lengths 72 5.4 Plasma Curve-fits 75 5.5 LLoD 77 5.6 Autofluorescence 78 5.7 Profiles 82 5.8 Diffusion Constant 89 5.9 Penetration Ratio 95 i i i 5.10 LUND. BAS Results 104 5.11 Other Parameters 106 Chapter VI CONCLUSION 109 APPENDIX A COMPUTER PROGRAMMES I l l A.l DAS.PRN I l l A. 2 SCANMENU. BAS 113 A. 3 VITSCAN.BAS 114 A. 4 PLASCAN. BAS 116 A. 5 BATCHRUN. BAS 117 A. 6 REDUCE. BAS 118 A. 7 B/G.BAS 119 A. 8 MINUS. BAS 120 A. 9 SUDATA.BAS 120 A. 10 BLOOD. BAS 123 A. 11 C/VAZ.BAS 127 A. 12 SLOPES. BAS 130 A. 13 PLOT. BAS 133 A. 14 DRAW. BAS 135 A. 15 LUND. BAS 137 A. 16 F i l e Formats 140 A. 17 Sample CONSENT FORM 142 APPENDIX B MATERIAL USED 143 B. 1 Electronics 143 B. 2 Equipment 144 B. 3 Model Eye 145 APPENDIX C CALIBRATION RESULTS 146 C. 1 Pod-DAS 146 C.2 Logarithmic Amplifier 148 C. 3 pH Dependence 149 C.4 R/M-Log Amp-DAS 151 C.5 Attenuation 153 C.6 Performance Data 156 C.7 Model Eye Scan Profile 157 APPENDIX D GLOSSARY 158 APPENDIX E ABBREVIATIONS USED 160 REFERENCES 162 iv L - I S T O F * T A B L E S 1. Average F-numbers using a Gullstrands emmetropic model eye 17 2. Distribution of subjects 68 3. Detail subject classifications 69 4. Average F-numbers and their S.D 71 5. Average lengths of the intra-ocular media 72 6. Significance levels for intra-ocular lengths tests 74 7. Results of linear, least-squares f i t of axial length, and lens thickness in mm to the subjects's age 74 8. Plasma f i ts 75 9. Effect of the first blood sampling time, t * 1 ' 77 10. Average LLoDs and S. D. s 77 11. Lens autofluorescence, S.D. vs age 80 12. Autof luorescence, S. D. vs DRP states 80 13. Autof luorescence, S. D. vs MS states 81 14. D averaged over a l l measurement scans 90 15. Tests of D vs age and duration 91 16. Diffusion constant, D by sex-disease states 91 17. PR3 averaged over a l l 55-70 minute scans, after background subtraction only 96 18. PR3 Average • / - S.D. of the various groups 98 19. Significance level to reject H* between diabetic and normal groups 100 v 20. Significance level to reject H1 betveen MS and normal groups 101 21. Significance level to reject H* between diabetic and MS states 102 22. Results of LUND. BAS 105 23. Electronic components of blue circuit board of the DAS.. 143 24. Results of Pod calibrations 147 25. Results of Log Amp test 148 26. Water and buffer sample differences 149 27. Calibration results of the R/M-Log Amp-DAS 152 28. Concentration and gradients of attenuated samples 155 29. Performance characteristics of the VF system 156 vi L I S T O F " F I G U R E S 1. The "diamond" is the intersection of the beam and the probe 2 2. Diamond dimensions and effects 8 3. Diamond displacement vith s l i t lamp translation 16 4. Variation of F vith s l i t lamp movement 17 5. Different scanning axes 19 6. Structural formula of Sodium Fluorescein 27 7. Excitation and de-excitation peaks of fluorescein in blood 28 8. BLOCK Diagram of the VITREOUS FLUOROPHOTOMETER 31 9. The Slit Lamp 32 10. Excitor f i l ter holder-slide and intensity monitor 34 11. Pod asembly and circuit 35 12. Pin diagram of the blue circuit board in the DAS 36 13. Filter transmission profiles 40 14. Filter Overlap 41 15. FLOWCHART for REDUCE. BAS 48 16. Overhead viev of the angles at which the s l i t lamp was set for LEFT eye scans 62 17. Localizing techniques 64 18. Flowchart of the scanning procedure 66 19. Plot of F-number results 71 20. Intra-ocular lengths 73 v i i 21. Autof luorescence 79 22. Bolus effects 83 23. Sample profiles 84 24. Comparison of MS profiles 87 25. Comparison of diabetic subjects' profiles 88 26. Overall comparison 89 27. Diffusion constant vs age 93 28. Diffusion constant vs duration 94 29. Penetration Ratio 97 30. Cross-section of the model eye 145 31. Calibration curve of the Pod-DAS 146 32. Performance of the Log Amp 149 33. pH dependence 150 34. Calibration curve for R/M-Log Amp-DAS 151 35. Refraction at sample cel l surfaces 153 36. Attenuation in sample solutions 154 37. Model eye sample profile 157 v i i i A C K N O W L E D G E M E N T L A M E N T "Pangs of hunger, Pang?..." No! Pangs of loneliness. Pangs of being different; Pangs Of behaving differently. Pangs of being misunderstood. Pangs of having different ideas; Pangs of having different ideals. Pangs of being misunderstood. Pangs of being in a different vorld. Pangs of wanting to be the same; yet. Pangs of needing to be understood to be different. Pangs of being misunderstood. Pangs of alvays different interpretations. Pangs of very different experiences; Pangs of a different upbringing; Pangs of being misunderstood. Pangs of a small vorld; but, We are of this one world. It hurts. My Father's hopes are my reality. Thank you, my FRIENDS. Art thou for something rare and profitable? Wouldst thou see a truth within a fable? From "A Pilgrim's Progress" by John Bunyan. I - I N T R O D U C T I O N 1 . 1 Vitreous Fluorophotometry Vitreous Fluorophotometry <VF) is a cl inical research technique first described by Maurice in 1963 [11. Its objective is to provide a non-invasive, standardized, reproducible procedure for examining the integrity of the blood-retinal barrier. Initial uses included the investigation of diabetic retinopathy (DRP), a retinal vascular disease which can cause blindness. Several other retinal vascular diseases have subsequently been investigated. VF is a method of sampling the vitreous close to the retina to assess the state of intactness of the tissue. In the procedure, a fluorescent dye called sodium fluorescein is injected intravenously into a subject and its entry into the vitreous is measured by pro- jecting a beam of light into the vitreous compartment and monitoring the resulting fluorescence. A profile of the amount of fluorescence is established by scanning Blong the axis of the eye. By relating the intensity of fluorescence to the amount of dye present at each posi- tion along a scan, the total mass of fluorescein that has entered the vitreous is then a measure of the permeability of the blood-retinal barrier. The basic components of the fluorophotometer are: a) A source of focussed light to excite the dye. b) A probe to detect the amount of fluorescence. c) A photomultiplier/radiometer system to convert the signals. d) A data acquisition system to store the converted signals. 1 The light source usually employed i s a tungsten bulb that can be varied in intensity. Its output into the eye i s directed through a focussing system of lenses, and a s l i t . The probe consists of a fibre optic conduit that i B placed at the focus of the slit-lamp microscope objective 121. It i s posi- tioned at an angle to the beam so that i t i s focussed on a cross- section of the beam, detecting the fluorescence from the "inter- secting" volume called the "diamond". (See Figure 1.) The fibre optic probe conducts the collected light to a radio- metric detection system which then outputs to a data acquisition system. Recording and/or data-storing device(s) then store the con- verted signals. Figure 1. The "diamond" i s the intersection of the beam and the probe. 2 1 . 2 B l o o d - R e t i n a l B a r r i e r In the human eye, there are two barriers to the transfer of molecules into and out of the vitreous and aqueous media. These are the blood-aqueous barrier (BAB) and the blood-retinal barrier (BRB). The BRB is a situation of restricted permeability between the blood and the retina. It functions at the level of the retinal pigment epithelium (RPE) and the retinal vessels. Restricted permeability serves to maintain the regulated physical and chemical environment of the retinal neural tissues, i .e . , homeostasis of the retina. The RPE, considered the outer BRB C3], forms a uniform, conti- nuous, single layer of cells united laterally by zonulae occludens. (See Appendix 0.) The retinal vessels forming the inner BRB 131, are lined by a continuous layer of non-fenestrated endothelial cells which are joined near their luminal surfaces by zonulae occludens, a junctional complex in which there is complete fusion of the outer leaflet of neighbouring cel l membranes. The integrity or "tightness" of the BRB may be compromised by disease processes which affect any of i ts components - the RPE, the retinal capillaries, arteries and veins. In the present study, VF instrumentation and analysis tech- niques were developed and applied to an investigation of the BRB in multiple sclerosis (MS), an application that had not previously been investigated. The results were compared to normal controls and graded severities of DRP. 3 1 - 3 App 1 i c a t i.oris In VF studies of diabetes mel l i tu s , researchers found abnormal leakage of dye even when there was no v i s i b l e DRP [4,53. This suggested that VF would be useful to detect the onset of the break- down of the BRB. Diabetic persons with wel l established DRP showed large amounts of the dye i n the vi treous . These re su l t s implied that VF could possibly be used to monitor the leakage component of s u b c l i - n i c a l retinopathy progression to severe stages. The technique has since been used to study other r e t i n a l vas- cular diseases, for example, hypertension and pars p l a n i t i s [6], and the ef fects of drugs such as sulindac which has recently been found to be e f fec t ive i n reversing abnormal ear ly leakage i n diabetes [71. 1.4 M u l t i p l e S c l e r o s i s The l i m i t e d understanding of the disease processes i n MS ar ises from the lack of good corre la t ions between relapses and disease progression. This imposes d i f f i c u l t y i n the development of an effec- t i v e treatment. The perivenular c e l l i n f i l t r a t e i n the cerebrum that has been described as an ear ly event i n the formation of an MS plaque [81, appears s i m i l a r to the r e t i n a l perivenular i n f i l t r a t e found i n the eye. In laboratory studies of t i s sue preparations (immunoperoxidase), i t was found that there was abnormal r e t i n a l venous permeability i n areas with and without v i s i b l e p e r i p h l e b i t i s [91, and that such ef fects were much more frequent and extensive than previously cons i - dered. Engel l and Andersen [101 estimated that almost a l l MS patients 4 would develop retinal periphlebitis at some point in their lifetime. If this is correct, i t can be expected that following the onset of periphlebitis, there is breakdown of the BRB (to molecules much larger than fluorescein) which persists despite cl inical and histo- pathologic resolution of the lesions. Clinically, retinal periphlebitis in MS may affect one, several or a l l of the retinal veins and appears either as an active lesion with patches of fluffy white haziness surrounding the veins or as an inactive venous sclerosis when there is halo sheathing. The course is mild, asymptomatic and transient with activity lasting weeks, months or up to two years. Resolution leaves no sequelae or else replacement sclerosis. Inflammatory activity can be confirmed clinically by fluo- rescein angiography photography which shows leakage, whereas inactive venous sclerosis shows no leakage [11,12]. Photographic information, however, is limited by the sensitivity of the emulsion and would not show a subtle breakdown of the BRB that might persist after reso- lution of inflammation. A significant proportion of MS patients with active periphle- bitis also show abnormal brain scans when compared to those with inactive periphlebitis [13]. An explicit relationship between the activity at the two sites - the BRB and the blood-brain barrier with regards to relapses and disease progression has not been inves- tigated. 5 1 . 5 A i m The objectives of t h i s study are the f o l l o w i n g : (1) To use VF as a sens i t ive system to study possible BRB changes i n MS cases showing act ive and inac t ive p e r i p h l e b i t i s with regard to quant i ta t ive differences i n leakage. A p a r a l l e l study on persons with diabetes i s conducted as a contro l on the performance of the VF system. (2) To document the frequency and sever i ty of leakage i n r e l a t i o n to the c l i n i c a l grading of the cer ta inty of the diagnosis, and i n r e l a t i o n to the the standard c l i n i c a l a c t i v i t y categories with the aim of e s tab l i sh ing re la t ionships between the ocular and centra l nervous system a c t i v i t i e s , p a r t i c u l a r l y relapses and disease progres- s ion. (3) To assess the u t i l i t y of the procedure as a non-invasive technique i n the diagnosis of MS, and as an i n d i r e c t monitoring method of grading the centra l nervous system a c t i v i t y . 6 I I - T H E O R Y A N D A L G O R I T H M S In this chapter, the details of the VF technique are elabo- rated. The algorithms proposed by two groups of investigators were closely followed in this experiment in order to compare results. The methods of these two groups are detailed in the following sections. The first section explains the "systematic errors" inherent in each VF scan. These errors arise from the limitations of the measu- ring instruments, as well as from the complex optical, biological system of the human eye. Thereafter, there is an explanation of the earlier, more basic models used in analysis. The second set of sections discusses the two more elaborate methods of analysis that were used in this study. The algorithm to accommodate the "systematic errors" was proposed by J.G. Cunha-Vaz and co-workers and is referred to as "The C-V Group".•• The state of the BRB is then embodied in a single number called the Penetration Ratio. "The L-A Group" described a more mathematically involved solution to find the Permeability and Diffusion coefficients. This algorithm is due to H. Lund-Andersen and co-workers. In the development of the VF system, alterations to the proto- col and the algorithm necessarily occur because of differences in procedures and instrumentation. Some modifications to the methods of the two groups are discussed in Section 2.5. * * " . . . Group" refers to a general geographical distribution of the various investigators, and is also used to distinguish between those using one method of analysis and those using the other. 7 2 - 1 Systematic Errors Figure 1 shows that the "diamond" i s of f i n i t e dimensions. As i t i s moved across an interface from a compartment of high dye con- centration into a compartment without dye, a non-zero signal i s obtained from the empty compartment. One such interface i s at the choriod-retina <CR) and the vitreous, where this non-zero signal effect i s especially significant during the f i r s t few minutes after the introduction of fluorescein into the blood. This "false" signal i s due to the depth or length of the diamond as illustrated below. It persists for some distance into the dye-free compartment. This effect i s sometimes called the "tailings" or "spread function" due to the associated peak because i t s strength depends on the peak signal at the interface [14.]. Choriod Posterior I Retina Vitreous D ' d*cos 8 g + ?*cos 9t V : 8.2512*11 0 P e a k a. Estimates of Axial Resolution and diamond volume. b. The "tailings" due to the choroid-retinal peak [18]. Figure 2. Diamond dimensions and effects. 8 There are other sources of "false" signals, especially very close to the retina. One effect is halation [15]. When focussed on the retina, the edges of the s l i t are not distinct. This is due to the transparent depth of the retina (about 0.5mm), and to scattering. There is also a possible dependence on retinal pigmentation [161. There may be signals from light reflected off the walls of the vitreous cavity although i t is unlikely that, away from the retina, the reflected light intensity can be sufficiently high, or be in the direction of the probe. This is also apparent from the relative volumes of the diamond and the vitreal cavity. As tailings are strongest near the retina, the solution is to consider data that are collected at a more remote point where the spread function is small. However, the retina and its associated CR peak must always be included in any scan as they constitute a refe- rence zero-position by which displacement can be measured. The distance from the retina within which data cannot be accepted depends on the dimensions of the diamond. The in vivo axial resolution (AR) is defined as the distance from the CR peak where the signal is a small fraction of the peak [141. The fraction chosen must not be so small that the signal is at the noise level of the detector. A practical definition of AR is the ratio of the signal at a fixed distance from the retina to the CR peak signal [171. Figure 2a estimates the AR and the volume of the diamond. They depend on the angle between the directions of the excitation beam and probe, the width of the s l i t , and the diameter of the probe. These parameters must be varied until an optimized AR is attained [181. 9 Reducing the s l i t width and/or using a smaller probe diameter need not necessarily result in better ARs, as a reduction of either or both reduces the amount of fluorescence detected. The noise of the detection system then limits the reduction that is possible. However, i t is necessary to have signals distinctly above the noise levels so that the data can be analyzed with greater confidence. Also, a small s l i t width cannot be measured (i .e. , calibrated) with precision. (Refer to Sections 3.2, 4.1 and Figure 17.) Increasing the angle shown in Figure 2a improves the AR and reduces the volume of the diamond. However, this angle is limited by the diameter to which the pupil of the eye can be dilated. The cornea's curvature and its varying thickness also distort the finite beam and s l i t , thereby reducing the probe's focus on the beam. The formulae in Figure 2a are therefore approximations as the in vivo dimensions of the diamond cannot be measured directly. The maximum dilation diameter of the pupil varies among sub- jects. When the pupil cannot be dilated to a minimum acceptable diameter, the procedure cannot be used. The power of the convex cornea curvature, however, can be offset by a plano-concave lens placed on the cornea. The plane surface of the lens provides a "window" to view the fundus. (See Figure 3.) Figure 3 also shows the importance of AR. The angle between the beam and the probe in the vitreous varies during a scan. The AR changes with i t . (See Figure 2a.) The AR is larger in the posterior vitreous than in the anterior segments. As the larger AR near the CR causes more severe tailings, i ts optimization is thus vital to the 10 instrument's performance. Many biological substances (tissues) are known to be auto- fluorescent. They fluoresce at certain incident wavelengths. The crystalline lens and the cornea (as well as the retina) are auto- fluorescent. They give off false signals in their vicinity (i .e. , tailings), B S well as absorb part of the excitation beam which must pass through them. The autofluorescence of the crystalline lens is unavoidable when the vitreous is to be scanned. It depends on the age of the subject as well as the disease process [19]. Cataracts (or opacities) in the lens also cause loss of input intensity, which limits the application of VF to eyes with clear media. In the later scans, there is usually a high concentration of dye in the anterior chamber which can cause a loss of incident light due to absorption, or stray signals from scattering. However, the expected levels of fluorescein in the aqueous at 60 minutes after injection are sufficiently small that they do not attenuate the excitation beam appreciably. (See Appendix C.5.) •ther sources of error are related to the apparatus. In Section 3.1, where the properties of fluorescein are discussed, the problem of f i l ter overlap will be mentioned. Problems in the radiometric system such as dark current noise due to random photon events in the photomultiplier tube (PMT) are minimized by constantly checking instrument zero adjustments. Al l the above sources of variation are affected by the inten- sity of the excitation beam, which affects the amount of scattering 11 and therefore autofluorescence, tailings, and AR. Hence, optimi- zation of the apparatus involves the adjustment of a l l parameters which contribute to the quality of the data acquisition. 2 . 2 T h e o r y The permeability of the BRB is related to the diffusion of the dye across i t . Concentration differences and electric potentials are some of the forces driving the diffusion phenomenon. In the first two hours after the injection of the dye, passive diffusion governs its penetration through the BRB from the blood into the vitreous [ 2 0 , 2 1 ] . This means that Fick's Law, in which a concentration gradient is the driving force, can be applied. Hence, to measure the permeability of the BRB, the change in the blood-dye concentration over time must be known. The concentrations of dye on both sides of BRB may then be related by a proportionality constant which represents the permeabi- l i ty . The simplest mathematical model is that of a plane retina [ 2 2 ] . The equivalent one-dimensional problem is (1) D * d f tc(r,t)/drB = dc<r,t)/dt , where the concentration, c, is position-dependent and time-dependent, r is the distance from the middle of the vitreous. D is the diffusion coefficient which is assumed to tbe independent of c, and hence, may alternatively be called the diffusion constant. The assumptions of such a model are that (a) fluorescein can only diffuse towards the middle of the vitreous after penetrating the 12 BRB plane, (i.e., a one-directional transport process); and, (b) the boundary conditions, which are: (2) dc(0,t)/dr = 0 , (3) D * dc(a,t)/dr = P lc p<t) , where the retina i s at a distance a from the mid-vitreous, and c" i s the concentration of fluorescein in the blood at time t. Like D, the diffusion constant, P1, the "permeability coefficient" i s independent of c, and i s referred to as the permeability index [23]. The solution to Eq. 1, using Eqs. 2 and 3 i s (4) c(r,t) dr r - t T • t c p (T) dT T • a P1 relates the total amount of dye in the blood available to pene- trate the BRB at the post-injection (p.i.) time, t, to the total amount in the vitreous at t. Mathematically, the numerator i s the area under the concentration profile (taken at t) between the mid- vitreous and the retina. The denominator i s the area under the plasma-fluorescein concentration profile up to t. Several problems arise in solving Eq. 4. The processes at work at the BRB are not simple. It has been established that "active transport phenomena" drive solutes against concentration gradients [20,21], The forces driving these active transport processes are, in fact, greater than those for passive diffusion by about 31 times. 13 They come into effect after the first two hours p . i . Unless these outward active processes are to be studied the last scan is usually taken at approximately one hour p . i . The time taken for the dye to reach the eye depends on the site of injection, e.g., the dye appears at the retina 10s sooner when i t is injected into a carotid artery than when i t is injected into a peripheral vein in the arm [24,25]. Injecting the dye slowly or quickly also affects the time of appearance at the BRB and the profile of the bolus which is attenuated even after a fast intra- venous injection due to mixing with blood. Venous blood samples are drawn throughout the procedure to obtain the plasma profile. The first sample is usually taken after several minutes p . i . and the number of samples required depends on the method of solving the integral, e.g. more samples are needed within the 60 minutes i f the area is found using the trapezoidal rule. Assumptions must also be made about the profile between 0 and t ' 1 1 , the f irst sampling time when the latter method is used [261. Curve-fitting techniques may be used in solving the plasma integral. In pharmacokinetic studies [27], i t has been found that the time-course of fluorescein in the blood is best approximated by a two-compartment model of mixing. This requires a mode of curve- fitting a sum of two negative exponentials to the data [28,29] which includes a fast and a slow decay in the levels of fluorescein in the blood. (Refer to Section 3.1.) The integral in the numerator in Eq. 4 assumes in its lower limit, that dye penetrating the BAB and dye penetrating the BRB have 14 not mixed within one hour p . i . 127] Tailings of the CR peak res- trict the upper limit of the integral as previously mentioned. How- ever, i t can be expected that most of the fluorescein in the vitreous is in the vicinity of the retina as will be explained. Attenuation of the excitation source by fluorescein in the anterior chamber was briefly mentioned in the previous section. Because of its small volume, the dye quickly f i l l s the anterior chamber by way of the i r i s and ciliary body. Assuming a uniform distribution (because of its small volume), the attenuation may be expressed in the form of the Beer-Lambert law: (5) c(measured) = c(true) * exp< - b » cft * d ) . This assumes that only the concentration, c", in the aqueous chamber of depth, d, attenuates the excitation beam. The dye in the anterior segment of the vitreous (adjacent next to the posterior surface of the lens) and lens autofluorescence are not included. b is the attenuation (or extinction) coefficient [183. This attenuation is less than 10% for c"<1000' at the one hour scan. (See Appendix C. 5. ) Figure 3 below shows how rays are refracted at the interfaces in the eye. As the diamond is moved anteriorly along the optical axis, the angles of incidence change at the various surfaces so that AR is also position-dependent. Figure 3 also demontrates that a 1-mm translation of the s l i t lamp does not correspondingly produce a 1-mm diamond displacement in the medium in which i t is focussed. It is thus necessary to translate 15 Figure 3. Diamond displacement vith s l i t lamp translation. s l i t lamp translations to displacements of the diamond since the latter are not directly measurable. S l i t lamp translations, d are related to diamond displacements, x, by [30] x = F * d . The "F-number* represents the effects of refraction at each inter- face. F i t s e l f depends on x. Table 1 and Figure 4 show the results for a model eye by Krogsaa, et al [30]. Note that F i s different in the compartments because there are'fewer interfaces to traverse as the diamond i s moved towards the cornea. Also, F>1 in a l l cases demonstrating the power of the ocular system. 16 Model Eye Slit Lamp Compartment Distances Movements F-number in mm in mm Aqueous Chamber 3.60 2.49 1.45 Crystalline Lens 3.60 2.20 1.64 Vitreous Chamber 16.97 11.75 1.44 Table 1. Average F-numbers using a Gullstrands emmetropic model eye. [301 1.6 - 1.5 - 1.4 • AQUEOUS LENS VITREOUS SLIT LAMP TRANSLATIONS Figure 4. Variation of F vith s l i t lamp movement. [301 17 2 . 3 T h e C — V G r o u p A simple way to quantify the permeability of the BRB is to measure the amount of fluorescein in the posterior vitreous segment. However, different instruments with different sensitivities and ARs, measure different strengths of CR tailings. Therefore, a number that quantifies the permeability must be independent of instrument diffe- rences. To offset errors due to CR peak tailings, a background, pre- injection scan is subtracted from a l l other post-injection (p.i.) scans. Lens autofluorescence is also eliminated by this subtraction. The subtraction is carried out by aligning "landmarks" such as the lens and CR peaks of each scan. This method, however, does not consider possible shifts in the CR peak position with time, i . e . , the CR peak may not be the true position of the retina in the later scans. As fluorescein is continuously removed from the blood, a point may be reached when the signal from the posterior vitreous is greater than that from the CR (in high leakage cases). That is, the CR peak appears in front of the retina. Also, due to different fixation or scanning axes, the distances between peaks may vary. An alternative method is to align only the retinal position. The retina may be located visually at the start of each scan. This assumes that the starting points of each scan are at the same loca- tion on the retina. The macula, for example, may be used as such a starting landmark. An advantage of this method is that errors due to different alignments (shown in Figure 5b) are smaller near the CR. 18 a. Different starting points. b. Macula alignment. Figure 5. Different scanning axes. However, the data collected from the anterior half of the eye may no longer be analyzed vith confidence. The d i f f i c u l t y in this method i s the accuracy with which the CR can be located due to halation, scattering, and the centring of a small s l i t on a larger probe. In order to account for the possible effects of CR tailings in the later scans, a "bolus" scan i s made within 3 to 5 minutes p.i. The dye i s not expected to have penetrated the BRB (in most cases) within the f i r s t 10 minutes p.i. The bolus scan then provides the strength of the CR tailings at specified distances from the retina after the background has been subtracted. The distance from the retina at which calculations are done i s usually the 3-mm point. It i s expected that the ARs of most instru- ments are smaller than this [311. Thus, CR tailings should not be 19 very significant at this point. Any point where the tailings are not substantial can be chosen. However, to study leakage from the BRB, i t is important to scan close to the retina. Therefore, 3mm from the latter is a suitable choice (after AR considerations). The CR tailings are subtracted from later measurement scans** only i f the former are significant. Ishimoto, et al [323, suggested a criterion for this bolus CR correction with a recommendation for implementation i f (6) (CR Peak value)/(3-mm Vitreous value) > 10 . The above equation was based on the commercial Fluorotron* Master fluorophotometer. The condition may be different for other instru- ments. The measurement scans are taken at specified time intervals before the outward transport processes become significant. After alignment and background subtraction, (and bolus CR correction, i f necessary), the average value of the dye concentration around the 3- mm position is found [323. The result is divided by the plasma inte- gral up to the p . i . time of the scan. The final result is called the penetration ratio, PR3, of units, s"1. The advantage of using PR3 instead of P1 in Eq. 4, is that the 3-mm value from the retina may be used for comparison between diffe- rent patients and instruments. There are no problems of CR tailings, and i t is not required to integrate very close to the CR. * * Al l scans other than the pre-injection and bolus scans are referred to as measurement scans. 20 The averaging i s carried out between the 2-mm and the 4-mm points E323. This reduces errors usually caused by random fluctua- tions in the data which can persist, even after subtraction, for a penetration ratio at the 3-mm point, PR3*. The averaging also dimi- nishes the errors due to the alignment problems mentioned before. 2 . 4 T h e I_ — A G r o u p Much of the scanning and data-correction techniques used by this group are the same as that of the C-V Group (by virtue of similar instrumentation). However, instead of using PR3 or P1, a more elaborate, mathematical model of the transport of fluorescein in the eye i s constructed to estimate the vitreous diffusion constant, D, and the permeability constant, P, of the BRB E33], The analysis involves solving the diffusion equation, ( 7 ) V • (D7c) = 6c/6t , as a boundary-value problem directly. The d i f f i c u l t y in such an approach i s that fluorescein i s being continuously removed from circulation. This means that the transport problem i s a "transient" one. Several assumptions of the previously mentioned plane-retina model of the C-V group apply to this model. (1) Dye from the BAB and from the BRB have not mixed uniformly at 60 minutes p.i. (2) The retina i s assumed to be spherical with radius of curvature a so that the diffusion i s radially inward, towards the centre of the vitreous chamber E233. Furthermore, i f D i s independent of c, then 21 Eq. 7 reduces to an r-dependence only: (8) (D/r)-{ 2 6/6r • r 6*/6r« } c(r,t> = dc(r,t)/6t , for 0 £ r £ a Due to symmetry, the boundary condition i s (9) oc(0,t)/6r = 0 . The i n i t i a l condition i s (10) c(r,0) = 0 , for 0 < r < a Thus far, the model assumes that the concentrations on both sides of the BRB are related through a proportionality constant, P which represents the permeability of the entire BRB. P makes no reference to the location of breakdown within the BRB (inner or outer). Thus, at any time, t, the amount of dye available for diffu- sion towards the mid-vitreous, depends on the amount of dye c(a,t) that has already penetrated the BRB. But c(a,t) depends on c"(t) by assumption. The barrier condition i s then (ID - D 6c(a,t)/6r = P-( c(a,t) - C (t) ) , (12) { 1 + <D/P)-6/or ) c(a,t) = C<t) . Using the method of Laplace transforms, the solution i s (13) c' (r,s) = c''<s) F' <r,s) where 22 Q4) F'(r,s) = u(r) sinh v(r,s) , § sinh w(a,s) • J"s cosh w(a,s) <15) u(r) = aP/r/D , for 0 < r S a , (16) w(r,s) = W<s/D) , for 0 < r < a , (17) 5 = u(a) - /D/a . s i s the transform variable; prime implies the transformed functions. Transforming back to t-space, a slowly-convergent series results for small t C33], However, i f Eq. 14 is expanded for large real part of s, and then inverted, the approximate solution to Eq. 7 i s (18) c(r,t) * c" (t-t) F(r,T) d T , T - • where (19) F(r,T) = u<r)-( (exp<-M<-1)8) - exp(-M(1)*))//(in ) - 5-exp(5 2T)•(exp(2g(T)M(-l))-erfc(g<T)+M<-1)) - exp(2g(T)«(l))•erfc(g(T)*M(l))) } , and (20) W(j) = M(r,T,j) = <a-jr)/(2V(DT)) , j = - l , l (21) g ( T ) = 5/T . (22) erfc(x) = (2/Vn) exp(-u*) d p i j II • • Eq. 22 i s the complementary error function. Eq. 18 may then be used as the theoretical solution in a non- linear curve-fitting calculation to the experimental data. The good- ness of f i t may be tested by k - N (23) S8 = E <cMr(k),t) - c(r (k), t))» • 6 (k)° , k - > where 23 (24) e(k)» = max « fl" , c*<r(k),t) 4) serves as a 'weighting factor', and. (25) fi a lowest concentration that can be detected . c*(r(k),t) i s the concentration measured at (a-r(k)) mm from the retina, of the scan taken at time t. The index k represents the order of the data-points along the axis of a scan. Non-linear curve- f i t t i n g i s then carried out by the Marquardt algorithm [34]. A search through the parameter space of P and D- i s done and the best f i t i s determined when S* is a minimum for one set of P and D values. 2. 5 Other- Methods The algorithms used by other groups of investigators are usually variations of those discussed above. For example, Eq. 19 may be simplified for computational purposes [23]. Most modifications, however, arise owing to differences in instrumentation which necessi- tate differences in protocol and algorithm. Different methods have been proposed for the bolus CR peak correction mentioned in Section 2.3. Since many investigators use the Fluorotron* Master, one method of correction suggested for this apparatus was to multiply the bolus profile (between 2 and 4 mm from the retina) by the ratio of the CR peaks of the measurement and the bolus scans [153. The modified bolus profile was then subtracted from the measurement scan. This seems reasonable as tailings depend on the peaks causing them. Bursell, et al [353 argued that this algorithm over-corrected 24 the errors due to tailings because the CR peak values included fluo- rescence from the vitreous because of the f i n i t e diamond. (See Figures 1 and 2.) The correcting ratio should in fact be slightly smaller than that between CR peaks. However, the variations of the dimensions of the diamond in vivo cannot be determined, and the appropriate correction factor i s unknown. It has been found that retinal blood flow increased by 40 to 70X in the transition from light to darkness [36]. Although this i s not l i k e l y to affect most scans, except possibly the bolus and the plasma integral, some investigators use more intense sources of excitation such as xenon flash tubes to attain signal levels v e i l above the dark current noise so that room lighting need only be dimmed and not completely turned off [37], For instruments that scan continuously, i.e., collect data con- tinuously along the scanning path, the signals from adjacent positions overlap because of the f i n i t e diamond. The methods used in "smoothing" the signals (integral and curve-fitting methods) are then important [38]. In other instruments that employ the "spot" method using "chopped" or flash excitation sources, data are collected at specific points along the scanning axis, e.g. at every 1mm interval. These tvo methods of data collection determined by characteristics such as AR also determine whether PR3 or PR3» i s to be employed. To compare the results obtained by various fluorophotometers, a set of instrument characteristics i s used to describe each instru- ment's capabilities. This set of performance data includes para- meters such as angle between the beam and the probe directions, AR, 25 f i l t e r overlap, lower limit of detection (LLoD), reproducibility (R) and error of measurement < EoM). The in vivo LLoD may be defined as the lowest detected (or detectable) concentration plus twice i t s standard deviation [32]. Practically, i t means that, for example, at the mid-vitreous of a background scan, the LLoD i s the average value of the concentration detected in that region plus twice the standard deviation of that average. The sensitivity of the detection system i s defined as the ab i l i t y of the detection system to differentiate changes in adjacent concentration volumes. It i s invariably dependent on the diamond and the ambient concentration. (See Appendix C.6.) R and EoM also have the same dependences. Their definitions are: (26) EoM = { c(measured)/c(true) } - 1 ; (27) R = standard deviation of repeated measurements. The problem with these definitions i s that the in vivo "true" concen- trations cannot be determined. 26 I I I . T H E A P P A R A T U S 3 . 1 S o d i u m F l u o r e s c e i n Sodium fluorescein was f i r s t synthesized from resorcinol and phthalic anhydride in 1872 [391. It i s a very weak dibasic acid with a molecular weight of 376.27. Its solubility i s increased as a sodium salt. Na* -0 C00-Na* RESORCINOLPHTHALEIN SODIUM - C..Hlt0,Na« Figure 6. Structural formula of Sodium Fluorescein E40], Its suitability as an indicator in ophthalmological research i s due to the fact that the peak excitation wavelength (490nm) i s di f f e - rent from the peak emission wavelength (520nm). In addition, the de- excitation time i s short: approximately 4ns. Hence, with a suitable combination of f i l t e r s to separate the two wavelengths of light, the concentration may be deduced from the amount of fluorescence. 27 1 2 4000 5000 6000 7000 Wavelength (AngstrSm Units) Figure 7. Excitation (1) and de-excitation (2,3,4) peaks of fluorescein in blood [25]. The choice of f i l t e r combinations, however, i s made d i f f i c u l t by the fact that the ranges of excitation and emission wavelengths shift towards the red end of the spectrum when fluorescein i s mea- sured in blood compared to measurements in water solutions [413. This effect may be caused by multiple scattering, absorption and autofluo- rescence of the tissues that are scanned. The optimum "cross-over* point for the f i l t e r combination should be about 525nm. Fluorescein diffuses readily from the blood into a l l extra- cellular fluids except across the retina (BRB), and the brain (blood- brain barrier). As i t i s a weak acid, i t does not bind with (or stain) normal v i t a l tissues, and i s highly fluorescent in alkaline media. In aqueous solution, about 80*/. of incident light i s converted to fluorescent radiation [39,403. However, the dye only returns approximately 2G7. fluorescence when dissolved in blood. This loss i s due to binding to proteins (serum albumin) and red blood c e l l 28 membranes. Another effect is quenching by the haemoglobin. The absor- ption spectrum of haemoglobin is about identical to that of fluores- cein. This can be demonstrated in severely anaemic patients vhere there is a stronger fluorescence as proportinately less of the dose is quenched by the haemoglobin. By means of equilibrium dialysis or ultra-filtration [40], i t is estimated that between 50 to 84% of fluorescein is bound. It is however, the unbound fluorescein (17%) that diffuses across cellular membranes and the BRB i f i t is disrupted. This effect must be taken into account when analyzing plasma scans. Fluorescein has low toxicity which is probably due to its inability to bind with vital tissues [393. In animal experiments, lethal doses were at 2 to 3 grammes per kilogramme body weight. In this VF study and other investigations on human subjects, the dose administered is calculated at 14"1 body weight using pharmaceu- t i c a l ^ prepared ampoules of 25% concentration (2.5"1). Fluorescein is well tolerated but there are occasional side effects such as transient nausea or vomiting immediately after injec- tion. Yellowish tinting of the skin lasts for several hours after injection and the urine is yellow for about two days. Allergic reac- tions are rare. (See Appendix A.17.) Other important dependences of its fluorescent property are on the pH, concentration and temperature. Only in an alkaline medium is its fluorescent property enhanced [39,40]. It was found that pH 7.4 is the level at which the dye fluoresces most efficiently [42]. This is approximately the pH of the cellular fluids of the body which 29 varies l i t t l e . Importantly, the pH of the calibration sample solu- tions as v e i l as the buffer for diluting plasma samples must be specially prepared. The dependence on the ambient concentration i s a result of scattering of the incident beam at the focus of the probe. At high concentrations, the excitation beam cannot penetrate the volume of dye. Attenuation of the incident beam causes loss of signal at the detector. In this study, the upper limit i s about 0.01"1. 3 . 2 T h e ? H a r d w a r e This section describes the instrumentation and the modifi- cations that were made to the instruments. Figure 8 shows the block diagram of the VF system assembled for this study. (Appendix B.2) The principal component of the fluorophotometer i s the s l i t lamp microscope. (Figure 9.) The b u i l t - i n power supply (from mains) with specific intensity settings was replaced by a regulated d.c. supply because random variations in beam intensity were found to occur. These fluctuations were believed to arise from variations of the line voltage when the number of users increased (i.e., unregu- lated line). In order to continuously monitor the intensity, a photovoltaic c e l l was placed along the path of the beam before i t was focussed through the s l i t and prism system. This c e l l , placed close to the bulb, did not block the beam's path. This method of monitoring lamp intensity, which i s dependent on optical alignment, was compared with another method which monitors the intensity of the output of the s l i t 30 Regulated Lamp Power Supply r Tungsten Bulb DVM Excitor F i l t e r / I n t e n s i t y Monitor Slide RETINA S L I T L A M P A S S E M B L Y Eye-piece 4 Fibre Optic i, to y i i_ P O D : = CONTACT LENSES Barrier F i l t e r Shutter IF M D A S Ribbon Cable to Computer L O G A M P R / M O S B O R N E M I C R O C O M P U T E R Dot-Matrix F>R I N T E R F i g u r e 8. BLOCK Diagram of the VITREOUS FLUOROPHOTOMETER 31 2 6 2 5 2 4 1 . F i x a t i o n l a m p 2 . H r u b y l e n s 3 . H r u b y l e n s g u i d e r a i l 4 . F o r e h e a d r e s t 5 . C h i n r e s t 6 . K n o b f o r c h i n r e s t h e i g h t a d j u s t m e n t 7 . A r m c l a m p i n g s c r e w s 8 . G r i p b a r 9 . C o r d f o r f i x a t i o n l a m p 1 0 . S o c k e t f o r f i x a t i o n l a m p c o r d 1 1 . G e a r b o x c o v e r 1 2 . Z o o m leve r 1 3 . S l i t t i l t i n g r i n g 1 4 . •• F i l t e r s l i d e 1 5 . S l i t r o t a t i o n k n o b 1 6 . A p e r t u r e d i a p h r a g m 1 7 . C o a x i a l k n o b f o r s l i t c o n t r o l a n d r o t a t i o n 1 8 . S o c k e t f o r s l i t l a m p h o u s i n g 1 9 . K n o b f o r h e i g h t a d j u s t m e n t o f m i c r o s c o p e s l i t l a m p a s s e m b l y 2 0 . L e v e r f o r c r o s s - s l i d e m o t i o n , c o a r s e a n d f i n e , o f t h e t a b l e 2 1 . P o w e r c o r d 2 2 . M a i n s w i t c h a n d c o n t r o l f o r s e c o n d a r y v o l t a g e o u t p u t 2 3 . P i l o t l a m p 2 4 . C r o s s - s l i d e t a b l e 2 5 . B a s e p l a t e 2 6 . S w i v e l a r m c o n n e c t i n g m i c r o s c o p e a n d s l i t l a m p a s s e m b l y t o c r o s s - s l i d e t a b l e Figure 9 . The S l i t Lamp. (From NIKON Zoom-Photo S l i t Lamp Microscope Bench Type Instructions Manual) 3 2 illumination using a photocell. Testing showed that the latter method was more sensitive to variations in beam intensity. At high intensities (i.e., running high currents through the filament), fluctuations were noted when used on the a.c. mains. These fluctuations were reduced (halved) when the regulated supply was installed. It should be noted that these intensity fluctuations appear as variations in the signal about an average because the system i s continuously exciting and detecting the fluorescence in overlapping volumes (because of the diamond). (Refer to Section 2.5.) Therefore, provided that the intensity of the bulb does not vary significantly from the same average value during each scan, the fluorescence fluc- tuations can be interpreted as deviations about an average con- centration at any position in the scan. It was also found that maintaining a constant, high intensity illumination was d i f f i c u l t because the high currents (and tempera- tures) cause the bulb intensity to f a l l continuously. Stable high intensities could only be attained after a "warming-up" period usually about 30 minutes. However, excessive, long periods at high intensities caused a reduction in the lifetime of the bulb. The configuration to monitor the intensity employs the unused side of the excitor f i l t e r holder-slide. A small solar c e l l was glued to a microscope cover-glass. The assembly was, in turn, glued to a washer that f i t t e d into the slide. This method does not monitor the beam during a scan because the chip cuts off the beam when i t i s in operation. 33 Intensity checks are carried out immediately prior to scanning. The output of the solar c e l l i s measured on an LED voltmeter. The intensity output is always adjusted to the value (on the voltmeter) at which calibrations were carried out. S l i t lamp translations are measured by constructing a potentio- meter with a 10-turn rotary potentiometer/resistor. Fitted with a gear on i t s shaft, the pot is held by an arm attached to the grip bar of the s l i t lamp. (See Figures 8 and 9. ) The assembly i s referred to as the Pod. The gear rests on a rack which i s mechanically coupled to the body of the s l i t lamp. When the rack moves with the s l i t lamp during a scan, the Pod produces the analogue voltage signals which are read by the microcomputer. (See Figure 8. ) The most important modification to the s l i t lamp concerns the oculars or eye-pieces. One ocular i s replaced by a special adaptor with a fibre optic conduit at the focal plane of the microscope objective. The fibre optic collects the fluorescence from the diamond and conducts the light to the Photo-Multiplier Tube (PMT). CHIP FILTER Figure 10. Excitor f i l t e r holder-slide and intensity monitor. 34 Figure 11. Pod assembly and c i r c u i t . An e lec tronic shutter and the bar r i e r (green) f i l t e r are placed between the output of the f i b r e opt ic conduit and the PMT. (See Figures 8.) The PMT then relays to the radiometer (R/M) whose output voltage varies l i n e a r l y with the amount of input fluorescent l i g h t . The R/M has several exponent set t ings , inc luding an AUTO- adjusting exponent option. The AUTO se t t ing , which keeps outputs between 0 and +100 mV, was found to be unstable at the "cross-over" points where output voltages greater than +100 mV were scaled down. Hence, the R/M i s set i n the (most sens i t ive) 0-exponent range. To offset high R/M outputs when high concentrations are scanned, a logarithmic ampl i f ier (ca l led LOG AMP i n Figure 8) i s used. The Log Amp was ca l ibrated i n conjunction with the Data Acqui- s i t i o n System (DAS). This i s to ensure that the ampl i f icat ion of small input s ignals i s such that i t s outputs remain logarithmic. 35 DATA LINES DI09-7 FROM/TO CPU IR5 TTTTTTTT J_L 13 11 18 3 5 Q ADC 3 4 20 7 a & 1 2 10 - f sv +2.5V REF C3 H H C5 C5 cs H H C5 Pl +5V IEF P2 -15V +15V 15V POWER SUPPLY Figure 12. Pin diagram of the blue c i r c u i t board i n the DAS. C3* i s 3 C2-capacitors i n s e r i e s . Unused S/H pins are not shown. Offset trimpots, Rl, f o r A are not shown. (See Appendix B.1 f o r part numbers.) 36 The outputs of the Pod and the Log Amp are connected to the DAS. Two circuit-boards comprise in the DAS. A 15-V power supply (white) board produces the necessary power for the components on the principal (blue) board. The latter c i r c u i t i s shown in Figure 12. (Part numbers are given in Appendix B.l.) The Pod and Log Amp outputs are connected to the inputs of two operational amplifiers (A) which are each followed by a sample-and- hold chip (S/H). The outputs from the latter are then connected to two channels of an 8-channel analogue multiplexer (MUX) which i s an electronic switching device. The other six channels, that are not used are grounded. The MUX selects one output of the S/Hs after the other, passing the signal onto the next chip. The next chip i s an 8-bit Analogue-to-Digital Converter (ADC) which i s the principal component of the DAS. Its purpose i s to con- vert the analogue inputs, held steady by the S/Hs, to the d i g i t a l format that the microcomputer understands. As i t can only convert one input at the time, the S/Hs become necessary for maintaining those input voltages until the MUX selects them for the ADC. The output of the ADC goes out on a ribbon cable to the micro- computer. The cable also carries the sequence of instructions from the computer to the various DAS components in order to properly organize the conversion of the signals. Besides powering the As, the MUX and the S/Hs, the +/- 15-V supply also operates the shutter in the PMT. The latter opens when the voltage i s changed from -15 to +15V, and closes when the polarity i s reversed. The shutter movements are co-ordinated with the 37 activation of the DAS by a mechanical toggle svitch called Switch A on the front face of the DAS box. Lastly, the +5V logic level required in CMOS di g i t a l elec- tronics [433 i s obtained by connecting the +15-V line to a voltage regulator chip, <P2 in Figure 12). The +5V output also provides the voltage drop across the Pod circ u i t . It i s also, in turn, connected to another voltage regulator-reference chip, PI to produce the 2.5V required for the ADC reference. The microcomputer i s the 8-bit, 64-K Osborne 1 with CPM opera- ting system. Its main advantages are i t s parallel and seria l ports. The disadvantage of this machine i s that i t i s configured with three memory banks. Bank #1 contains the usual transient programming area, while Bank #2 partially shadows i t . A l l ports are accessible from Bank #2 only [44,453. Bank #3 i s video memory. As the Osborne controls the DAS directly through the parallel port, considerations such as the transition time between bank had to be taken into account during programming in order to optimize the data acquisition. Although the ingenuity of the electronic design of this configuration did not go unappreciated, the time required to understand and work with i t could have been put to better use. The last component in the system i s a dot-matrix printer that produces the hard-copy results. It i s also accessed through the parallel port (by Centronics communication protocol). As no printout i s ever required until a l l scanning has been completed, there i s no competition for the port between the printer and the DAS. 38 3 . 3 O t h e r M a - f c e? r i a J _ Two contact lenses are used during scanning. One i s a hydro- ph i l i c soft contact lens which i s f i t t e d f i r s t . Its purpose i s to alleviate the discomfort of the second lens without using repeated i n s t i l l a t i o n of topical anaesthetic. The second lens (Luma" lens) i s plano-concave to offset the power of the cornea. It i s made of a mildly pliable plastic, and provides the "window" for viewing the fundus. The soft lens i s attached to the cornea by the surface tension of tears and the hard lens attached to the soft lens by the surface tension of viscous methyl-cellulose. Saline solution was tried but did not hold the Luma" lens in position. It also caused i r r i t a t i o n in some subjects. The two band-pass f i l t e r s used are Spectrotech SE4 and SB5. Their transmission profiles and overlap are shown in Figures 13 and 14. Their cut-off wavelengths are in accordance with specifications for fluorescein in water. (Refer to Section 3.1.) The SE4 excitation f i l t e r passes wavelengths between 453 and 493 nm only. Similarly, only the main emission peaks between 509 and 612 nm are transmitted by the SB5 barrier f i l t e r . The smaller (blue) SE4 i s mounted in the holder-slide which also holds the intensity monitoring chip. It i s inserted into the path of the beam before scanning. The larger green f i l t e r (SB5) i s placed between the output of the fibre optic and the electronic shutter above the PMT. Its large aperture ensures that a l l signals from the conduit pass through i t before activating the PMT. 39 40 ,T'° rrrri Figure 14. F i l t e r Overlap. E2 and SB were used. 41 A model eye was also constructed. Its purpose was to calibrate s l i t lamp translations. However, the calibrations were made with a more precise micrometer translation stage fixed on an optical bench. (Refer to Section 4.1.) The model eye i s also used as a sample c e l l (for plasma scanning). A cross-sectional profile i s shown in Appendix B.3. 3 . 4 T h e S o f t w a r e Most programmes were written in Microsoft BASIC (MBASIC). Although a compiled version called CBASIC was available, MBASIC was chosen because i t had many bui l t - i n functions for f i l e and string manipulation. The one exception was the programme for the control of the DAS. Written in 8080 Assembly Language codes [46], i t s "l i s t i n g s " f i l e i s called DAS.PRN. Certain programming "habits" were developed because of restrictions (and economy) in the use of memory space. For example, many of the MBASIC statements written were concatenated, as i s allowed by the language. "Free" variables were re-used wherever possible. (Refer to Appendix A for a l l programme listings.) Specific subroutines are called from menu programmes. The f i r s t such menu, SCANMENU.BAS directs control to one of three subprogrammes for scanning and f i l i n g . These are VITSCAN.BAS and PLASCAN.BAS which are stored on the disk in Drive B, and a subroutine (in SCANMENU.BAS i t s e l f , ) called SUBJECT DATA ENTRY. A subroutine, when called, i s merged above SCANMENU.BAS in the memory bank. When a scan i s ended, control i s returned to the menu. Prompts to operate the data acquisition were written into this 42 set of subroutines to enhance "user-friendliness". NO and YES res- ponses are indicated by hitting the ESC and ANY (other) key respec- tively. This association is appropriate as the ESC key i s located at the upper l e f t corner of the keyboard and has l i t t l e probability of being mistakenly activated, especially when lights are dimmed during scanning. A version called DAS.ASM was f i r s t prepared using the specifi- cation sheets of the various electronic components in the DAS as guides [433. It vas then assembled by the 8080 two-pass assembler provided vith the Osborne 1, producing the listings, DAS.PRN as shovn in Appendix A.l. Note that the l e f t four hexadecimal (hex) numbers denote the memory addresses (in Banks #1 and #2) of the machine language codes given by the next 2 to 6 hex digits. Entry and exit loops to Bank #2 are clearly marked. The port status test i s executed only on the f i r s t entry where, i f necessary, the port-controlling. Peripheral Interface Adaptor, MC6821, i s re-configured to suit the DAS. The strategy of this subroutine i s simple. After preparing the port, poll Switch A until the toggle i s up. Tell the S/Hs to sample their inputs simultaneously, then hold them. Ascertain that the ADC i s not busy. Next, order the MUX to switch on and the ADC to begin digitizing the pod-S/H output. When the conversion i s done, store the result at a specific address (D1D2 hex) in memory. Make sure the ADC i s ready. Now order the MUX and ADC to do the same for the R/M-Log Amp-S/H output. This time, store the answer at D1D4 hex. Test the interrupt status line (Switch A) and put the result at D1D0 hex. Go 43 back to the MBASIC calling subroutine. When SCANMENU.BAS i s loaded, i t reserves the area above D1CF hex for the machine language code numbers (from DAS.PRN), and the abovementioned results after each c a l l . The machine language codes are loaded only i f one of the f i r s t two subprogrammes i s called. They are placed into the reserved memory by the DATA-READ-POKE sequence of commands, starting at memory address D1D6 hex. If VITSCAN.BAS i s loaded, a checklist of the VF system i s immediately displayed and the f i r s t c a l l to DAS.PRN i s immediately made; usually to measure the various intra-ocular distances. VITSCAN . BAS marks a position by sounding a "beep" when any key is depressed. When "landmarking" i s ended, i t displays the difference between positions corresponding to consecutive beeps, then asks whether to repeat landmarking or continue on to scanning. Before scanning begins, a prompt to ascertain the eye to be scanned i s given: ESC for the right eye; ANY (other) key for the l e f t . Once answered, a set of "SCANNING INSTRUCTIONS" i s displayed. Switch A i s toggled up to i n i t i a l i z e the DAS in a 3-second loop. (This delay time may be-varied by software to suit the time constant of the DAS.) The digitized outputs of the Pod and the Log Amp are displayed in (approximately) the f i r s t second of this delay loop, after which, the screen i s blanked. The remaining time of the loop i s for "dark adaptation" by the subject and the system to reduce the noise picked up by the PMT. A "beep" sounds to mark the position of the retina (where every 44 scan must begin), as well as to cue the user to begin scanning. Control toggles back and forth between Banks #1 and #2. After each return from DAS.PRN, the abovementioned specified addresses are PEEKed. The data are transferred to elements of two 2x1600 arrays (depending on the eye being scanned). Switch A i s tested to ascertain that the scan is to continue or to stop. The maximum time available for scanning before the arrays are f i l l e d i s approximately 25s. If they are used up, VITSCAN.BAS auto- matically exits from the scanning loop to flash "You are out of memory...". Alternatively, Switch A may be toggled down to end a scan. Upon leaving the scanning loop, VITSCAN.BAS goes into a plot- ting and f i l i n g subroutine. Left-eye data are f i l e d on the disk in the l e f t drive; right-eye on the right. In this way, a one-diskette- per-subject-eye system of data storage i s maintained. This f l e x i b l i t y allows for both eyes of one subject to be tested, or, two subjects to be examined within the same period by assigning one diskette to one subject-eye. (As many subjects as time restrictions allow may be examined, but only the diskette in the LOGGED drive may be changed - a quirk of MBASIC.) An interrupt i s included to enable an "ABORT" during plotting and f i l i n g . It i s activated by depressing any key. Two prompts to confirm the "ABORT" appear. An positive response produces another prompt to continue scanning. A negative reply returns control to the position within the plotting and f i l i n g subroutine where the inter- rupt was activated. 45 The temporary f i l e that stores the data i s renamed when the p l o t t i n g and f i l i n g subroutine has been completed and i t i s confirmed that the data i s to be saved. The filename i s entered when the response i s a f f i rmat ive ; otherwise, the temporary data f i l e i s over- wri t ten or erased l a t e r . F i l e s are named by the p . i . time. For example, the 3-minute p . i . bolus scan i s named 3.DAT where the f i l e - type, ".DAT", i s automatically juxtaposed to the "3" entered. The prompt to continue scanning appears next. I f scanning i s to be continued, there i s a choice to begin again at SYSTEMS CHECKS or at SCANNING INSTRUCTIONS. If the scanning mode i s to be terminated, contro l i s returned to SCANMENU.BAS. In the plasma scanning option, s i m i l a r c a l l s to DAS.PRN are made. No p lo t t ing loop i s required i n t h i s case as the s l i t lamp i s f ixed i n pos i t ion . (See Section 4.3.) The arrays used are smaller: two arrays of 1000 elements. An averaging subroutine i s immediately entered when the scanning loop i s exi ted. A s i m i l a r interrupt capabi- l i t y i s also i n s t a l l e d . The resu l t s of the averaging and the standard deviations are displayed. The time of the blood sampling i n minutes p . i . i s then entered. Three 55-element arrays are used to hold the resul t s of each plasma sample scanned. At the end of scanning, the arrays are f i l e d i n PLASMA.DAT i n the appropriate subject-eye disket te . The SUBJECT DATA ENTRY subroutine i s usual ly ca l led after • VITSCAN.BAS and PLASCAN.BAS have been executed. This subroutine i s used to enter pertinent subject information: name, age, eye that was scanned, date, lengths of intra-ocular distances recorded by both the 46 s l i t lamp and ultra-sound scans, volume of dye that was injected, comments, and observations. The last category i s used to enter notes on a particular scan or the subject's medical history. A l l informa- tion entered i s f i l e d in SUBJECT.DAT in the appropriate diskette. At the end of an examination period, there should be a SUBJECT .DAT f i l e , a PLASMA.DAT f i l e , and the scanning .DAT f i l e s on a subject-eye diskette. A l l other non-.DAT f i l e s are erased. For later analysis, a working copy of the data diskette i s always made. The second menu routine i s BATCHRUN.BAS. It allows the execution of individual subprogrammes, or a specified sequence of subprogrammes. On activation, a menu shows a l l the subprogrammes with brief descriptions of their purposes. The order of the subprogrammes to be run i s entered and confirmed. The f i l e directory of the data diskette next appears, and the f i l e s to be analyzed are then entered. After printer status and paper supply have been checked, execution i s begun by CHAIN MERGEing the f i r s t subprogramme in the sequence. Control i s then transferred to the task. On completion, the next subprogramme in the sequence i s loaded over the f i r s t , and so on until the series i s completed. The following describes the subpro- grammes and their tasks. REDUCE.BAS averages the raw data in each . DAT f i l e , reducing the number of data-points from a possible maximum of 2x1600 per f i l e to a possible averaged maximum of 3x256 - Pod output, averaged R/M output and i t s standard deviation. The algorithm, although slow because of memory space restrictions, i s not d i f f i c u l t to follow. A flowchart i s shown in Figure 15. 47 jRead in .DAT f i l e : X0; <X,Y) pairs; find X-4" and X"*". © Bin a l l Y's of this X. X0=Pod starting position; XX=final X=X-X0; YY=Averaged Y; Z2=YY's S.D. at XX; U=YY at previous X, i.e. X-l; T=chosen Y to average about; If X=X-1", set U=T; L=averaging interval about T. Store XX,YY, ZZ in .AVG f i l e s . •j STOP ) Yes t £ ) Find the Scatter of these Y's. No Choose the Y that occurs most times and i s closest to U. Set this Y=T. © 4 Yes L=all Y's. No T > U Yes Set L=[T-5, T*5], No Yes Set L=CU,T+5], Set L=CT-5, U3. Average over a l l the Y's in L to find YY and ZZ, and XX=X-X0. Set U=YY. Figure 15. FLOWCHART for REDUCE.BAS. 48 Basically, REDUCE.BAS considers the scatter of the R/M outputs at each Pod output. It sets an interval within which a l l R/M outputs are included in the averaging. This interval i s dependent on the result of the averaging of the Pod output before i t . Hence, the averaging process i s carried out point by point in ascending order of Pod output from the posterior vitreous to the cornea. The results are zeroed by the retinal position (at the "beep") and converted to concentration units before being stored in a .AVG f i l e with the same p.i. time filename. After a resultant .AVG f i l e has been entered, the associated •DAT f i l e i s destroyed. This action frees diskette space for later f i l e s . This i s the reason for creating a working copy of the original data diskette. This technique of data reduction or "smoothing" was chosen over the usual methods such as curve-fitting because of the method of scanning. Also, as mentioned before, this VF system i s a time- averaging DAS as opposed to the "spot" system [383. It i s "backward- biased", as i t i s dependent on the average of the previous Pod posi- tion because of the capacitive time constant (approximately 2.2s) in the electronics. Several background scans are usually made. They are .DAT f i l e s identified as "0", "00", "000", etc. B/G.BAS contains a background- file - t e s t i n g loop to ascertain that multiple pre-injection scans were made. If the loop finds several such scans, B/G.BAS takes these already averaged background scans, aligns (by the location of retinal position) and averages over a l l of them to produce one f i n a l 0.AVG 49 f i l e . A l l other background .DAT and .AVG f i l e s are then deleted to free diskette space. If only one background scan was made, B/G.BAS returns to BATCHRUN.BAS to the next subprogramme in the series. MINUS.BAS subtracts the f i n a l 0. AVG f i l e from a l l other .AVG f i l e s . Two sets of results are produced by two types of alignment: a) by matching the positions of the retina as they were marked on each scan (at the beep); or, b) by searching out and matching the CR peaks. Subtraction i s then carried out after each alignment. The results by (a) and (b) are f i l e d in .RET and . CRP f i l e s respectively. Again the filenames are the p.i. times of scan. SUDATA.BAS prints out the subject and scanning data. A formatted output of the information in the SUBJECT.DAT f i l e i s produced. The F-numbers (in Section 2.2) are calculated and printed. The routine reads through every .AVG, .RET and .CRP f i l e , compiling the concentrations at a set of specified points in each profile. These points are the retina, the CR, lens and corneal peaks, the minima and the centres of the vitreous and the aqueous chambers, and points (that are in the vitreous chamber only) which are 0, 3, 6, 9, 12, 15... mm from the retina. At the end of compilation, SUDATA.BAS calculates, when appropriate, the performance specifications: LLoD, AR and R. BLOOD.BAS i s the curve-fitting subprogramme that calculates the areas under the plasma profile. PLASMA. DAT i s read, and the data are converted to concentration units with a constant that accounts for buffer dilution and the unbound fluorescein factor - 17'/.. (See 50 Section 3.1 and C47],) To calculate this constant, i t i s assumed that after centri- fuging, a l l the red blood cells have been removed to one end of the haematocrit but that the fluorescein i s homogeneously distributed throughout the plasma. If c* (t) i s the plasma fluorescein concen- tration at sampling time, t, and only 17% i s available to penetrate the BRB at a l l times, the area under the plasma-fluorescein profile i s given by (23) I(t) = 0.17 C " ( T ) dT T • * If X ml of each spun sample i s diluted with Y ml of buffer, then the resultant concentration, c»(t), measured by the s l i t lamp i s given by (29) c» <t) * X • c» <t) / { X • Y ) . Eq. 28 becomes (30) I(t) = 0.17 { 1 • Y/X } f>T - « C P ( T ) dT T • • The factor before the integral depends on the amount of free fluores- cein (17%), and the volumes of plasma sample and the buffer used in the dilution process. Note that any amounts of X and Y may be used. A set of four polynomials of order 2 i s f i t t e d to the data, i.e., (2+1)-parameter curve-fitting [34]. The polynomials are: 0) y = A + B » x + C » x* ; 1) y = A • B * log x + C » ( log x )• ; 51 2) y = A * exp { B * x • C » x* } ; and, 3) y = A • B / x + C / x * They were chosen to approximate the removal of the dye from the blood. The d i f f i c u l t y in producing a fast and efficient algorithm in a non-linear f i t to the sum of two negative exponentials (expected of the fast and slow decays of the two-compartment model mentioned in Section 2.2 and [28]) forced such a method for estimating the plasma integral. It i s also d i f f i c u l t to justify such a non-linear f i t to a few data-points only. (Refer to Sections 4.2 and 4.3.) Areas under the best f i t are calulated as follows - (1) the integration lower limit is t=0. 5 minutes p. i . (2) The upper limits are the measurement scans' p.i. times (given by the filenames). (3) The area from 0 to 0.5 minutes is approximated by the area of a right-angle triangle with base 0.5 and height equal to the value of the best f i t at 0.5. (Refer to Section 2.2.) Plots with the f i t s superimposed on the raw data are produced. The goodness of f i t i s determined by the reduced chi-square of each f i t . At the end of the four f i t s , the one with the lowest reduced chi-square i s chosen. Its code number (0-3) i s entered in a PLASMA. FIT f i l e , followed by the coefficients (A, B and C) and their calcu- lated errors, the upper limits of each integration, the areas up to those limits, and their errors. C/VAZ.BAS follows closely the algorithm of Section 2.3. This subprogramme uses .RET and .CRP f i l e s only. When .RET f i l e s are used, a l l profiles are aligned for subtraction by the positions of the retina as marked at the beginning of each scan. Similarly,- .CRP 52 f i l e s are aligned by the CR peaks before any subtraction. The bolus .RET f i l e and a measurement .RET f i l e are read simul- taneously. The permeability indices, P1 (of Section 2.2) and the penetration ratios, PR3 (of Section 2.3), as well as PR3» are calcu- lated for each profile. The CR correction condition, the left-hand side of Eq. 6, is also calculated. Two types of bolus correction are undertaken. The f i r s t involves subtraction of the bolus from the measurement scan without applying the CR bolus ratio. (Refer to Section 2.5.) A plot of the results between 0 and 11 mm from the retina i s produced, followed by calculations of P1, PR3 and PR3*. For the second type pf correction, the subtraction i s repeated but with the bolus profile modified by the CR bolus ratio. A plot i s produced, and the three parameters calculated. The procedure i s carried out for the . CRP f i l e s with the same p.i. times. If more than 10 non-negative points exist within the posterior half of the vitreous chamber after each subtraction, C/VAZ.BAS f i l e s them in .CV# f i l e s on the data diskette. The f i l e s are named with the same p.i. time filenames, but, with the following filetypes: .CV1 'unaltered bolus subtraction of .RET f i l e s " ; . CV2 "modified bolus subtraction of .RET f i l e s " ; . CV3 "unaltered bolus subtraction of .CRP f i l e s " ; . CV4 "modified bolus subtraction of .CRP f i l e s " . If many measurement scans are made, the data diskette can run out of storage space. To prevent this from halting the calculations, C/VAZ.BAS contains an error-trapping subroutine which automatically 53 f i l e s the .CV# f i l e s on the diskette which contains the MBASIC pro- grammes in the other drive. By studying the plots produced by C/VAZ . BAS, some of the . CV# f i l e s can be deleted. The .CV# f i l e s on the programme diskette can then be transferred back onto the data diskette. SLOPES.BAS checks the results of C/VAZ.BAS by curve-fitting at the . AVG-file level, using the same set of (2+1)-parameter functions as those in BLOOD.BAS. Only vitreous data are considered in the curve-fitting. A vitreous profile i s divided into three sections: posterior, mid-, and anterior vitreous. The four functions are fitt e d to one section at a time. The f i t vith the lowest reduced chi-square i s chosen to represent that section in a l l later calculations. When a l l the best f i t s have been calculated for each section, SLOPES.BAS uses them to estimate the following: a) the concentrations after background subtraction, c(x,t). b) the penetration ratios, PR3», before and after bolus subtraction. c) the gradients, c 9(x,t). d) the diffusion constants, D. The values chosen for x are 3 and 9 mm from the retina, as well as 3mm from the posterior surface of the lens. PR3* i s found at the 3-mm point from the retina and the lens without CR bolus correction. Alignment i s by RET only. D i s approximated by converting the diffusion equation, Eq. 1, . into a difference equation [34,48]. This approximation uses the calculations at the 3-mm and 9-mm points. It i s 54 (31) c<9,t)/t = D • f c 8(9,t) - c'(3,t) )/< 9 - 3 ) . t i s in minutes p.i. Solving for D, (32) D = c<9,t) » t-' * 10-1 / { c 8(9,t) - c a(3,t) } em's". The 10-1 factor appears when converting to centimetres and seconds. These values of PR3* and D may be used as i n i t i a l estimates for LUND.BAS curve-fitting as well as checking the solutions of RET C/VAZ.BAS. PLOT.BAS i s an earlier version of SUDATA.BAS. It produces individual scaled-down plots of any .AVG, .RET or .CRP f i l e . A l l the information gathered by SUDATA.BAS i s also collected and printed below each plot. Comparison of data and plots by inspection i s d i f f i - cult. DRAW.BAS is the plotting subprogramme that complements SUDATA .BAS. This subprogramme uses the f i r s t entered f i l e to produce semi- log plots recorded data-point by data-point. A l l subsequent plots are scaled and superimposed on the f i r s t plot. Its principal advantage i s that i t can superimpose the profiles from different subjects as well as producing "fully stretched" profiles. Limited by the loss of clar i t y when too many f i l e s are superimposed, the routine provides a means for easy comparisons of the profiles from different subjects, and the evolution of fluorescein profile of one subject. The last routine written was LUND.BAS. The algorithm i s that . stated in Section 2.4. The routine - largely translated from the FORTRAN subroutines in Bevington [34] - uses the Marquardt gradient- expansion method of non-linear curve-fitting. Although i t i s able to 55 read data from any file(-type), LUND.BAS i s usually run on data from .CV# f i l e s . The following are required at each run: a) chosen f i l e from which data are read. b) interval within which data are to be fitted. c) i n i t i a l estimate of P, and i t s step increment. d) i n i t i a l estimate of D, and i t s step increment. The chosen interval i s usually between 1.5 and 6 mm from the retina. I n i t i a l P and D values depend on the expectations of the results for the particular subject i.e., the results of C/VAZ.BAS and SLOPES. BAS. Increments are usually 107. of the i n i t i a l estimates. The subprogramme tests the f i t by calculating a reduced chi- square rather than the S* of Eq. 23 in Section 2.4. The weights are the S.D. of the data as calculated by REDUCE. BAS (or others), i.e., at each Pod position. It i s also possible to use the LLoD (i f known) in Eq. 24. The reduced chi-square for the i n i t i a l values entered i s found. The subprogramme then goes on to find better f i t s by varying P and D by the increments entered. After each f i t , a new reduced chi-square i s found. If i t i s smaller, the subprogramme can be terminated. If not, i t i s used to determine the direction to vary D and P for the next f i t . The search can only be terminated in three ways - (1) hardware interrupt (abort) or, (2) overflow errors (division-by- zero), (3) reduced chi-square less than 2. As poor i n i t i a l estimates may result in slow convergence and long computing time, this subprogramme i s only used in overnight runs. 56 X V . C A L I B F £ A T O M S A N D P R O T O C O L In this chapter, the calibrations of various components of the VF system and the preparation of subjects for scanning are described. The results of the former are given in Appendix C. As the ADC is an 8-bit converter for a maximum input of 5V, one digit increment of i t s output 256 "Osborne/DAS" units - 0 to 255 - i s a *19-mV change of input voltage. A l l calibrations must be in ranges that are expected during a scan. The gains at every stage in the DAS must be carefully adjusted so that the f u l l range of the ADC can be used. In this way, the sensitivity of the VF system i s optimized. • 4 . 1 InstruTTient Preparation The operational amplifier (A) receiving the Pod output, and the S/H were offset and zeroed. The gain was adjusted to approximately 12. This allowed a net displacement of the s l i t lamp of approximately 22 mm. The focus of the probe in the vitreous chamber however, was displaced more than this value because of the F-numbers (in Section 2. 2). The Pod (or s l i t lamp translation) was calibrated by placing the s l i t lamp microscope assembly on an optical bench on which a micrometer translation stage was fixed. The micrometer was used to displace the s l i t lamp in precise, 0.025-in increments. The result of each displacement was recorded from the computer monitor. A linear least-squares f i t was carried out on the data collected. Only the slope of the f i t i s required as the zero position changes with each 57 scan. (See Appendix C.1.) As the Log Amp and the DAS are both capable of amplifying any input signal their gains have to be adjusted in tandem allowing low pre-injection scan signals to be differentiated from noise but not allowing high concentration signals to be amplified to greater than the maximum 5V that the ADC can convert. The f i r s t step was to ascertain the range over which the Log Amp remained logarithmic. At low voltage inputs, the Log Amp did not maintain i t s function, because i t was performing close to "zero* at 1 mV (the logarithm of 0 was of course not defined). For input values up to 10V, the Log Amp performed satisfactorily. Other conditions were: a) R/M set at 0-exponent outputs a maximum signal of about 10V for a maximum concentration of about 100000"1 ; b) Log Amp output in the range of 0 to 5V for the ADC using 1 mV to 10 V as the input range from the R/M to the Log Amp. The Log Amp was adjusted so that i t s output was logarithmic in the range of input voltages. (See Appendix C.2. ) The operational amplifier to which the output of the Log Amp was connected, was then adjusted to a gain of 1. Samples of various concentrations of sodium fluorescein were freshly made at each calibration because solutions lose their fluore- scence efficiency over long storage periods. As the pH of the solvent affects i t s efficiency, the appropriate buffer solution (pH 7.4, discussed in Section 3.1) was also prepared. A pH meter was calibrated at 20C and pH 7.0 using a standard pH 58 7.0 solution. The Sorensen's Phosphate buffer solution at pH 7.4 was then made up as follows - a) Prepare a 0.9077. weight by volume (or, 0.067 moles) solution of monobasic potassium phosphate in demineralized d i s t i l l e d water. Solution "A". b) Prepare a 2.397. weight by volume (or, 0.067 moles) solution of sodium phosphate in demineralized d i s t i l l e d water. - Solution "B". c) To get 100 ml of a pH 7.4 buffer solution, for example, mix 19.7ml of A and 80.3ml of B, i.e., mix in A:B volume ratio of 1:4.076. As an alternative, once the pH meter was calibrated, the pH 7.0 buffer concentrate, (the standard used to calibrate the pH meter) was used as a "base" to make the pH 7.4 buffer. This method was possible because the concentrate was made from sodium phosphate. Hence, by adding potassium phosphate to the solution at pH 7.0, a pH 7.4 buffer was obtained. This method was used to prepare the large volume of the pH 7.4 solution that was required for the sample concentrations needed for calibration, and also for diluting plasma samples from blood taken during scanning. Sample concentrations were made by a dilution method. a) Weigh an empty test-tube with seal in place. b) Add an amount of sodium fluorescein powder, then weigh again. c) F i l l the test-tube with the pH 7.4 buffer and weigh again. The f i r s t prepared sample was a "master solution" from which a l l . other samples were derived. Its concentration was estimated by assuming that the mass and volume of sodium fluorescein were negli- gible compared to the mass and volume of added buffer. The 59 concentration of the master solution was then (33) c" a = { (b) - (a) } / { (c) - (b) }"1 where (a), (b) and (c) are the results of the stages mentioned before. The mass of the dissolved salts was also assumed to be negligible. The following procedure was used to make other samples: a) Weigh an empty test-tube with i t s seal. b) Draw a small volume of c"a with a 1-ml tuberculin syringe, place i t in the test-tube and weigh i t . c) Add buffer then weigh again. The concentration of the new solution was (34) c = c"' * { (b) - (c) ) / { (c) - (a) >"1 . where (a), (b) and (c) are the results of the steps discussed imme- diately above. Other samples were made by varying the volumes of c" 8 and the buffer in order to prepare concentrations between 5 and 9000"1. Note that the error due to the mass of the fluorescein became less with greater dilution. The prepared concentrations were placed in several sample cells (cuvettes) which were clamped in front of the s l i t lamp. Each c e l l was scanned through i t s 1-cm depth. This was carried out to study the effects of attenuation. (See Appendix C.5.) The peak in each concen-. tration profile was used to find the R/M-Log Amp-DAS calibration curve. The reason for this was explained in Figure 2. The R/M output increased linearly with the concentration that 60 i s scanned but not the output of the Log Amp. Hence, the ca l ib ra t ion equation was expected to be an exponential that translates the Osborne/DAS units to concentration values. The c u r v e - f i t t i n g functions used were those in BLOOD.BAS. (Refer to Section 3.4.) The best f i t was the one with minimum reduced chi-square. The mathemati- c a l so lut ion i s given i n Appendix C.4. Other s l i t lamp ca l ibra t ions included s l i t dimensions, angle between exc i ta t ion beam and probe, and the in tens i ty of the beam. The s l i t s ize used was 2 X 0.1 mm. As the length did not affect the AR ( in Figure 2), th i s magnitude was chosen to provide better v i s i b i l i t y when focussing at the re t ina . The width of the s l i t was defined by the ru l ings on a red blood c e l l counter (or hemacytometer). The v e r t i c a l and hor izonta l l i n e s formed a gr id of 1 mm* with subdivisions of 1/400 mm2. The s l i t was focussed on the g r id . The microscope oculars were adjusted to the maximum (35X) magnification and the s l i t width adjusted to the desired s i ze . The slit-adjustment knobs were then secured to prevent accidental reset. S l i t- lamp in tens i ty , monitored by the s o l a r - c e l l chip, was set at a chip output voltage of 141 +/- 1 mV which was the lamp " i n t e n s i t y " used to carry out the concentration ca l ib ra t ions . A l l scans had to be made at th i s LED voltmeter reading. The output had to be checked before and after each scan to ascertain that the exc i t a - t i o n beam did not f luctuate s i g n i f i c a n t l y during scanning. When the angle between the d i rec t ions of the probe and the exc i t a t ion beam was maximum, the AR was at a minimum. Using the £1 average depths and refractive indices of the model eye - Gullstrands schematic eye [30], this angle was found to be approximately 16°. (Other investigators' instruments are usually set at 14". [31]) The entire s l i t lamp (objectives and bulb housing) was rotated, then lacked at 8° from the translation axis of the s l i t lamp assembly. This latter adjustment could be changed for particular subject to overcome i r i s clipping or specular reflection off the plane surface of the Luma* Lens onto the probe. To test the suitability of a l l instrument adjustments, a scan of the model eye was carried out. Different concentrations were used in the various compartments. The profile i s shown in Appendix C.7. Bulb and Focussing FOCUS Probe Pick-up Path Right Ocular with Fibre Optic Probe Figure 16. Overhead view of the angles at which the s l i t lamp was set for LEFT eye scans. 62 4. 2 Sub j ect, P r e p a r a t i o n A subject i s given a CONSENT FORM (shown in Appendix A.17) to read and sign. The attendant ophthalmologist answers additional ques- tions the subject may have and countersigns the form. The age and the weight of the subject are noted and the latter used to calculate the amount of sodium fluorescein for injection (at 14 body weight). The pupils are dilated using drops of tropicamide 1% and pheny- lephrine 57., repeated after 10 minutes. After a 30-minute waiting period for the drugs to take effect, the pupil diameter i s measured. If i t i s 7 mm or larger, the scanning procedure i s started. If the dilation i s s t i l l inadequate for scanning after additional tropica- mide and phenylephrine, the procedure i s cancelled. The subject's seated position i s explained and tried. The s l i t lamp and chin rest are adjusted to the subject's height to minimize discomfort during the scanning. The s t e r i l i z e d soft-contact lens i s then f i t t e d to the l e f t eye, then the s t e r i l i z e d Luma" Lens i s mounted with a drop of methyl-cellulose. Each mounting i s checked for trapped air bubbles. The subject then places his/her head in position on the s l i t lamp assembly with instruction to press the forehead against the headband. The f i r s t scan for "landmarking" uses unfiltered light to localize the surface positions of the retina, the posterior and anterior lens, and posterior cornea (along the optic axis). The depth or thickness of each section i s displayed and noted (in Osborne /DAS units) by the programme, VITSCAN.BAS. The measurements are 63 usually done three times to check reproducibility. The retina i s marked by two methods. One way i s to centre the smaller, 0.1mm s l i t on the larger, 0.45mm diameter probe. However, the effects of scattering (halation) makes this method d i f f i c u l t . Another method consists of placing the l e f t edge of the s l i t on the right edge of the probe; then the right edge of the s l i t on the l e f t edge of the probe. The two results are later averaged to "define" the retinal position. Both methods are used within the three measure- ments. (a) Retinal landmarking (b) Other interfaces Figure 17. Localizing techniques. Other surfaces are marked by the contrast at the interfaces. For example, at the posterior surface of the lens, a sharp demar- cation i s seen as there i s autofluorescence and scattering from within the lens tissue, and reflection at the interface between the lens and the "dark" vitreous volume. (Refer to Section 2.1.) The smaller depth of the anterior chamber makes the measuring of that 64 compartment more d i f f i c u l t . The pre-injection scan(s) are done immediately after the length measurements are completed. The probe i s f i r s t focussed on the retina, and s l i t intensity is checked. The blue excitation f i l t e r i s s l i d into the path of the beam; Switch A is toggled up. Scanning is started at the "beep", and at the end of each scan, the visual plot i s studied for signs of clipping or subject movement. Up to four background scans can be made. (Refer to Section 3.4.) When background scanning i s completed, the Luma* lens is removed and the subject brought to the photography room where the fluorescein i s injected and photographs of both maculae and optic discs are taken. The injection i s given by a registered nurse who also advises the subject of the dye's possible immediate effects (nausea), and delayed effects such as yellow-coloured urine. The dye i s injected quickly. Two stopwatches are started when the syringe i s half-emptied. The photography is usually completed 55 to 70 s after the injection. The bolus scan i s the c r i t i c a l scan as there i s a maximum time interval within which i t must be made. (Refer to the theory in Sec- tion 2.3.) The protocol requires a 3-minute p.i. bolus scan. However, events such as instrument intensity re-adjustments may delay this scan. A bolus i s never accepted past 7 minutes p.i. Other "bolus" scans are taken to study the early changes in the levels of plasma fluorescein, and, therefore, CR bolus tailings. Plotting and f i l i n g usually take up to 90 s. Immediately after bolus scanning and the removal of the Luma* 65 lens, a blood sample i s collected. A finger i s pricked with a s t e r i l e lancet and approximately 0.1 ml of blood i s collected in a hepa- rinized capillary tube (haematocrit). The p.i. time i s noted when half the tube i s f i l l e d . Another sampling i s done at about 15 minutes p. i . The f i r s t measurement scan i s taken around 30 minutes p.i. Other scans can also be taken for reproducibility studies. The third blood sample i s taken after the Luma Lens has been removed. This procedure i s repeated around 60 minutes p.i. The soft lens i s also removed as scanning with the s l i t lamp i s ended after the 1-hour measurement scan(s). The last parts of the examination are monochromatic nerve-fibre photography and ultrasound measurements of intra-ocular distances. The latter i s used to compare against the distances of the various segments of the eye measured by the s l i t lamp. The ultrasound results are also used to scale the profiles collected, and to calculate the F-numbers for the optics in this VF system. START Read and sign CONSENT Form. Answer questions. -30 min.p.i. Dilate pupils with tropicamide. -5 min.p.i. Mount lenses. Measure intra-ocular distances. Background scanning. 0 min.p.i. Injection and photography. 3-7 min.p.i. Bolus scanning and blood sampling. 15 min.p.i. Blood sampling. 28-33 min.p.i. Measurement scanning and blood sampling. 55-65 min.p.i. Measurement scanning and blood sampling. Nerve-fibre photography; Ultrasound scanning. Plasma-fluorescein scanning. Figure 18. Flowchart of the scanning procedure. 66 4 . 3 B l o o d - P l a s m a P r e p a r a t i o n The four blood samples collected are centrifuged immediately after the ultra-sound measurements have been made. The speed i s set at 2000 rpm for approximately 12 minutes. This adequately separates the red blood cel l s . 2 ml of pH 7.4 buffer solution are measured out by a volumetric pipette and placed in each of five (cuvette) dry sample cel l s . One c e l l i s used as the background sample. 0.01 ml are dravn from each haematocrit using a 0.025-ml pipette and emptied into each c e l l . The constant outside the integral in Eq. 30 i s then 34.17. The c e l l s are then scanned by the s l i t lamp. Each c e l l i s placed so that no light i s reflected off any of i t s surfaces onto the probe. The probe i s positioned visually, in such a way that i t i s not focussed on air bubbles on the inner surfaces of the cells. The focus of the probe i s slightly behind the inner front wall of each c e l l . This i s due to possible attenuation of the incident beam owing to the high concentrations. Also, from Figure 2b, the probe must be totally included in the sample volume. The f i r s t sample to be scanned i s always the background c e l l . It i s scanned once only. Samples with fluorescein can be scanned more than once by varying the position of the probe's focus slightly each time. Also, plasma fluorescein samples are usually scanned in chrono- logical order of their p.i . times. SUBJECT DATA ENTRY i s usually executed last. 67 V - A N A L Y S I S A N D D I S C U S S I O N 5 . 1 C l a s s i f i c a t i o n s The procedure vas completed on the following sample population: TYPE NUMBER AGE RANGE MALE Diabetic 10+2 21 - 64 FEMALE Diabetic 5 18 - 57 MALE Normal 4 27 - 38 MALE MS 3 32 - 39 FEMALE MS 13 20 - 59 TOTAL 35*2 Table 2. Distribution of subjects. Two male diabetic subjects were recalled at one and three months after their f i r s t scans to study the reproducibility. The two sets of results for these two subjects were averaged in the following analysis. Only three blood samples were taken from the f i r s t two male and the f i r s t two female diabetic subjects on whom the procedure was carried out. Whenever possible each parameter studied was classified and tested on the basis of disease (MS or diabetes), age, and sex. Only l e f t eyes were scanned because of the time restriction for the bolus, scans i.e., a one-eye design [49], Diabetics with non-proliferative retinopathy were separated into three groups according to the severity of capillary disease as 68 shown by fundus photography and fluorescein angiography. Group D(l) consisted of subjects with no or early DRP (0 to 5 microaneurysms); group D(2), subjects with mild to moderate non-proliferative DRP, and group D(3), subjects with severe non-proliferative DRP. DISEASE STATES NUMBER D(l) - Diabetic with zero to early 7M DRP ( 0 - 5 microaneurysms). IF D(2) - Diabetic with mild to IM moderate DRP. 2F DO) - Diabetic with 2M severe DRP. 2F TOTAL DIABETIC SUBJECTS AVAILABLE 10M:5F MSU) : stable 2M:3F recovering from relapse 1M:2F MS(2) : slowly progressive 0M:4F in relapse 0M-.4F TOTAL NUMBER of MS subjects AVAILABLE 3M:13F MS(3) : benign 1M:1F relapsing-remitting 1M-.6F MSU) : relapsing-progressive 1M:5F chronic progressive 0M:1F TOTAL NUMBER of MS subjects AVAILABLE 3M-.13F MS(5) : no periphlebitis 2M:5F MS(6) : active periphlebitis 0M:2F MS(7) : inactive periphlebitis 1M:3F TOTAL NUMBER of MS subjects AVAILABLE 3M:10F Table 3. Detail subject classifications. 69 MS subjects were grouped for analysis i n three ways. The f i r s t was by the a c t i v i t y of MS at the time of the scans. The subdivisions comprised MS<1) - subjects who were e i ther stable or recovering from a relapse or had recovered completely from a recent relapse at the time of scanning and MS(2) - subjects who were e i ther i n slow progression or i n relapse. The second was by the standard c l i n i c a l a c t i v i t y categories of MS as reported i n case records. The subjects were again subdivided into two sections - MS(3) and MS(4). MS(3) included those c l a s s i f i e d as benign (only one attack) or re laps ing- remit t ing (almost complete recovery from each attack) . MS(4) subjects were c l a s s i f i e d as relapsing-progressive, or chronic-progressive. The t h i r d grouping separated the subjects according to the absence or presence of e i ther active or inac t ive r e t i n a l p e r i p h l e b i t i s . These were ca l l ed MS(5), MS(6) and MS(7) respect ive ly . As not a l l subjects were examined for these states, the sample s i ze for t h i s grouping was reduced. 5 . 2 F - N u m b e r s The F-numbers i n Section 2.2, were calculated by SUDATA.BAS for each subject. The average and standard deviat ion were found for each in t ra -ocu la r compartment and tested against the averages from G u l l - strands emmetropic model eye i n Table 1. The re su l t s were calculated using the data from 34 subjects only. Three subjects (from Table 3) were excluded because they did not have the ultra-sound scans. As F- numbers are character i s t ic s of the system, the subjects were not separated in to age, sex or disease states . 70 MEDIUM AVERAGE S.D. Aqueous 1.735 0. 359 Lens 1.511 .; 0.157 Vitreous 1.245 0.089 Table 4. Average F-numbers and their S.D. Since 34 measurements were made, the mean values in Table 4 were tested against those in Table 1 using a normal(0,1) distribution test. A P=l% level of significance was imposed. The results showed that the F-numbers from Table 1 were not applicable to this VF system: the F-numbers in the two tables were significantly different. Hence, the Luma" contact lens (which replaces the Goldmann contact lens in Lund-Andersen's calculations) changes the F-numbers s i g n i f i - cantly. 2.4- 2.2- 2-1 ce CD I 1.6-1 u. 1.2- a • 8* CP a Legend A V I T R E O U S X L E N S a A Q U E O U S 20 SO 40 SO AGE in y e a r s 60 70 Figure 19. Plot of F-number results. 71 5 . 3 I n t r a - o c u l a r L e n g t h s The averages for the vitreous and aqueous chamber depths, the lens thickness and the total axial length as measured by ultrasound for male and female subjects are shown below. The sample sizes were 17 in each group. AGE RANGE SEX VITREOUS LENS AQUEOUS AXIAL NUMBER 16-20 F 15.76 3. 38 4.45 23.59 3 21-30 H 17.96 3. 67 3. 60 25.23 7 F 16,24 3.38 3. 96 23.58 6 31-40 M 16. 44 3.81 3.56 23. 81 9 F 15. 94 3.96 3.42 23. 31 2 41-50 F 16.13 3. 84 3.48 23. 46 2 51-60 F 14.45 4.32 3.22 21.99 2 61-65 M 14. 78 4. 54 3. 06 22. 38 1 TOTAL 23.74 34 Table 5. Average lengths of the intra-ocular media (in mm). Although the sample sizes were small, and the various disease states were not taken into account, the above table suggests either that axial length decreases with age, or, that the lens thickens with age. [50] Linear least-squares f i t s were done on each treatment. The null hypothesis, H*, in each case was that the slope of the straight line was zero. The alternative hypothesis, H", for the axial length test was that i t decreased with age, (i.e. a negative slope); while for the lens, i t thickened with age (i.e. a positive slope). The levels of significance <P) at which H" would be rejected were found in a t(17-2) test. The results are shown in the Table 6. 72 a. M A L E I N T R A - O C U L A R L E N G T H S 30 - i 25 E 2 0 E c X '5 I— o 2 a a a a 3 n a A ^ 20 30 40 50 A G E in y e a r s x • 60 70 L e g e n d A VITREOUS X LENS • AQUEOUS 3 AXIAL b, F E M A L E I N T R A - O C U L A R L E N G T H S 30 -, 25 £ 2 0 t .c ir is-t— o ~z. UJ ~" 10- A A A a 3 S3 A A X CP 20 30 40 SO A G E i n y e a r s 60 L e g e n d A VITREOUS X LENS a AQUEOUS 3 AXIAL Figure 20. Intra-ocular Lengths. 73 TEST MALE FEMALE Lens vs Age < 1% < IX 3.79 */- 0.35 3.72 •/- 0.55 Axial vs Age < 5Y. < 2.57. 24.31 */- 1.65 23.16 + /- 1.05 Table 6. Significance levels (P) for intra-ocular lengths tests. Average S.D. (mm) calculated from Table 5. MALE FEMALE L = 26.79 - 0.07»A r = 0.45 L = 24.49 - 0.04»A r = 0. 52 T = 3.11 • 0.02»A r = 0.59 T = 2.76 + 0.03»A r = 0.72 Table 7. Results of linear, least- squares f i t of axial length (L), and lens thickness (T) in mm to the subject's age (A in yrs). r = linear correlation coefficient. Hence, i t would seem that the crystalline lens thickened with. age. The spread of the data was significant when comparing case by case. This i s observed in the low value of the linear correlation coefficients, r of the f i t s in Table 7. 74 5 . 4 R l a s m a . CULT - V e> — £ d_ "t s The four (2+1)-parameter f i t t i n g polynomials in BLOOD.BAS can be reduced to two. The f i t s to the 35+2 cases in Table 2 showed that most plasma profiles were (as expected) best described by the loga- rithmic or the exponential forms. The simple parabola was a poor f i t for what was expected to be a fast exponential-decay-type behaviour of the two-compartment model. It produced a minimum between t=30 and t=S0 minutes p.i. in almost a l l cases. The one that was accepted had a minimum that was situated in the neighbourhood of t=60 minutes p.i. Fit #3, the second-order polynomial in 1/t, usually rose from negative values to a maximum in 0. 5 < t < t " 1 ' , the f i r s t sampling time. This resulted in negative areas upon integration. Such results were rejected even though the reduced chi-square might have been the smallest among the f i t s . SUBJECT TYPE FIT FUNCTION TOTAL #0 . #1 . #2 . #3 Male Diabetic Female Diabetic 0 0 6 3 4 1 2 1 12 5 Male Normal 0 3 1 0 4 Male MS Female MS 0 1 3 7 0 3 0 2 3 13 TOTAL 1 22 9 5 37 Table 8. Plasma f i t s . 75 Testing the above table in a two-way classification without interaction at P=57. shows that the types of disease (rows) did not have significant effect on which function was the best f i t . There was significant difference between the type of best-fit functions: function #1 was significantly the most probable result. As mentioned before, the amount of dye in the blood rises from zero at t=0 (injection) to a maximum in less than one minute then begins to f a l l . The area depends on the f i t which i s defined by the number of samples collected. The more samples taken within the one- hour p.i. interval, the better defined the f i t . However, with only four samples, (or less in 4 diabetic subjects), the area between t=0 and t ( l > may have significant effect on the integration. Hence, the effect of the time at which the f i r s t blood sample was taken on the resulting f i t was investigated. The treatments were the intervals containing t , l ) . The replications [51,52] were the f i t t i n g functions. Sex was not considered. At P=10X, t ' 1 ' did not have a significant effect; the types of f i t were significant (as shown before). There did not seem to be any interaction between the types of disease and t ' 1 ' . The above results imply that the curve-fitting in BLOOD.BAS may be shortened to save running and printing time: the f i r s t and the last function(s) may be omitted from consideration. Table 9 shows the distribution of the 35+2 f i t s in the t' 1 ' intervals. Note that these intervals were arbitrarily chosen. 76 INTERVAL of t< » ' in minutes p.i. FIT FUNCTION #0 #1 #2 #3 TOTAL < 6 0 2M 0 0 0 3D 3D 2D 0 2N IN 0 13 6 to < 10 0 8M 3M 2M 0 3D 2D 0 0 IN 0 0 19 > 10 in* 0 0 0 0 2D 2D 0 0 0 0 0 TOTAL 21 11 37 Table 9. Effect of the f i r s t blood sampling time, t l 1 1 M = MS; D = Diabetic; N = Normal LLoD TYPE AVERAGE S.D. NUMBER Male Diabetic 4.32 1.25 10+2 Female Diabetic 4. 57 1. 84 5 Male Normal 3.99 0.68 4 Male MS 4.62 3.19 3 Female MS 4. 51 2.10 13 FINAL 4.41 1.72 35+2 Table 10. Average LLoDs and S.D.s. (ng. ml"' ) 77 The LLoD was calculated from the background scans (0.AVG f i l e s ) i n the i n t e r v a l between 8 and 10 mm from the r e t i n a . S p e c i f i c a l l y , the i n vivo LLoD i s defined i n t h i s study as the average concentra- t i o n i n the i n t e r v a l plus twice the root-mean-square value of the standard deviations of the data col lected i n that i n t e r v a l . An analysis of variance at P=257. showed that the types of disease did not affect the averages i n each treatment. This was expected because the pre- in ject ion scans were dye-free. Hence, the average LLoD was equivalent to a concentration of 4.41 +/- 1.72 ng .ml- ' . This value compares favourably with the i n v i t r o LLoD of approximately 5" 1 that was estimated during concentration c a l i - brations, and as stated i n [14]. 5 - S A u t o f l u o r e s c e n c e Plots of lens autofluorescence vs age of subject, and duration of disease are shown on the next page. Least-squares s t r a i g h t - l i n e f i t s were found for each disease category. t(n-2) tests on the slopes of the f i t s were performed on H * : the slope was 0 i n each case; and on H * : the slope was pos i t ive i n each case. Table 11 shows the resul t s of the tests against age only. These resu l t s imply that lens autofluorescence increased with age [19,35]. Also, the averages, intercepts and approximately equal slopes of the f i t s suggested that lens autofluorescence was higher and occurred e a r l i e r i n the diabetic subjects than the normal and MS subjects. However, t h i s trend was not c l e a r l y defined between the MS and the normal subjects as the slopes were d i f fe rent . 78 a. 300 LENS AUTOFLUORESCENCE vs AGE 250 A A C y 200 o in £ 150 o z> —1 O 100 -n < on £ 50' x 2< c ^ • A 10 20 30 40 50 AGE in years SO 70 Legend A DIABETIC X NORMAL • MS ' b. LENS AUTOFLUORESCENCE vs DURATION 300- A A * 250- o> c 200-.c A O X !< 150- X A x CJ ~r 100- A O a 50- * A X A & . X A X Legend A X ' A 0IA8ETIC A X MS 0 - 0 5 10 15 20 DURATION in years Figure 21. Autofluorescence. 79 DIABETIC NORMAL MS Sample size, n 15 4 16 Correlation, r 0.65 0. 93 0.84 Average age (yrs) 30.2 32.0 37.9 Average reading 121.8 55.4 69.4 Standard Deviation 83. a 20. 4 43.4 Slope ("'.yr"1) 4.11 4.20 2. 93 Intercept ( -2.32 -78. 0 -41. 7 P(reject H*) 17. 57. 17. Table 11. Lens autofluorescence, S.D. ("1) vs age. The duration of diabetes did not have a significant effect on lens autof luorescence (P>25V.), and was not included. Autof luorescence versus state of DRP was tested. D<1> D(2) DO) Number 8 3 4 Average 69.94 184.9 178.3 S.D. 35.36 73.7 104.0 Table 12. Autofluorescence, S.D. ("') vs DRP states. t-tests on differences showed that D(l) and D(2), D<1) and D(3) were significantly different (P=107.); while D<2) and D(3) were not (P>257.). Note that the S.D.s were sizeable fractions of the averages, i.e., there was a large variation from case to case. (An analysis of variance at P=2.5% showed that at least two of the means were not the same.) One conclusion i s that lens autofluorescence increased 60 s i g n i f i c a n t l y v i t h progression from no DRP to severe DRP [19,35], These re su l t s par t ly explain the lov cor re l a t ion coe f f i c ient , r=0.65, for the c u r v e - f i t t i n g i n Table 11. Diabetic subjects betveen the age of 18 and 27 years v i t h severe DRP reduced the goodness of the s t r a i g h t - l i n e f i t . Hovever, i t has not been e x p l i c i t l y shovn that the increase i n lens autofluorescence i s due to the thickening of the lens v i t h age. This requires a much larger sample s i ze of normal subjects spanning a large age range. MS STATES AVERAGE S.D. NUMBER MS(1) 69.96 48. 93 8 MS(2) 68. 85 40.45 8 MS (3) 47. 38 27.69 9 MS<4) 97.71 44. 93 7 MS<5) 69. 74 54.35 7 MS(6) 73.03 49.66 3 MS(7) 67. 30 29- 60 3 Table 13. Autofluorescence, S.D. (" 1 ) vs MS states. There vas no s ign i f i cant difference betveen the means of MS(1) and MS(2), or that of MS<5), MS(6) and MS<7), (P>25*4). Hovever, the means of MS(3) and MS(4) vere s i g n i f i c a n t l y d i f ferent (P=2.57.). Note the large S.D. Although the trend of increasing autofluorescence v i t h age vas. seen i n Table 11, the c l a s s i f i c a t i o n i n terms of graded severi ty of recent c l i n i c a l a c t i v i t y did not shov trends s i m i l a r to the cases v i t h diabetes. The d i s t r i b u t i o n of those older subjects among the MS 81 groups had "biased" the results. No trends were observed except that autofluorescence was higher than for normals. The trends established thus far are in accord vith other researchers' findings. Hovever, comparing average readings, the values here are much lover than from [351. One possible reason for this i s that the instrument characteristics and calibrations vere different. For example, the angle betveen beam and probe vas different. Hovever, i t i s more li k e l y that alignment errors vere the cause of the differences. 5 . "7 P r o f i l e s The effects of CR tailings vere studied in a D(l) subject. (Figure 22.) Some of the scanning problems discussed in Section 2.1 are demonstrated in these profiles. The excitation beam in the 2-minute scan vas probably partially clipped by the i r i s as suggested by the plateau shovn vithin the f i r s t 2 mm from the retina. The tailings hovever, coincide with the 4-minute scan for distances greater than 2 mm from the retina. The 4-minute scan peaked at about the position of the retina that was located visually. Note that the tailings at this time persisted well into the vitreous where, for this subject at this p . i . time, no dye was expected to have penetrated. The 6-minute profile shows the d i f f i c u l t y in visually locating the retina. The peak i s approximately 1 mm anterior to the retina. This particular bolus alignment error can be corrected by setting the zero at the CR peaks, but for later measurement scans the actual 82 100CH DISTANCE FROM RETINA in m m Figure 22. Bolus effects. position of the retina by CR peaks can not be clearly defined in practice or in theory. Figures 23a-23g show the evolution of the dye profile in the posterior vitreous of subjects in various classifications. Note that the vertical log scales are different. The important points are as follows - a) the prominance of the bolus profiles and CR peaks. b) the change of slope with time about 3mm. c) alignment and peak shifts anteriorly. d) the difference in the concentrations farther from the retina. The profiles of the normal (Figure 23a) are "noisier" than 83 cn c _c z g o o L e g e n d A s - u i N u r c x 32-uiNlirE Q 0 - U I N U I E a S O - M I N U T E — I 1 : 1 r - t 2 J 4 5 DISTANCE FROM RETINA in mm c c < Z UJ CJ z O <J X X a , L e g e n d A 4-UINUTE X 37-UINUTE a S2-UINUTE *x a a a a A X ^ ( V x̂ 0 1 2 3 * 3 S OISTANCE "ROM RETINA in mm a. Normal. b. DU) - no DRP. c o 5? CJ z o <J x g^-^o^ q_p a a r &<. L e g e n d A 3-UINUTE X 31-UINUTE 1 0 0 a S I - U I N I J T E c 1 a O - U I N U T E -S **** o "HOT, 4 1 2 3 4 5 DISTANCE FROM RETINA in mm p: 10' < L e g e n d A 4-UINU1E X 51-UINUTE • 53-UINUTE a 0-MINUTC ., _ 31., ^ 0 1 2 3 4 5 DISTANCE FROM RETINA in mm c. D<2) - moderate DRP. d. D(3) - severe DRP. Figure 23. Sample profiles. 84 < C E U z o o L e g e n d A 3-uiNUTE X 30-UINUTE • SO-UINUTE 9 O-UINUTE Hi U1 (TA A 1 2 3 * 5 DISTANCE FROM RETINA in mm o> c = to-< es o z o u " * ^ x L e g e n d x 31-uiNUTE 0 SI-MINUTE a O - U I N U T E 3 1 2 3 * S DISTANCE "ROW RETINA in mm e. Stable MS. f. Relapsing MS. 5 'oo J o i= 10 < O L e g e n d A 5-MINUTE X 27-UINUTE O 60-MINUTE 8 0-MINUTE 0 1 2 3 4 5 DISTANCE FROM RETINA in mm f. Liquefied vitreous. Figure 23. Sample profiles (continued), 85 others as one might expect from the lower concentrations of dye. A l l profiles more or less coincide at 6mm unlike for others, especially the DO) and the MS cases in Figures 23f,g. The special case of the female, MS subject with the liquefied vitreous in Figure 23g is worth noting. The gradient of the 1-hour profile (at the posterior vitreous) i s small compared to the bolus at various positions from the retina. The 60-minute CR peak i s thus not well defined. However, as the profile about the 3-mm point was •f l a t " , misalignment should not produce large errors in this special case. However, regardless of the state of the vitreous, Figure 23g shows that a large amount of dye had indeed entered the posterior vitreous through the BRB. Figure 24a compares the 1-hour scans of two stable, MS sub- jects' profiles. The male, MS(4) profile i s very similar in slope and magnitude to the MSO), female subject. Both are elevated above the normal subject's profile. The plateau between 2 and 3 mm i s probably due to movement by the subject during the scan C321. The male subject had no periphlebitis; the female subject was not examined for this. Figure 24b compares the 1-hour scans of two female, MS subjects, with active and inactive periphlebitis respectively. The latter subject was the person with the liquefied vitreous. Both belonged to the MS(2) and the MS(4) groups. Note that the MS profiles are clearly above the normal. Figures 25a,b,c compare a male and a female diabetic person's 1-hour profiles in the D(l), D(2) and DO) groups respectively. The magnitude of leakage (vertical axis) i s progressively greater from 86 cn c z o cc I— z UJ o z o X* A * ^ A " A A, • *S A L e g e n d A N O R M A L X 1 4 5 . M S • U ' . M S A A ^bg«S A A " . * ^ A A A A < £ 1 2 3 4 5 DISTANCE FROM RETINA in mm z o < z L L J o O o CD CP x*x*x L e g e n d A N O R M A L X 2 4 6 . M S Q 2 4 7 . M S A " A A A A A A A - V * X A A A A A ^ 1 2 3 4 5 DISTANCE FROM RETINA in mm Relapsing-Remitting vs Relapsing-Progressive. b. Relapsing-Progressive vs liquefied vitreous. Figure 24. Comparison of MS profiles. (Refer to Table 14 for number codes.) D<1) to DO). This demonstrates good correspondence vith c l i n i c a l grading of severity of DRP. The differences in the magnitude of leakage in Figure 25a of D(l) profiles may have been accentuated by poor control of diabetes in one case despite the presence of less than 5 microaneurysms. The leakage i s similar to that of a normal subject vhen there i s good control of diabetes. Figure 26 compares the 1-hour scans of the following subjects: a) a male subject in the MS<1), MS(4) and MS(5) groups. b) a female subject in the MS(2), MS(4) and MS(6) groups. c) a female subject in the D(l) group. d) a female subject in the D(2) group. e) a male subject in the DO) group. These illustrations of profiles are intended to demonstrate 87 CO c c z o a: o z o z o < O z o K t — Z UJ o z o *x A ^ L e g e n d A NORMAL X 1F.DIAB A , A X A ^ ^ A A A A * AAA 1 2 3 4 5 DISTANCE FROM RETINA in mm' xxx 0 L e g e n d A NORMAL X 2F.DIAS • 2M.0IAB \ A~" ̂ A A A A I S A A A AA A 1 2 3 4 5 DISTANCE FROM RETINA in mm L e g e n d A ^ A A A V %3? • a ^ AA AA A 1 2 3 4 5 DISTANCE FROM RETINA in mm a. Male vs Female D(l), b. Male vs Female 0(2). c. Male vs Female D(3). Figure 25. Comparison of diabetic subjects' profiles. 88 4 g> 100 < z UJ u o L e g e n d a. 1 J J C A 0 I 2 3 4. 5 6 DISTANCE FROM RETINA in mm Figure 26. Overall comparison. (Refer to Table 14 for number codes.) different magnitudes of leakage vithin MS and diabetes. Profiles provide qualitative comparisons of the integrity of the BRB - such as visib l y different gradients (at some point), or, that one l i e s above or belov the other. In the calculations of PR3 or P1, the division by the results of the plasma integral may produce quite different quan- t i t a t i v e descriptions of the BRB. 5 . S D i f f u s i o n C o n s - t e r n - t The diffusion constant, D vas calculated by SLOPES.BAS. Table 14 gives the results of averaging over a l l measurement scans made. Note that D and a l l calculations pertaining to i t are alvays given in units of *10-' em's"1. Least-squares f i t s to D = A + B»(age), and D = X • Y»(duration). vere carried out on each disease group vith H* : slope = 0, and, HA : slope < 0 or > 0, i.e., one-sided t-tests depending on the coeffi- cient, B or Y of the f i t s . 89 AGE DIABETIC NORMAL MS 18 4.44 (Fl) 19 0. 85 (F3) 20 4. 42 (F235) 21 3.73 (Ml) 22 3.69 (Ml) 1. 35 (F23?) 9. 78 (Ml) 23 0. 55 (Ml) 25 5.18 (F135) 26 2. 94 (F3) 27 7.38 (M) 17.13 (F135) 28 7.18 (Ml) 2. 54 (M3) 29 5.22 (Ml) 30 1.84 (F2) 6. 42 (F13?) 31 6.02 (Ml) 6.72 (M) 32 4.12 (M) 3.62 (M137) 35 3.58 (M2) 7.11 (M145) 9. 33 (F236) 37 5.22 (F247) 38 15.15 (M> 4. 34 (F246) 39 11. 49 (M135) 44 29.90» (F247) 49 8.72 (F247) 56 27. 59 (F135) 57 2.69 (F2) 59 18.68 (F145) 13.13 (F24?) 64 2. 31 (M3) AVERAGE 3.83 8.34 10.86 S.D. 2. 44 4. 75 8. 49 NUMBER 10M:5F 4M:0F 3M:13F D averaged over a l l measurement scans. CODES : (XA) for diabetics; (XBCD) for MS where X = Male or Female A = 1,2,3 for D(i),D(2), D(3) respectively; B = 1,2 for MS(1),MS(2) respectively; C = 3,4 for MS(3),MS(4) respectively; and, D = 5,6,7 for MS(5),MS(6), MS(7) respectively. D = ? means subject not examined. » i s the subject with liquefied vitreous. 90 DIABETIC NORMAL MS TOTAL Sample Size 15 4 16 35 Coefficient, A 4.82 -16.06 -4.96 -1.08 Coefficient, B -0.03 0. 76 0.42 0.25 Correlation, r 0.18 0. 73 0.62 0.46 P(reject H«) > 25% 257. 17. < 0.57. Sample Size 13 9 Coefficient, X 5.84 4.61 Coefficient, Y -0.22 1.14 Correlation, r 0. 53 0.63 P(reject H«) 57. 57. Table 15. Tests of D vs age and duration. AVERAGE + /- S.D. NUMBER P D<1) 5. 08 2. 73 8 257. D(2) 2.70 0.87 3 2. 57. D(3) 2.16 0. 91 4 0. 57. Male Diabetic 4.46 2.67 10 107. Female Diabetic 2.55 1.33 5 0.57. DIABETIC AVERAGE 3.83 2. 44 15 0. 57. Male Normal 8.34 4.75 4 257. MSU) 12. 15 8. 33 8 57. MS(2) 9.58 9. 01 8 257. MS<3) 9.61 8.25 9 257. MS(4) 12. 47 9.17 7 107. MS<5) 13.08 8. 50 7 57. MS<6) 6.84 2.50 2 > 257. MS<7) 11.87 12.21 4 257. Male MS 7.41 3. 94 3 > 257. Female MS 11. 66 9.15 13 2. 57. MS AVERAGE 19.86 8. 49 16 2.57. FINAL AVERAGE 7. 56 6.90 35 97. Table 16. Diffusion constant, D by sex-disease states. 91 The results in Tables 14 and 15 show that D tended to increase with age, This result was dominated by the larger MS sample with numerically more older subjects. Also, the B-coefficient for dia- betic subjects was negative but not significantly so. Others had also noted this trend but were unable to prove i t s t a t i s t i c a l l y C533. The tests of D against the duration of diabetes (Table 15) were inconclusive because the sample sizes were small, and the Y-coeffi- cients for diabetes and MS were of opposite signs, showing opposite trends. This may be due to intra-group variation in severity. Larger samples of each disease state are needed to establish the existence of any trend. Note the significant scatter of data as the corre- lation coefficients, r, are not close to 1. (Figure 27.) The extreme right column in Table 16 shows the results of the tests for H* : average D in each group = D" , where D"=6 [20] i s the diffusion constant of the dye in water. Despite the large spread of the data in each group, one notable result was that the D's for diabetics were significantly lower than D", for this sample. The reason for this i s not known. Similarly, the explanation for the D(2) and the D(3) averages being half that of D(l) i s not apparent. It should be noted that the D(l) sample was twice the size of D(2) or D(3). In contrast, the average D for MS subjects was significantly, higher than DM. As stated before, the high values came from most of the older, MS subjects, notably, the female, MS subject with lique- fied vitreous who had the highest value. This result may be due to 92 z o a 5 •£ 3.0- Lscsnd 4 M-auacnc • NORMAL :0 20 JO *a SO 50 70 AGZ ;n y e a r s < 23.0- & l/l £ u 20.0- Z < 13.0- C u z ta.o-c u. DIF  3.0- 0.0- Legend A NORMAL X M-MS C F-MS 20 20 40 30 SO AGZ in y e c r s a. Diabetic subjects. b. MS subjects. Figure 27. Diffusion ocnstant vs age. the effect of mechanical mixing on the difference equation, Eq. 32. Other tests on the results in Table IS shoved that the value of D for D(l) vas significantly different from those of D(2) and D(3) (P<5/£). There vas no significant difference betveen D(2) and D(3) <P>257.). Betveen MSU) and MS(2), MS(3) and MS(4), and MSC5), MS(6) and MS(7), there vas no difference (P>25/C). These results vere due to the large spread of data in each group vith high values distributed throughout a l l groups. The diffusion coefficient in the normal eye vas found by others to be a) 13.3 >/- 6.8 and 11.9 +/- 5.4 (Chahal, et al. [23]), b) 13.2 +/- 4.3 (Ogura, et al. [531), c) 7.4 •/- 3.4 (Lund-Andersen, et al. [54]). 93 E ° 20.0 c z < ( / I z o o 1/1 3 L e g e n d «. DIABETIC x MS 0 5 10 15 20 DURATION OF DISEASE in years Figure 28. Diffusion constant vs duration. For diabetic eyes, D = 9.6 +/- 2.0 [54]. Note that the large S.D.s allow for much overlap. Comparing these results to those in Tables 15 and 16, for P=55C, there was no difference between the average found here and those derived by others for normals. Individual results in Table 15 were within the range in (c); but, somewhat lower than those values in (a) and (b). Comparing the averages of any diabetic group, or, of individual cases, a l l were found to be significantly lower (P<0.5%) than the value stated by Lund-Andersen, et al [54], Many of the D values for MS subjects were within the quoted values for normal and diabetic persons. The exception was the lique- fied vitreous case for which D was higher than a l l others. D values were also elevated for several older MS subjects. No explanation in terms of c l i n i c a l activity i s known. (Refer to Table 14.) 94 The results of D in Table 14 were derived for profiles aligned by RET only. Misalignment might account for some of the extreme values. As the average over a l l measurement scans was used, this should alleviate the alignment errors. Also, no correlation was seen between D and the PR3 values studied below. 5 - 9 P e n e t r a t i o n R a t i o Recall that PR3 (in units of »10"* s"1 ) was calculated in two methods of alignment by the programme, C/VAZ.BAS: . by RET, and, by CRP. (Refer to Section 3.4.) The results are presented in Table 17. In some cases, PR3 was more than halved when the alignment was changed from RET to CRP; in others, i t remained approximately the same. These changes are evidence of the d i f f i c u l t y of locating the retina by sight. A case in point was the 22-year old, MS subject whose 1-hour, PR3 value was 48.2 by RET - the highest of a l l MS subjects; while, by CRP, i t f e l l to 13.1. Her fluorescein profile at 1-hour p.i. showed that the CR peak was more than 2 mm from the located zero position. Although she was not examined for the presence or activity of peri- phlebitis, her profiles were not distinctly different from other MS subjects in the MS(2) and MS(3) groups, but, were when compared to the MS(6) category. Another noteworthy result i s that of the 32-year old normal who • i s the brother of the 35-year old, MS subject. His PR3 values were about 3 times that of other normals. His sister, who, at the time of scanning, was in relapse, had lower PR3 values than his (in either 95 DIABETIC NORMAL MS nut RET CRP RET CRP RET CRP 18 2.4 (Fl) 4.8 19 294.2 (F3) 103.1 20 16.2 (F235) 15.7 21 29.5 (Mi) 18. 0 22 11.2 (Ml) 12.1 48.2 (F23?) 13.1 10. 0 (Ml) 32.7 23 1.4 (Ml) 16.6 25 30. 6 (F135) 17.6 26 144.4 (F3) 63.8 27 6.2 (M) 3.1 4.6 (F135) 4.2 28 1458.5 (M3) 886.8 27.5 (Mi) 22.1 29 10.4 (Ml) 6.5 30 11.1 (F2) 8.3 22.8 (F13?) 11.7 31 8.5 (Ml) 14.6 15.4 (M) 15.8 32 32.3 (M) 22.3 4.0 (M137) 1.6 35 30. 0 (M2) 24. 6 11.2 (M145) 7.7 25.2 (F236) 18.2 37 12. 2 (F247) 16. 2 38 8.4 (M) 8.9 37.0 (F246) 33. 5 39 14. 7 (M135) 6. 7 44 31. 9» (F247) •29.0 49 9. 5 (F247) 9. 5 56 9.9 (F135) 10.1 57 13. 0 (F2) 12.9 59 2.4 (F145) 1.9 5.2 (F24?) 5.8 64 138.2 (M3) 66.1 No. 10M:5F 4M:0F 3M:13F Table 17. PR3 averaged over a l l 55-70 minute scans, after background subtraction only. CODES : (XA) for diabetics; (XBCD) for MS vhere X = Male or Female A = 1,2,3 for D(1),D(2),D<3) respectively; B = 1,2 for MS(1),MS(2) respectively; C = 3,4 for MS(3),MS<4) respectively; and, D = 5,6,7 for MS<5),MS(6),MS(7) respectively. D = ? means subject not examined. • i s the subject with liquefied vitreous. 96 z o z UJ 0 . Legend a N O R M A L x l-o .0 2-0 a j -o 3 M S x " " a 3 30 »0 50 60 AGE in years O 5 z o z UJ a. Legend 4 1-0 x 2-0 a 3-0 5 N O R M A L a M S 30 40 so so AGE in years a. By RET. b. By CRP. Figure 29. Penetration Ratio. alignment)! This anomaly might have been due to improper instrument calibrations (settings) at that time; otherwise, i t cannot be explained. This subject (and his sister) w i l l have to be recalled for further testing. His reading vas omitted from analysis. (Note that other data such as F-number calculations vere s t i l l admissible.) The results of statistical-inference testing vere anticipated by ranking PR3 results in ascending order (by RET and by CRP sepa- rately) for MS subjects. The f i r s t noticeable point vas that a l l MS(6) subjects had higher values. The case of (F247), the subject vith the liquefied vitreous, also had a high PR3. A l l other MS clas- sifications vere distributed throughout the order vith no obvious •clustering*. This implied that there vas no detected difference betveen MS(1) and MS(2), MS(3) and MS<4), MS(5) and MS(7). 97 The highest PR3 results calculated was that of a 22-year old, male, D(3) subject. Although he has severe DRP, his reading was about 10 times higher than other D(3) subjects which may be an intra- group variation. The 1-hour profile showed indisputable, elevated levels of dye in the vitreous. His PR3 value was also omitted from a l l testing. The ordering of diabetic PR3 results showed that the results of the D<3) group were consistently the highest values. Cases with obviously high amounts of leakage were clearly detected (and detec- table) by this fluorophotometer. This " j u s t i f i e s " the omission.of.the above D(3) case because the other D(3) results, taken individually or together, were already significantly higher than those of the other two diabetic groups. 1 1 •' GROUP RET CRP MEAN +/- S.D. # MEAN +/- S.D. # D(l) D(2) D(3) 12.6 10.5 8 18.0 10.4 3 192.3 88.3 3 15.9 8.9 8 15.3 8.4 3 77.7 22.1 3 NORMAL 10.0 4.8 3 9.3 6.4 3 MSQ) MS(2) MS(3) MS(4) MS(5) MS<6) MS(7) 12.5 9.9 8 19.6 12.0 7 16.0 9.7 8 15.6 13.4 7 12.8 9.3 7 31.1 8.3 2 14.4 12.2 4 7.7 5.4 8 17.6 9.4 8 11.0 5.9 9 14.8 12.1 7 9.1 5.8 7 25.8 10.8 2 14.1 11.6 4 Table 18. PR3 Average •/- S.D. of the various groups. 98 The DU) and D(2) classifications, l i k e the MS(1), MS(2), and MS(3), MS(4) groups, did not distinctly separate out in either the RET or the CRP sorts. This could imply that intra-group and inter- group fluctuations vere significant. Poor alignment and/or an algo- rithm not optimized for such calculations could also have prevented the appearance of any expected order. PR3* results by SLOPES.BAS ordered in the same manner as those by RET. This vas expected because SLOPES.BAS vas vritten to approxi- mate RET results by curve-fitting. (Refer to Section 3.4.) The order for the D<3) group vas exactly the same but the PR3 values vere about 10'/. greater than the PR3* values. On average, hovever, PR3» values vere neither alvays greater than nor alvays lover than PR3 (by the sign test at P=5V.). PR3* results from SLOPES. BAS could then be used to check C/VAZ.BAS's RET PR3 results. Either set of results could be consistently used to represent the penetration ratio vhen only align- ment by RET vas considered; computing time could be shortened by choosing to run one of the tvo programmes only. It i s observed from Tables 17 and 18, that most individual and average results of diabetic and MS subjects vere higher than those of the normals, despite the large S.D.s. Although age-matching tests betveen the members of each group vere not possible because of the small sample sizes, such trends demonstrated that differences betveen groups and indivduals existed (for this sample) and vere detected! Tests of the significance of the differences betveen the PR3 means betveen any tvo groups vere carried out. Analysis of variance vas used to test 99 H* : the means of any 2 groups were the same, i.e. their difference vas 0, against Ha : the means of any 2 groups vere different, i.e. their difference vas not 0. The P(reject H*) are shovn in the respective tables belov. The entries in the upper triangles are for RET alignment. Those in the lover triangles are for CRP alignment. BY RET NORM DU) DO) DO) B NORM »•*» > 25 > 25 2.5 Y D(l) > 25 » »* * > 25 < .5 C D<2) > 25 > 25 * * » # 5 R P DO) 1 < .5 2.5 • *•» Table 19. Significance level (X) to reject H* betveen diabetic and normal groups. In the above table, only the DO) group vas clearly and signi- ficantly different from a l l other groups. But, the results in DU) and D(2) ranged from 1.4 to 30, and 6.5 to 32.7 for RET and CRP respectively. These indicate that breakdovn of the BRB had already occurred and vas detected in subjects vithout signs of DRP but vere. at the early stages [4,5]. The only significant difference for the groups in Table 20a belov i s that betveen MSU) and MS(2) in CRP. In Table 20b, MS(6) i s 100 significantly different from MS<5) and normal but not from MS(7). Betveen other groups, there i s no significant difference. Table 18 shovs the large S.D.s of these groups. BY RET NORMAL MS(1) MS(2) MS(3) MS(4) NORMAL **»*»» > 25 25 > 25 > 25 B Y MS<1) > 25 • • * » • 25 * * * * * ***** MS(2) 25 2.5 ***** ***** • *•*.* C R MS (3) > 25 ***** ***** * * * * * > 25 P MS(4) > 25 ***** • •••* > 25 * * * * * (b) BY RET NORMAL MS (5) MS(6) MS<7) B Y NORMAL MS(5) > 25 > 25 • •••• 5 5 > 25 > 25 C R P MS(6) MS(7) 25 > 25 2.5 > 25 • ••*» > 25 25 • •••• Table 20. Significance level (.'/.) to reject H* betveen MS and normal groups. One reason for the great dispersion of data in MS<7) i s the placement of the case of the liquefied vitreous. Her classifications, place her into the respective categories, but i t i s not clear i f her results should be included at a l l because of her unique case. When 101 MS(7) i a tested without i t , against MS(6) in RET, P f a l l s to 2.57. from 25%. A l l other comparisons remain the same. From the averages and S.D.s in Table 18, i t i s easy to see that between the two diseases, there i s no significant difference between D(l)-D(2), MSU)-MS<2) and MS(3)-MS<4) classifications when comparing within or between these groups. D(3) i s , of course, very much greater then a l l other groups. The only set which i s "internally" s i g n i f i - cantly distinct i s the MS(5)-MS(6)-MS'7) set. Hence, i t i s tested against D(i) and D(2) only. MS(5) MS (6) MS<7) R D(l) > 25 10 > 25 E T D(2) > 25 25 > 25 C D<1) > 25 25 > 25 R P D(2) 25 > 25 > 25 Table 21. Significance level (%) to reject H" between diabetic and MS states. There i s no significant difference (at P=5%) between DU), D(2) and a l l MS groups. If these treatments and results are correct, there i s no significant difference between an MS subject's PR3 and that of a diabetic with n i l to moderate DRP, or a normal. However, individual variations in PR3 values (Table 17) should be noted. The results are similar when the CR bolus corrections are applied, with and without the correcting peak-to-peak ratio. (See 1 0 2 Section 2.3.) It i s hence not possible to study the effects of applying these corrections. The results thus far indicate the problems that are inherent in the VF technique. Assuming no instrumental errors, the d i f f i c u l t i e s in the positioning of the retina and the alignment of profiles in analysis reduce the certainty of the PR3 results. The above results and tests show that MS(1) and MS(2), and MS(3) and MS(4) cannot be differentiated. This i s due to the large variations within each group. It does not mean that the VF technique i s not applicable to MS as the subjects in MS(6), with active peri- phlebitis, were discernible from others of the MS<5> and MS(7) groups as was the case for diabetes where the severity of leakage also corresponded to the severity of DRP. Comparing the results for normals in Table 17 to other investi- gators' results, which are from 3.5 to 5.3 [32], i t i s seen that the PR3 values calculated here are within or above this range. No compa- risons are available for MS PR3. However, abnormal leakage was seen in some MS subjects other than the two with active periphlebitis. (Refer to Table 17.) These elevated PR3 values cannot be explained by retinal vasculature appearance (photographs and c l i n i c a l examina- tion). They imply that the VF system may be useful as a sensitive technique to detect subclinical activity. However, the present study only included a small sample of subjects and i s not able to relate the c l i n i c a l gradings of activity of MS or the current activity at the time of the procedure. 103 5 . 1 0 L U N D . B A S R e s u l t s LUND.BAS was not tested on a l l subject. Fir s t l y , the gradient- expansion algorithm i s slow (on this computer), and i s dependent on the i n i t i a l i z i n g estimates of P and D. Convergence i s slow i f any of the input values are far from the "true" values. (Refer to Section 3.4 and [33,343.) Secondly, the conditions set to halt calculations when the reduced chi-square value begins to diverge from a minimum, or i s less than 1, are not amply stringent in terms of convergence to a f i n a l solution set. The programme outputs the residues of the f i n a l , best f i t but plot outputs to visually check the answers are not (yet) available. Several outputs were returned on this subprogramme for three subjects tested. They were the 27- and 38-year old normals, and the 38-year old MS subject in relapse. (See Table 17.) Table 22 shows the case-by-case results. The units of P are •10-• cm.s-1. D remains in units of »10-* cmas-'. Only data points that were between 1.5 mm from the retina and the mid-vitreous were accepted for curve-fitting to Eq. 18 (Section 2.4). The number of data-points that was accepted and f i t t e d in each case i s shown in the extreme right column. The results of P values from other investi- gators for normals and diabetic subjects are also included. In comparison with published results, the ones obtained in this study are just within the range or less than those in the references. The diffusion constants, D, also follow the same trends when compared to those calculated in the previous section. Again, the large S.D.s of the quoted results should be noted. 104 INITIAL ESTIMATES FINAL FITS CHI-SQ. # P D P D 27-year 10.0 7.0 old NORMAL: 6.8 2.0 60.RET 7.9 20.8 2.7 1.6 4.78 3.62 15 15 27-year 10.0 old NORMAL: 6.6 60.CRP 2.8 5.9 1.13 16 27-year 10. 0 old NORMAL: 6.6 68.RET 4.0 5.0 1.44 14 38-year 0.5 old NORMAL: 6.6 64.CRP 5.3 5.1 0.70 20 38-year 10.0 old MS, F246 10.0 : 61.CV2 8.1 14.4 0.63 38-year 10.0 old MS, F246 10.0 : 61.CV4 7.7 16. 0 0.72 • • References: NORMAL P a) 11.0 •/- 4. -values 0 by Lund-Andersen , et al [54] b) 30.0 +/- 8.3 by Ogura , et al [53] c) 7.2 •/- 4. 4 by Zeimer, et al [29] d) 19.1 +/- 9. 4 by Chahal, et al [23] References: DIABETIC P-values 71.0 •/- 38.0 by Lund-Andersen , et al [54] Table 22. Results of LUND.BAS. »* means unavailable. If the algorithm of LUND.BAS i s to be the adopted method by which different investigators compare P and D values, i t i s apparent . that the computational conditions on the reduced chi-square in LUND . BAS must be more restrictive and selective. Double-checking with output plots must also be implemented. 105 5 . 1 1 O - t h e r P a r a m e t e r s P1, the permeability index was also calculated by C/VAZ.BAS. (Refer to Section 3.4.) The results were, however, not useful. They often turned out negative and were rejected. This failure was probably due to the algorithm i t s e l f . The necessity to integrate very close to the retina or to find an approximation when integrating in that region was machine-(AR-)dependent [32]. Another penetration ratio not metioned thus far i s that of the BAB. This result was not investigated because the source of leakage was the i r i s and c i l i a r y body. The diamond does not scan close to the source of leakage and the models used in the algorithms employed here are not applicable. Another point i s that misalignment errors are greater farther away from the retina. (See Figure 5.) Hence, a "PR3" cannot be calculated for the BAB. These conclusions were borne out by SLOPES. BAS which includes such a calculation at 3 mm from the pos- terior surface of the lens. No correlations in any of the groups were found. Two other performance parameters were calculated from subject data. The f i r s t was the in vivo reproducibility (R). The definition in Eq. 27 in Section 2.5 was changed as this parameter was calculated from two sources. One way to test R i s to take scans within 3 minutes of one another; the averages about certain regions of each profile are found. R i s then defined as: R = 100 « I a(x,t») - a<x,ts) I / J a(x,t») + a<x,t") } %, where a(x,t') i s the average about x of the t 1-minute p. i . profile. R 106 i s simply half the deviation from the average divided by the average. The alternative i s to use the S.D. in the numerator, but as there are only tvo entries at each calculation, this vas thought to be unneces- sary. Note that small R-values imply good reproducibility. The region about the mid-vitreous of measurement scans taken at t>55 minutes p.i. vas selected. The 3-mm interval vas not used because of the influence of tailings or large dye concentrations at later p . i . times, vhen there i s leakage. Further, no mixing vas expected at the mid-vitreous at these times. Only CRP-aligned profiles vere considered oving to the problems in locating the retina surface. Hovever, the f i n a l R vas the average from a l l scans made after 55 minutes p.i. for vhich another scan vas made vithin 3 minutes from the f i r s t . R i s a systems characteristic, and no tests vere made against disease classifications, etc. Taking a l l the above into consideration, the reproducibility vas estimated in averaging over 25 cases, to be R = 19.0 12.7 % . The other method of estimating R i s to replace a(x,t) in the above formula vith the calculated values of D or PR3, or any calcu- lated characteristics. The conditions that the numbers must be from t > 55 minutes p.i. and vithin 3 minutes of one another s t i l l hold. The number found by replacing vith PR3 by CRP vas R = 15.8 14.2 7. . The last parameter considered vas the axial resolution (AR) alluded to in Section 2.1. It vas here defined as the ratio of the concentrations at 3mm to the CR peak of the bolus scan after the pre- 107 injection scan had been subtracted. This definition allowed only CRP- aligned profiles to be used. Also, diseased eyes were excluded owing to possible bolus effects. Note that not a l l bolus scans were made at exactly 3 minutes p.i . Depending on the subjects, bolus scans were made between 2 and 7 minutes p. i . Averaged over the three normals, AR = 0.032 •/- 0.026 . 108 V I . C Q N C L U S I O N The performance of the assembled vitreous fluorophotometer was in close agreement with the data that describe the type of light detection system used in this study. Hardware and software designed to interface the light detection system with a microcomputer provided the signal conversion, data analysis and storage capabilities. The inherent limitation of the optical system was the depen- dence on a plano-concave contact lens for scanning the vitreous chamber. Discomfort caused by this lens from sequential measurements was largely overcome by using a combination of a bandage soft contact lens and a plastic Luma" lens in place of the glass Goldmann lens used by other investigators. The effect of substituting the lenses was a different set of F-numbers which were re-calculated by measu- ring intraocular distances using ultrasound. The most outstanding problem was locating the same reference points along each scan for the purpose of alignment, reduction of data and subtraction in the subsequent analysis. The method used to lessen the effects of alignment errors was to average about an inter- val along the profile instead of selecting a specific point for comparison. The effectiveness of this algorithm however could not be determined because precise in vivo data on vitreous concentration could not be obtained. The correction algorithm for the choriod-retinal bolus t a i l i n g effects used in the Fluorotron" Master did not improve the separation between groups when applied to the fluorophotometer assembled for 109 this study. The penetration ratios in the 15 diabetic subjects were found to increase progressively with the severity of retinopathy (3 gra- dings), in agreement with published reports. However, there was significant dispersion of results about the averages in each group. In the sample of 16 multiple sclerosis subjects, the penetra- tion ratios were not significantly different between the two groups that represented standard c l i n i c a l activity categories or the two groups that represented current activity categories. The usefulness of vitreous fluorophotometry as a non-invasive test for monitoring central nervous system activity could not be ascertained because of the small sample sizes. Abnormal leakage was found in 4 of 15 cases with normal vitreous and either minimal or no evidence of retinal periphlebitis activity (2). The penetration ratios in active peri- phlebitis were elevated (3-4 times normal control). Abnormal leakage in the absence of active periphlebitis has not previously been recorded. An elevated penetration ratio vas also found in one case with vitreous liquefaction. Almost a l l subjects in the diabetes sample, irrespective of retinopathy severity, showed vitreous diffusion constants s i g n i f i - cantly less than the diffusion constant of sodium fluorescein in water. In the multiple sclerosis sample and controls, the diffusion constants in the vitreous and in water were not significantly d i f f e - rent. The diffusion constant in the vitreous was 2-4 times greater than the value for water in 4 older multiple sclerosis subjects and 5 times greater in the case with vitreous liquefaction. 110 A P P E N D I X A C O M P U T E R P R O G R A M M E S A.1 DAS.PRN This prograaae, DAS configures the PIA i n a triggering node of the Data Acquisition System. The control lines are needed for INTR, WR, MUX, svitch A and S/H. RD and CS are held lov, i.e. the output of the ADC i s alvays enabled. This routine, aa entered i n the HBASIC DATA statements, i s separated by a seni-colon and a nunber as seen belov. E.g. - ;#20 This version vas v r i t t e n i n July 1985 by PANG Kian Tiong. PA0-7 to be inputs. - l i n e #1-8 CA1 * not used. - l i n e 419 CA2 = not used. - l i n e #17 ••" » CRA vord * 0010 1113 PB8 PB1 PB2 PB3 PB4 PB5 PBS 1 to set PA0-7 as inputs. 1 for PB6, 7 to be inputs. 9 for PBS, 5 to be outputs. ? - WR control: - l i n e 49 0 « start conversion reset and v a i t DUX address: - l i n e #21 s pod * radiometer S/H address: - l i n e #11 » saitple ' hold Svitch A interrupt input 3 continue reading vait to s t a r t or stop 1 1 l i n e #15 PB7 * ? - INTR interrupt input - l i n e #13 0 » INTR h i 1 » INTR l o • • " » DRB vord: 00?? _?011 EF08 » 1 rlagpos equ 0ef08h ;raa/roa junp vector 2901 » era equ 02901h ;control r e g i s t e r A 2903 » crb equ 02903h ;control r e g i s t e r 3 2900 * dra equ 02900h ;data/direction r e g i s t e r A 2902 drb equ 02902h ;data/direction register 3 EFDE * status equ 0efdeh ;PIA status D1D6 org 0dld6h ;starti n g address Configuring * t the PIA begins. D1DS AF « •ode: xra a D1D7 3ADEEF Ida status D1DA FE02 cpi 2 D1DC ca rz ; i f already i n input mode ;#2 D1DD 3E2A • v i a,02ah ;0010_1010 D1DF 320129 sta era D1E2 3E00 nvi a.0 #3 D1E4 320029 sta dra D1E7 3E2E • v i a. 02eh ;0010.1110 D1E9 320129 sta era ;port A i s input #4 D1EC 3E00 nvi a.0 D1EE 320329 sta crb D1F1 3E3F • v i a, 03fh ;0011_U11 ;#5 D1F3 320229 sta drb D1FS 3E3S • v i a, 03Sh ;0011_0U0 D1F8 320329 sta crb ;port B i s set ;#6 111 D1FB 3EB2 D1FD 32DEEF D2B0 C9 mvi ata ret a, 2 atatua 0231 3A0229 D234 17 D235 D231D2 0238 3A3229 D23B 17 D23C DA38D2 D23F 3A3329 0212 2F D213 77 0214 3E2B D21S 323229 0219 C9 ; Reading and interrupt statua: INTR ;*7 input: Ida drb r a l ;*a high: ;#9 ;#13 Jnc Ida r a l Jc Ida cna sov avi ata ret ; i f INTR h i i i i INTR lo input drb high dra fl.a a,32bh ;a«13.1311 drb D21A F3 D21B 0330 021D 3E03 D21F 32B8EF 0222 32D3D1 0225 CDD6D1 0228 3A3229 D22B 17 D22C 17 0220 DA2302 D233 C33BD2 0233 F3 0234 D3B9 0236 3EB3 0233 3238EF D23B 3E1B D23D 323229 0243 21D401 D243 3E33 D245 323229 0248 CD31D2 D24B 21D2D1 D24E 3E23 0253 323229 0253 CD31D2 0256 3A3229 D259 17 D25A 17 D25B D263D2 D25E 3EF3 0263 320301 0263 D331 0265 3E31 0267 3238EF 026A FB 0263 C9 j Fi r a t entry point from BASIC programme. ;#U entry: di out mvi ata ;#12 ata c a l l ;#13 atart: Ida r a l r a l 3 a, 3 flagpos ;in Bank 2 3dld3h jstop code mode ;teat PIA atatua ;#14 start read ;poll Svitch A ; Subaequent entries from BASIC routine. ;#15 di out mvi ata ;#16 read: mvi eta a, 3 flagpoa ;in Bank 2 a,31bh drb ;3331 1311 Reading radiometer #17 Ixi h,3dld4h mvi ;#18 sta drb c a l l input a,333h ;3811_3B11 Reading pod. #19 l x i mvi ;#23 ata c a l l h,3dld2h a,fl23h ;0313_3311 drb input : Check for atop acquisition ;#21 Ida drb r a l r a l jnc ;#22 mvi ata ;#23 goback: mvi sta e i ret goback a r3f3h ;stop code 3dld3h out 1 a, 1 flagpoa ;back i n Bank 1 ;to BASIC routine 112 A.2 SCANMENU.BAS 133 WIDTH 52 : CLEAR. 4HD1CF : RE.1 Updated 230186 110 OEFSTR A 123 DEFINT I-N 130 A=" : PRINT A : PRIHT "SCAN MENU" : PRINT A : PRINT 143 RESET : PRINT "(0) Run VITREOUS SCANNING programme" : PRINT 150 PRINT "(1> Run PLASMA SCANNING programme" : PRINT 163 PRINT "(2) Run SUBJECT DATA ENTRY programme" : PRINT 170 PRINT : PRINT "Enter ANY OTHER number to EXIT." : PRINT 183 PRINT "Union programme do you »i3h to run *; : INPUT II 190 IF (II>2) THEN END ELSE PRINT 233 IF (11 = 2) GOTO 523 210 PRINT "Loading assembly-language subroutine, DAS. " 223 DATA 4HAF, 1H3A.4HDE.4HEF,4HFE.2, 4HC8 233 DATA 4H3E.lH2A.iH32.1..4H29, AH3E, 3 240 DATA 4H32.3,4H29.AH3E. 4H2E. SH32. i . 1H29 253 DATA 4H3E.3.4K32.3.4K29, SH3E. 4H3F 263 DATA 4H32. 2.SH29.SH3E.AH36.4H32.3,4H29 270 DATA 4H3E.2.AH32.4HDE.1HEF, 4HC9 283 DATA 4H3A.2,4H29.4H17, 4HD2. 1, 4HD2 293 DATA 4H3A,2.4H29,4H17, 4HDA, 3, 4KD2 333 DATA 4H3A.3,4H29,4H2F,4H77 313 DATA 4K3E. 1H23,4H32.2,4H29, 4HC9 323 DATA 4HF3. AHD3.3, AH3E. 3, 4H32.3, AHEF 333 DATA 1H32,4HD3,1HD1.4HCD. AHD6,AHD1 340 DATA 4H3A.2,4K29,4H17.1H17, 4HDA, 4H28. AHD2 353 DATA 4HC3.AH3B. 4HD2 363 DATA 4HF3.4HD3,3.4H3E. 3. 4H32, 3, 4HEF 373 DATA 4H3E.4H13.4H32, 2, 4H2S 330 DATA 4H21.4HD4,4HD1.4H3E. 4H33 393 DATA 4H32,2,4H29,4HCD, 1, 4HD2 430 DATA 4K21,4HD2.4KD1,4H3E. 1K23 413 DATA 4H32.2.4H29,4HCD,1, 4HD2 420 DATA 4H3A.2.4H29. 4H17. 4H17, 4HD2, 4H63. 4HD2 433 DATA 4H3E,4HF3. 4H32.4HD3. 4HD1 443 DATA 4HD3, 1, 4H3E, 1. 4H32, 3. 4HEF, 4HF3, 4HC9, 3, 3, 3 453 FOR 1=1 TO 151 463 READ J : K=4HD1D5»I : POKE K, J 473 NEXT I 480 PRINT : PRINT •=> DAS has been auccesafully loaded!" : PRINT 493 IF (11 = 1) GOTO 510 500 CHAIN MERGE "B:VITSCAN". 1013, ALL . 510 CHAIN MERGE "B:PLASCAN\ 1313. ALL 523 DIM AA(22),X(3) 533 A="-- : PRIHT : PRINT A 543 PRINT 'SUBJECT DATA" : PRINT A : PRINT S53 PRINT "Enter the folloving — > • : PRINT 563 LINE INPUT "Subject's NAME -> ";AA(01 : PRINT 573 LINE INPUT "Subject's AGE --> *;AA <1) : PRINT 580 LINE INPUT ' Scan EYE --> *;AA(2) : PRINT 590 LINE INPUT • Sca,n DATE -> ":AA(3) : PRINT : PRINT 603 FOR 1=3 TO 5 STEP 3 613 PRINT 'Enter the "; 620 IF (I<>0) GOTO 643 633 PRINT 'LENGTHS noted (in Osb/DAS units) —->• : GOTO 663 643 PRINT "ULTRA-SOUND scan results (in mm) --->• : PRINT 653 PRINT TAB(13);'If NO ultra-sound taken, enter 3." 663 PRINT : PRINT "VITREOUS ---> "; : INPUT X(I) 673 PRINT : PRINT • LENS ---> •; : INPUT X(I-l) 683 PRINT : PRINT • AOUEGUS ---> "; : INPUT XU-2) 693 PRINT : PRINT : PRINT : PRINT 733 NEXT I 710 PRINT 'Amount of FLUORESCEIN injected was •; 723 INPUT X(6) : PRINT : PRINT : PRINT : PRINT 733 PRINT "Enter any REMARKS, COMMENTS or OBSERVATIONS -->' : PRINT 743 PRINT TAB(10);"Press RETURN to exi t . " : PRINT : 1=3 753 LINE INPUT "Enter -> ';A 763 IF <A=") GOTO 783 773 1=1*1 : AA(I1=A : GOTO 753 783 PRINT : PRINT : PRINT 'The above i s to be f i l e d i n " 793 PRINT : PRINT TAB!13);"A for the LEFT Drive' 833 PRINT TAB(13);'B for the RIGHT Drive "; 813 A=INPUTSU) : PRINT : PRINT : KEY = 0 820 IF <A="A" OR A="a" OR A="3' OR A="b') THEN KEY=1 833 IF (KEY=0> GOTO 980 340 IF A=-'b" THEN A="B" 853 IF A="a" THEN A="A" 113 860 OPEN "a\U,A'":SU3JECT.DAT" : J=0 : !f = 3 : N=3 370 FOR K = J TO ,1 383 IF N=l THEN PRINT ELSE PRINT *1,AA(K) 390 NEXT K 900 N=N-1 913 ON N GOTO 923.930.940 920 M=6 : GOTO 370 933 J=4 : »=I : GOTO 373 943 PRINT : PRINT 'SUBJECT.DA? baa been completed.* 953 ERASE AA.X 960 PRINT : PRINT : PRINT : PRINT 'Da NOT forget to COPY onto a run Diskette before BATCHRUN !!" 970 PRINT : PRINT : GOTO 130 983 PRINT : PRINT : PRINT 'BAD Entry !!! Please try again.* 993 PRINT : PRINT : GOTO 733 A . 3 V I T S C A N . B A S 1030 REM Updated 343286 1013 DIN IXU633), IY( 1603), NL( 9) 1020 A="*»»-'">**->-*"»'-*" : PRINT : PRINT : PRINT A 1030 PRINT "VITREOUS SCANNING' : PRINT A : PRINT : PRINT 1040 PRINT "SYSTEMS CHECKS" : PRINT 1050 PRINT '(1) A l l equipment ON.' : PRINT '(2) A l l cables connected." 1060 PRINT '(3) S»itch A LOW. • : PRINT '(4) VOLTMETER/Intensity detector ON." 1073 PRINT '(5) Intensity = 141 •/- 1" : PRINT '(6) HV = 739" 1080 PRINT '(7) R/H exp = 0" : PRINT '<8) R/H-DAS output » 3' 1090 PRINT '(9) Switch A HIGH to begin." : PRINT 1100 K=&HD21A : IP=1HD233 : PRINT : PRINT "LANDMARK SCANNING" 1110 PRINT : CALL K : A=INKEYS : J«=-l 1120 CALL IP : PRINT "Intensity =•;PEEKf1HD1D4); 1133 K=PEEX1&HD1D2> : PRINT TABI30);"Pod =';K 1140 IF (INKEYS*'") GOTO 1163 1150 JM=jn*l : 1XIJH>=K : PRINT CHRSK7); 1163 IF (PEEK(4HD1D3)<>243) GOTO 1120 1170 IF J(1<1 THEN GOTO 1250 ELSE PRINT 1183 IF JH>8 THEN J(l = 8 1190 FOR K«I TO JH 1203 I = K-1 : HL(I)»IXCK)-IXU> 1213 PRINT TAB<18>;'Hark';K;'- Nark';I;'*•;BL(I) 1220 NEXT K 1230 PRINT : PRINT "Total LENGTH '';IX(JN)-IX(3) : PRINT 1243 PRINT "IMPORTANT: WRITE do»n these numberB.' : PRINT 1253 PRINT : PRINT "LANDHARKing »ill NOT be Repeated, OK?" : GOSUB 1738 1263 IF <A»CHRS<27>) GOTO 1100 1270 A*1NKEY9 : PRINT : PRINT "Which EYE i a to be scanned?" > PRINT 1280 PRINT ' Enter : ESC for the RIGHT eye, • 1290 PRINT TAB(16)j"ANY for the LEFT eye. "; : A=INPUTS(1) : PRINT : PRINT 1300 IF A=CHRS<27> THEN JJ=1 ELSE JJ=0 1310 IF JJ=0 THEN PRINT 'LEFT"; ELSE PRINT "RIGHT"; 1320 PRINT " Eye . i l l be scanned.• : PRINT : PRINT 1330 PRINT "SCANNING INSTRUCTIONS" : PRINT " = = = = = » = = = = .. = " = » = = = = =• 1343 PRINT '(1) Scan STEADILY (from the RETINA each time)." : PRINT 1350 PRINT "(2) Wait for the BEEP before s t a r t i n g . " : PRINT 1360 PRINT "(3) S»itch A to HIGH to BEGIN scanning." : PRINT 1370 K=&HD21A : IP=4HD233 : CALL X 1380 FOR 1 = 0 TO 123 1390 CALL IP : IM=PEEK < &HD1D2) 1400 IF K50 THEN PRINT "Pod *";IN, "Retina =" ;PEEK ( &HD1D4) 1413 IF 1=50 THEN PRINT CHRS<26> 1423 NEXT I 1430 I=-l : PRINT CHRS(7)j 1440 IF 1-1599 THEN GOTO 1470 ELSE I = I»1 1450 CALL IP : IX(I)= P£EK< &HD1D2) : IY(I)=PEEK(SHD1D4) 1460 IF (PEEK<&HD1D8><>248> GOTO 1440 ELSE GOTO 1488 1473 PRINT "You are out of MEMORY. S»itch A to LOW.• t PRINT 1488 N-I : GOSUB 1778 : PRINT : PRINT 1498 IF (KO<>8) GOTO 1598 1530 PRINT " l a the scan to be SAVEd?" : GOSUB 1738 : PRINT 114 1510 IF A=CHR9(27> THEN PRINT "NO I" E L S E PRINT "YES I" 1520 PRINT : PRINT 'Please CONFIRM vith the SAME • : GOSUB 1730 : PRINT 1530 IF A=CHRS(27> THEN GOTO 1590 ELSE PRINT 1540 PRINT "Enter the P.I. Scan TIJ1E" : PRINT TAB(20);"for the •; 1550 IF JJ = 0 THEN PRINT "LEFT"; ELSE PRINT "RIGHT"; 1560 LINE INPUT " eye. ";A : PRINT : PRINT 1S70 NAHE AA*"TEH.DAT* AS AA*A*".DAT" : PRINT 15B0 PRINT "===>> The f i l e has been successfully RENAHEd. " 1590 PRINT : PRINT "Is SCANNING to be CONTINUEd?" : GOSUB 1730 1600 IF (A=CHRS<27)) GOTO 1630 1610 PRINT "SYSTEMS .111 NOT be checked, OK?" : GOSUB 1730 1620 IF <A«CHRS<27>> GOTO 1270 E L S E GOTO 1040 1630 PRINT : PRINT : PRINT "The f l l e a in Drive A are :" 1640 PRINT : FILES • A: *.• • : PRINT : PRINT 1650 PRINT "The f i l e s in Drive B are : • : PRINT 1660 FILES " 8 : : PRINT : PRINT 1670 PRIHT "Do you vish to return to the SCAN MENU?" : GOSUB 1730 1680 IF <A=CHR3<27)> GOTO 1590 1690 ERASE IX,IY,ML 1700 ON ERROR GOTO 0 1710 CLOSE : GOTO 130 1720 REM Prompt options display. 1730 A=INKEYS : PRINT : PRINT TABI5);"Enter : ANY KEY f o r YES" 1740 PRINT TAB<17)j"ESC for NO "; : A=INPUTS<1> 1750 PRINT : PRINT : PRINT : RETURN 1760 REM Subroutine for video-plotting. 1770 WIDTH 127 1780 K=155 : W=0 : O=20!/<K-W> : IR=IX(N) : LL=IX<0>-10 : Kg=0 1790 IF LL<0 THEN LL=0 1800 IF (IR-LLX50 THEN U=ll ELSE U=1251 / (IR-LL) 1810 PRINT CHRS(26);CHR9(27)•")•;CHRS(27)•" = •-CHRS(54)«CHRS(36) ; 1820 PRINT " Press ";CHRS(27)••(";"ANY";CHRS<27)•">"; 1830 PRINT " KEY to ABORT this SCAN at any time.• : A=INKEY9 1B40 FOR 1=0 TO 20 1850 PRINT CHR9<27).'=".CHRS<32.I>>CHRS<32>; 1860 IF (1=0 OR 1=5 OR 1=10 OR 1=15 OR 1=20) THEN PRINT "«•; ELSE PRINT " I " ; 1870 NEXT I 1880 FOR 1=0 TO 12S 1890 PRINT CHRS(27>."=-"<-CHRS(53)-CHRS(33.I) ; 1900 IF (IHKEYSo"" ) GOTO 2180 1910 IF RIGHTSISTRSII),1>="0" THEN PRINT ••»•; ELSE PRINT "-"; 1920 NEXT I 1930 IF JJ = 0 THEN AA= "A: " ELSE AA="B:" 1940 OPEN "0",I3, AA*"TEH.DAT" : PRINT 13, IH : A=INKEY9 1950 PRIHT CHRS(27)•" = "*CHRS(54 > *CHRS(32) ;LL; : H»0 1960 PRINT CHR9(27)»"-"»CHSS(54).CHR9(150)jIR; 1970 FOR 1=0 TO N 1980 IF I I H K E Y 3 0 " " ) GOTO 2210 1990 IF (IY(I) = 128 OR IY(I) = 127) GOTO 2050 2000 IP=CINT((IX(I)-LL).U) : K=CINT( (IY(I)-W).0) 2010 PRINT »3,IXtl) : PRINT »3,IY<I> : K-H'l 2020 IF (1P>124 OR K>20) GOTO 2050 2030 IF (IP<0 OR K<0) GOTO 2050 2040 PRINT CHR9(27)."»".CHR9(52-K).CHR9(33»IP);"." 2050 NEXT I 2060 IP=IM-LL : K»CINT(IP"U) : PRINT CHR9(27)••("; 2070 PRINT CHRS<27)."="«CHRS<34>>CHRS(35>iM;"pairs were entered." 2080 PRINT CHRS ( 27) • " = " +CHR3 (53) *CHRS(33*K) ; "R" ; 2090 IF (JH<0> GOTO 2150 2100 FOR 1=0 TO JM 2110 I P = I P - H L U ) : K=CINT(IP«U) 2120 IF <K<0 OR K>124) GOTO 2140 2130 PRINT CHRS(27)."=".CHR9(53)«CHR9(33»K);"!"; 2140 NEXT I 2150 A=INKEYS : PRINT CHRS(27)•"="*CHRS(54)*CHRS(40)f 2160 PRINT "Press ANY key to return to PROMPT mode."; : A=INPUT9(1) 2170 CLOSE 2180 PRINT : PRINT : PRIHT CHRS<27>«"<" : PRINT 2190 WIDTH 52 2200 RETURN 2210 PRINT CHR9(27)."=".CHR9(35)'CHRS(35); 2220 PRINT "Please CONFIRM that you vant th i s scan ABORTEDl I " 2230 GOSUB 1730 : PRINT : ATT=INKEY9 : ATT=INKEY9 2240 IF <A=CHR9(27)> GOTO 1990 2250 K0=1 : GOTO 2170 115 A . 4 P L A S C A N . B A S 1000 REM Updated 180186 1010 DIM IXI1200). IY(1200),L(256), V( 256), S( 256), X(55), Y(55),Z(55) 1020 A="«->*»'<-»**»»»**»' : PRINT : PRINT A : KEY=-1 1030 PRINT 'PLASMA SCANNING" : PRINT A : PRINT : PRINT 1040 PRINT : PRINT 'SCANNING INSTRUCTIONS" : PRINT '===.===«»=..«==...' 1050 PRINT "(1) Aim PROBE with WHITE l i g h t . " ! PRINT 1060 PRINT "(2) Adjust c e l l l o r NO REFLECTION.' : PRINT 1070 PRINT "131 Move PROBE un t i l Just behind glass surface.' s PRINT 1080 PRINT "(4) Fix the POD position." : PRINT 1090 PRINT '(5) S«itch A to HIGH to BEGIN scanning.' 1100 K=IHD21A : IP=1HD233 : CALL X : PRINT CHRSI26); 1110 FOR 1=0 TO 99 1120 CALL IP : K=PEEK(&HD1D2) : K=PEEK(&HD1D4) 1130 NEXT I 1140 I=-l : PRINT CHRS(7); 1150 IF 1 = 1199 THEN GOTO 1180 ELSE I-I»l 1160 CALL IP : IX (I) =PEEK ( &HD1D2) s IY (I) =PEEK (&HD1D4) 1170 IF <PEEK<SHD1D0><>240) GOTO 1150 ELSE GOTO 1190 1180 PRINT "You are out of MEMORY. Svitch A to LOW. " : PRINT 1190 NN=I : PRINT "AVERAGING begins." : PRINT : PRINT 1200 PRINT TAB (5); "Press ANY key to interrupt." : PRINT : PRINT 1210 FOR J=0 TO 255 1220 L(J)=0 : V(J)=0l : S(J)=0l : A=INKEYS 1230 NEXT J 1240 FOR J=0 TO NN 1250 I=IX(J) : HM=IY(J) 1260 IF (INKEYSo"") GOTO 1450 1270 IF (MM=127 OR MM=128) GOTO 1290 1280 L(I)=L(I)«1 : V(I)=V(I).MM : S(I)=S(I)*MH*2 1290 NEXT J 1300 J=-l : JM=0 : KS=0 1310 FOR 1=0 TO 255 1320 MM=L(I) : KS=XS*MM 1330 IF (INKEYSo"") GOTO 1450 1340 IF (HM<2) GOTO 1390 1350 J = J»1 : V!J)=V(I)/MH : L(J)=I 1360 S<J)=<S(I)-MH«V(J)A2)/<HH-1> 1370 IF (MM<JM> GOTO 1390 1380 JH=MH : M=J 1390 NEXT I 1400 KEY=KEY*1 : Y<KEY)=V(M) ! Z(KEY)=SOR<S<H>> 1410 PRINT : PRINT "AverBge ••;V(M)j•«•/-•;Z<KEY) 1 PRINT 1420 PRINT TAB(10);"for";JM;"out of» ;KS s "points. • : PRINT 1430 PRINT : PRINT 'What i s the Sample TIME (in min. P.I.) "; 1440 INPUT X(KEY) : PRINT : PRINT : GOTO 1460 1450 PRINT TAB(5);"Averaging interrupted II* : PRINT : PRINT 1460 PRINT : PRINT 'Is SCANNING to be CONTINUEd?' : GOSUB 1720 1470 IF <A<>CHRS<27)> GOTO 1040 1480 PRINT : PRINT "Was the scan for (0) BOTH eyes 7" 1490 PRINT TABC18>;"(1> LEFT only 7* 1 PRIHT TAB(18>;'<2) RIGHT only •; 1500 INPUT KM : PRINT : PRIHT 1510 IF KM=2 THEN J=l ELSE J=0 1520 A= *:PLASMA. DAT* : PRINT "The PLASMA data are:" : PRINT 1530 IF J=0 THEN OPEN '0",#2, "A'-A ELSE OPEN "0",»2, "B'-A 1540 FOR 1=0 TO KEY 1550 PRINT #2,XII) : PRINT <2,Y(I) : PRINT #2,Z(I) 1560 PRINT I + l;') ';X(D, Yd) : ' •/- ';2(D 1570 NEXT I 1580 CLOSE : J=J-1 : PRINT : PRINT 1590 IF (KM=0 AND J=l) GOTO 1530 1600 PRINT "Plasma averages have been f i l e d . • : PRINT 1610 PRIHT 'The f i l e s in Drive A are : * 1 PRIHT : FILES 'A:«.»' 1620 PRINT : PRINT : PRINT 'The f i l e s i n Drive B are : • 1630 PRIHT : FILES 'B: •. • • : PRINT : PRINT : PRINT 1640 PRINT "Do you have another set of samples to da?" : GOSUB 1720 1650 IF A=CHRS<27> THEN GOTO 1690 ELSE RESET 1660 PRINT : PRINT "Enter ANY key after changing LOGGED diskette.• 1670 A=INPUTS(1) : RESET : PRINT : PRINT 1680 GOTO 1020 1690 ERASE IX, IY.L, V,S,X,Y,Z 1700 PRINT : PRINT : PRINT : PRINT 1710 CLOSE : GOTO 130 1720 A'INKEYS : PRIHT : PRINT TABI5)! "Enter : ANY KEY f o r YES" 1730 PRINT TABI17)j*ESC for NO ' j : A'INPUTSU) : PRINT : PRINT ! PRINT 1740 RETURN 116 A . 5 B A T C H R U N . B A S 100 WIDTH 52 ; RED Updated 120386 110 DEFINT I-N 120 DEFSTR A 130 DIN AN(13>, ITASKCU) 140 GOTO 170 150 GOTO 680 160 RESET 170 PRINT CHRSI26);"Are you using a COPY of the o r i g i n a l data?" 180 A='-»*»».»-«-»»«->'" : PRINT : PRINT A 198 PRINT "PROGRAMME MENU" : PRINT A : PRINT 200 PRINT "<0> Run REDUCE - Ra» data averaging" 210 PRINT "(1) Run B/G - Background averaging" 220 PRINT "(2) Run MINUS - Background subtraction" 230 PRINT "(3) Run SUDATA - Reprod. /LLoD/AxRes" 240 PRINT "(4) Run BLOOD - Plasma data integration" 250 PRINT "(5) Run C/VA2 - Cunha-Vaz's algorithm" 260 PRINT *<6) Run SLOPES - Curve-fitting method" 270 PRINT "(7) Run PLOT - Coarse p l o t t i n g " 280 PRINT "(8) Run DRAW - Super-impose pl o t t i n g " 290 PRINT "(9) Run LUND - Lund-Andersen's algorithm" 300 FOR 1=0 TO 10 310 ITASK(I)=-99 320 NEXT I 330 PRINT : PRINT "Enter the SEOUENCE to run ->" : PRINT 340 PRINT "Runs ALWAYS begins vith the l e f t to right. • : PRINT 353 PRINT TAB(16);"Yes = ANY key : No = ESC* 360 PRINT : JTASK=-99 : A=INKEY9 : A""" : NTASK=-99 370 PRINT "0 1 2 3 4 5 6 7 8 9* 380 FOR 1=0 TO 9 398 A=INPUTS(1) : PRINT * "; 400 IF A=CHR9<27) THEN GOTO 428 ELSE ITASKU>=55 413 IF I<7 THEN JTASK=5S ELSE NTASK=55 423 NEXT I 430 IF (JTASK>0 OR NTASK>0> GOTO 490 440 PRINT : PRINT : PRINT "NO task assigned!* : PRIHT 450 PRINT "Do you «ant to continue! <Y> *; 460 A=INPUT9(1) : PRINT : PRINT 470 IF (A="N* OR A="n") THEN LPRINT CHR9C27);CHR9(79)J ELSE GOTO 160 480 END 490 PRINT 503 FOR 1=0 TO 9 510 IF ITASK(I)>0 THEN A="Y* ELSE A="N" 520 PRINT A;* S30 NEXT I 540 PRINT : PRINT : PRINT TAB(15);"Please CONFIRM I (Y) *J : A=INPUT9(1) 550 IF (A="N* OR A="n") GOTO 170 560 IF (JTASK<0 AND NTASK=55> GOTO 640 570 PRINT : PRINT : PRINT " F i l e s in Drive B are:* 580 PRINT : FILES "B:*.«" : PRINT : PRIHT : PRINT 590 PRINT TAB!5);"Enter in CHRONOLOGICAL, ASCENDING order.* 608 PRINT : PRINT TAB<18);"Enter RETURN to e x i t . * : PRINT : HO--1 618 LINE INPUT 'Enter --> "jA 620 IF <A=""> GOTO 640 630 MQ=HO-l : AN< MQ) = A : GOTO 610 640 PRINT : PRINT "Check PRINTER/PAPER.* : PRINT 650 LPRINT CHRS< 27) ;CHRS (65) ; CHRS (12) ; CHRS ( 27) ;CHR9( 50) ; 660 LPRINT CHR9<18);CHR9(27);CHRS<73);CHR9(1); 670 LPRINT CHRS(27);CHR9(78);CHRS(6); : LTASK=0 680 NERR=0 : PSCALE'.095919 690 FOR KTASK=LTASK TO 9 700 IF ITASK ( KTASK) <3 THEN GOTO 820 ELSE LTASK=KTASK*1 710 ON KTASK GOTO 733,740,753,760,770,780,790,800,810,820 720 CHAIN MERGE * A:REDUCE*, 1010, ALL 730 CHAIN MERGE 'A:B/G", 1010, ALL 740 CHAIN MERGE "A:MINUS",1010,ALL 750 CHAIN MERGE "A:SUDATA",1010,ALL 760 CHAIN MERGE "A:BLOOD",1010,ALL 770 CHAIN MERGE "A:C/VA2", 1010,ALL 780 CHAIN MERGE "A:SLOPES", 1010, ALL 790 CHAIN MERGE "A:PLOT*, 1013, ALL 833 CHAIN MERGE "A:DRAW",1010,ALL 810 CHAIN MERGE *A:LUND", 1010, ALL 820 NEXT KTASK 830 PRINT s PRINT : GOTO 450 117 A . e R E D U C E . B A S 1000 REM Updated 210186 1010 A = " « : PRINT A : PRINT •REDUCE" : PRINT A 1020 DIN 1(256), IXI4000),IYI4000), IS<256), Y(256),WI256) 1030 DEF FNCQNC(X)=EXP(-109.339*SQR(11915.9»X/.114441)) 1040 ON ERROR GOTO 1790 1050 A="» 1060 FOR LT=0 TO HO 1070 FOR K=0 TO 255 1080 I(K>=0 : IS(K)=0 1090 NEXT K 1100 JC0=255 : JCR = 0 : OPEN "I",#1,"B:"-AN(LT)•\ DAT" i INPUT #1,N : K = - l 1110 IF EOFIDO0 THEN GOTO 1160 ELSE K = K-1 1120 INPUT #1,J : INPUT #1,IY(K) : K J X K J X l 1130 IF J>JCR THEN JCR=J 1140 IF J<JCO THEN JC0=J 1150 IXIKXJ : GOTO 1110 1160 CLOSE 11 : JVL=0 ; IHA=K ; NUN»-1 : JMV = 0 1170 FOR K=JCO TO JCR 1180 IF (IIXX2) GOTO 1530 1190 PRINT "For X =";K;", there are" ;KK); "points. " : PRINT 1200 IHAX=0 : IHIN=255 : J ' - l : JLE=I(K> : U=0i : V»0l : KP=0 1210 FOR KX=JHV TO IMA 1220 IF (KP=1 OR IX(KXX=K> GOTO 1240 1230 KP=-1 : IMY=KK 1240 IF I X(KKX> K THEN GOTO 1290 ELSE JAL=IY(KK> 1250 IF J=(JLE-1) THEN GOTO 1300 ELSE J=J-1 1260 IS(JAL) = IS(JALX1 : U=U'JAL : V=V'JALA2 1270 IF JAL>IMAX THEN IMAX=JAL 1280 IF JAL<IHIN THEN IMIN=JAL 1290 NEXT KK 1300 IF KP=t THEN JHV=IHV ELSE JMV'KK 1310 KK=I(1AX-II1IN : PRIHT "MAX =" i IMAX, "MIN = * ; IMIN, "MAX-NIN = ";KK 1320 IF XK<7 THEN GOTO 1530 ELSE JLE=0 1330 IF JVL=0 THEN P=U/(J*1> 1340 FOR KK=IMIN TO IMAX 1350 IF (ISIKKX0) GOTO 1430 1360 PRINT " For Y =•;KK;',";TAB(18))"there are";IS(KK>;"points. " 1370 IF (IS(KKXJLE) GOTO 1430 1380 IF IS(KK)>JLE THEN KJ=255 1390 ICO=ABS(CINT(KK-P)) 1400 IF (ICO>KJ) GOTO 1420 1410 KJ'ICO : IHV=KK 1420 JLE=IS(KK) 1430 NEXT K K 1440 KK'IHV : PRINT : PRINT "Estimate used i s " ; K K i V-0t t U-01 l JLE-9 1450 ICO=KK»5 : ICR=KK-5 : KK=CINT<KK-P> 1460 IF KK>5 THEH ICR=CINT(P) 1470 IF KK<-5 THEN ICO=CINT(P> 1480 FOR KK=IHIN TO IHAX 1490 IF <KK>ICO OR KK<ICR) GOTO 1510 1500 JLE=JLE.IS(KK) : V»V-IS(KK>»KKA2 i U=U*KK"IS(KK) 1510 NEXT KK 1520 PRINT : PRINT "Upper Bound *•;ICO, "Lover Bound =";ICR 1530 U=U/JLE : PRINT : PRINT TAB(10);"No. of points used -•;JLE;"of•;I(K) 1540 FOR KK=IHIN TO IMAX 1550 IS(KK)=0 1560 NEXT KK 1570 IF JLE<2 THEN GOTO 1630 ELSE V=(V-JLE«UA2>/(JLE-1> 1580 IF V=0l THEN V= 11 1590 IF V<0l THEN V=ABS(V) 1600 NUM=NUM'l : V=SQR(Y) : HNUMXK ; Y(NUH)=U : W(NUM)=V 1610 PRINT : PRINT •(•;AN(LT);•) Ans i";K,UJ•-/-•;V 1620 JVL=1 : P=U : PRIHT A 1630 HEXT K 1640 OPEN "0",#2. "Br"*AH(LT)*".AVG" : PRINT " F i l i n g ";AN(LT)J". AVG" 1650 FOR K=0 TO NUN 1660 KK=I(K)-N : U=FNCONC(Y(K)) : V'FNCONC<W(K>) 1670 PRINT #2, KK : PRINT 12, U : PRINT 12, V 1680 NEXT X 1690 CLOSE : AA="B:•'AN(LT)•". DAT" 1700 NAME AA AS "B:TEM. TEH" 1710 PRINT : PRINT AN(LT)j". DAT has been renamed." 1720 KILL "B:TEM.TEM" 1730 PRINT : PRINT AN(LT);". DAT has been eraBed.• ! PRINT A 1740 NEXT LT 1750 ERASE I,IX,IY,IS,Y,W 118 1769 ON ERROR GOTO 9 1770 CLOSE : GOTO 150 1780 REN Error subroutine to print positions vhere errors occur. 1790 IF NERRM0 THEN END ELSE NERR=N£RR*1 1800 LPRINT "REDUCE Error Code «";ERR;"in Loop *";LTj"at Line #";ERL 1810 RESUME NEXT •7 B / G . B A S 1000 REM Updated 210186 1010 DIN X(3,256),Y(3,256),Z(3,256),L(4),H<256),C(256>,D<256> 1020 PRINT : PRINT "B/G" : PRINT •••»• : PRINT : NN=-1 1030 FOR 1=0 TO MO 1040 IF VAL(AN<I))=0 THEN NN=NN*1 1050 NEXT I 1060 IF (NN=0) GOTO 1570 1070 FOR 1=0 TO 255 1080 M(I)=0 : C(I)=0! : D(I)=0l 1090 NEXT I 1100 ON ERROR GOTO 1610 : XHIN=2551 1110 FOR N=0 TO NN 1120 ON N GOTO 1140,1150 1130 A="0" : GOTO 1160 1140 A="00" : GOTO 1160 1150 A="000" 1160 OPEN "I",#1,"B:"'A'".AVG" : I=-l 1170 PRINT : PRINT TAB(10);"Reading ";A;".AYG" : PRINT 1180 1=1*1 : INPUT #1,X(N,I) : INPUT #1,Y(N,I> : INPUT «1,Z(N,I> 1190 IF X(N, IXXMIN THEN XMIN = X(N, I) 1200 IF <EOF(1)=0) GOTO 1180 1210 CLOSE : L(N)=I 1220 NEXT N 1230 PRINT : XMIH=ABS(XMIN) : PRINT TAB!10);"Min =";XMIN : PRINT 1240 FOR 1=0 TO N-l 1250 FOR J=0 TO L(I) 1260 K=CINT(X<I, Jl'XMIN) : C(K)=C(K)*Y(I,J) 1270 D(X)=D(K)»Z(I, J)"2 : H<K)=M(K)*1 1280 NEXT J 1290 NEXT I 1300 K=-l 1310 FOR 1=0 TO 255 1320 IF H(I)=0 THEN GOTO 1350 ELSE K = K-1 1339 IF H<I>=1 THEN J=2 ELSE J=MII) 1340 C(K)=C(I)/M<I) : D(K)=SOR(D(I)/(J-D) : M(K)-I-XBIH 1350 HEXT I 1360 OPEH "O",12,"B:B/G.DAT" 1370 FOR 1=0 TO K 1380 PRINT #2,M(I) I PRINT #2,C(I) : PRINT #2,D(I) 1390 NEXT I 1400 CLOSE 1410 KILL "B:0.AVG" 1420 NAME "B:B/G. DAT" AS "B:0.AVG" 1430 FOR 1=1 TO NN 1440 ON I GOTO 1450,1470,1490 1450 KILL "B:00.AVG" 1460 GOTO 1500 1470 KILL "B:000. AVG" 1480 GOTO 1500 1490 KILL "B:0000.AVG" 1500 NEXT I 1510 N=0 : AN(N)="0" 1520 FOR 1=1 TO MO 1530 IF (VAL(AN(I))=0) GOTO 1550 1540 N=N*1 : AN(N)=AN(I) 1550 NEXT I 1560 MO=H 1570 ERASE X, Y, Z, L, H, C, D 1580 PRINT "B/G - average completed and f i l e d . " : PRINT 1590 ON ERROR GOTO 0 1600 CLOSE : GOTO 150 1610 IF (ERL»1160 AND ERR=53) THEN RESUME 1230 1620 IF NERR>10 THEN END ELSE NERR=HERR*1 1630 LPRINT "B/G ERROR Code #";ERR;" i n Line #";ERL 1640 RESUME HEXT 119 A . a M I N U S . B A S 1000 REM Updated 210186 1010 DIM X(2,256),Y(2.256),Z(2.256),M(2),W<2) 1020 A="-*-.." : PRINT A : PRINT "MINUS" : PRINT A : PRINT 1030 ON ERROR GOTO 1350 1040 FOR LT=0 TO HQ. 1050 IF LT=0 THEN 1=0 ELSE 1*1 1060 OPEN "I\#1,"B:*-AN(LT)-".AVG" : K»-l : U = 0! 1070 K=K«1 : INPUT #1,X(I,K> : INPUT #1,Y(I,K) 1080 INPUT #1,P : ZiI,K)=P A2 1090 IF <X(I,K>>20 OR YU.KXU) GOTO 1110 1100 U=Y(I,K) : W<I>=XU,X> 1110 IF EOF(1)<>0 THEN CLOSE ELSE GOTO 1070 1120 H<I)=K 1130 IF (LT=0) GOTO 1300 1140 J=0 : ICR=3 : ICO=0 1150 IF J=8 THEN A=". RET" ELSE A=".CRP" 1160 OPEN "0",#2, "3:"*AN(LT)-A : JCO'-l : JAL=0 : JVL=X (0, M (0))-ICO 1170 FOR K=0 TO M(I) 1180 JMV=X<I,X)-ICR 1190 IF (JMV>JVL) GOTO 1270 1200 FOR JCR=JAL TO M(0) 1210 JLE=X(0,JCR1-ICO 1220 IF JMVoJLE THEN GOTO 1250 ELSE JCO=JC0*l 1230 U=Y(I,K)-Y(0,JCR) : V=S0R(Z(I, X)*Z(0,JCR) ) : P'JMV 1240 PRINT #2,P : PRINT #2, U : PRINT #2, V : JAL=JCR-1 : GOTO 1260 1250 NEXT JCR 1260 NEXT K 1270 PRINT : PRINT TAB( 10) ;AH(LT) <-A;1 has"; JC0*1; "data-seta. " 1280 CLOSE : J=J-1 : ICR=W(I) : ICO=W<0> 1290 IF (J=l) GOTO 1150 1300 NEXT LT 1310 ERASE X, Y,Z,N,W 1320 ON ERROR GOTO 0 : PRINT : PRINT 1330 CLOSE : GOTO 150 1340 REM Error subroutine to print positions where errors occur. 1350 IF NERRM0 THEN END ELSE HERR=NERR»1 1360 LPRINT "MINUS ERROR Code #";ERR;"in Loop #";LT;"at Line =";ERL 1370 RESUME NEXT A . 9 S U D A T A . B A S 1000 REM Updated 070386 1010 DIM XN(3),B(3),C<3) 1020 A="--- .-.." : AA="SUBJECT DATA" : PRINT 1030 PRINT A : PRINT AA : PRINT A : PRINT 1040 LPRINT SPC(27)CHRS(14);AA : LPRINT : LPRINT 10S0 OPEN "I",#1."3:SUaJECT.DAT" : LPRINT : LPRINT 1060 INPUT #1.AM : LPRINT S?C(30)"Name : ";AM : LPRINT 1070 INPUT >1,A : LPRINT S?C(30)" Age : ";A : LPRINT 1080 INPUT #l,A : LPRINT SPC(30)" Eye : ";A : LPRINT 1090 INPUT #1.A : LPRINT SPC(30)"Date : ";A : LPRINT : LPRINT 1100 INPUT #1,XN13) : INPUT #1.AN(1) : INPUT #1,XN(2) 1110 INPUT #1,3(0) : INPUT #1,B(1) : INPUT #1,3(2) 1123 C(3)=3(0)/XN(3) : C! 1)=B111/XN(1) : C(2)=B(2)/XN(2) 1130 INPUT #1,00 : aa=2S3!.QQ : LPRINT 3PC(18)"FLUORESCEIN injected" 1143 GOSUB 1373 : LPRINT : LPRINT 1153 LPRINT SPC!10)"Scaling by Lund-Andersen's F-numbers :" 1163 QQ=.13812»XN(0> : GOSUB 1333 : GOSUB 1390 : PP=QH 1170 BQ=.15731»XN(1> : GOSUB 1340 : GOSUB 1390 : PP=PP*GO 1180 Q0=.13938«XN(2) : GOSUB 13S3 : GOSUB 1390 : QO=PP-GO 1193 GOSUB 1363 : GOSUB 1393 : LPRINT : LPRINT 1233 LPRINT SPCI13)"Ultra-sound results :• 1213 00=313) : GOSUB 1333 : GOSUB 1390 : PP=QO 1220 00=3(1) : GOSUB 1340 : GOSUB 1393 : PP = PP-CO. 1230 00=3(2) : GOSUB 1353 : GOSUB 1390 : 00=PP-00 1240 GOSUB 1360 : GOSUB 1390 : LPRINT : LPRINT 1250 LPRINT SPC(10)"Recalculated F-numbers :• 1 2 0 12S0 O.Q=C(0)/PSCALE : GQSU3 1330 : LPRINT 1270 SG=C(1)/PSCAL£ : GOSUB 1340 : LPRINT 1280 Q0=C(2)/PSCAL£ : GOSUB 1350 : LPRINT : LPRINT : LPRINT 1290 LPRINT SPCI10) "REMARKS Si COMMENTS :•;CHRS<27);CHRS(58) 1300 IF (EOFtl)<>0) GOTO 1320 1310 INPUT #1,A : LPRINT SPC1201A : GOTO 1300 1320 CLOSE : LPRINT CHRS(18);CHRS(12); : GOTO 1400 1330 LPRINT S P C ( 2 2 ) " V i t r e o u a Chamber"; : GOTO 1370 1340 LPRINT SPC134)"Len3*; : GOTO 1370 1350 LPRINT SPC(23)"Aqueous Chamber"; : GOTO 1370 1360 LPRINT SP C ( 4 1 ) ' • : LPRINT SPCI26) " A x i a l L e n g t h " ; 1370 LPRINT " ";CHRS(247);" "; 1380 LPRINT USING " ####. ###" ;QQ; : RETURN 1390 LPRINT " mm" : RETURN 1400 DEF FNSLIJ,?)=C(J'1)«(P-XN<J> >-3<J) 1410 ON ERROR GOTO 2770 1420 XN(1)=XHU)-XN(0) : XN( 2) = XN< 2)-XN (1) 1430 B<1)=3<1)'B(0> : B<2>=3(2)•B<1) 1440 XMV=XN(0)/2 : XMA=(XN(l)*XN(2))/2 : XL=XMV-10! 1450 YL=XMV-10! : PRINT : PRINT : KEY=0 : I=MQ-1 : K2=0 1460 DIM SX(20,I),SY(20,I),SZC20, I) 1470 LPRINT : LPRINT : LPRINT : LPRINT : LPRINT 1480 ON KEY GOTO 1500, 1510, 2560 1500 AF=".RET" : GOTO 1520 1510 AF=".CSP" 1520 FOR I=KZ TO MO 1530 SY(0,I)=-10! : SY<2,I)=lE-22 : S Y ( 3 , I ) = - I 0 ! : IM=9 1540 SY(4,I)=SY(2,I) : SY(6,I)=-10! : ZP=255! : YP=ZP 1550 PP=0l : ?T=0! : PD=0! : PZ=0! : OT=01 : OD=01 : QZ=0! 1560 JJ=-1 : PRINT TAB(18);"Reading ";AN(I);AF;" "; 1570 OPEN "I",#3,"9:"*AN(I)*AF 1580 INPUT #3,PX : INPUT #3,3 ; INPUT 13.R 1590 JJ=JJ»1 : WP=A3S(PX-XMV) : VP=A3S<XMA-?X) 1600 IF (PX<XL OR ?X>YL) GOTO 1620 1618 PT=PT-1! : PD=PD-Q : PZ=PZ-S A2 1620 IF (V(P>YP) GOTO 1640 1630 SXC1,I)=PX-C!0> : SY(1,I)=0 : SZ(1,I)=R : YP=WP 1640 IF (VP>ZP) GOTO 1660 1650 SX(5.I)»FNSL(1, PX) : SY(5,I)=Q : SZ(5.I1=R : ZP = VP 1660 IF (PX>XMV OR S<SY(0,I1) GOTO 1680 1670 SX(0,I)=PX'C(0) : SY(0,I)=Q : SZ(0,I)=S 1680 IF (PX<0! OR ?X>XN(0) OR Q>SY<2,1)) GOTO 1700 1690 SX(2.I)=PX-C(0) : SY(2.I)=a : SZ(2,I)»R 1780 IF (PX<XMV OR ?X>XN(1) OR Q<SY(3,I)) GOTO 1730 1710 IF ?X< = XN(0) THEN SX(3.1)=PX-C<8) ELSE SX(3,1)=FNSL(0, PX1 1720 SY(3.I)=g : SZ(3,I>=R 1730 IF (PX<XN(1) OR PX>XN(2) OR 0>SY(4,I)) GOTO 1750 1740 SX(4,I)=FNSL(1,PX) : SY(4,I)=0 : SZ(4,I)=R 1750 IF (PX<XMA OR Q<SY(6,I)) GOTO 1770 1760 SX(6,I)=FNSL(1,PX) : SY(6,i)»Q : SZ(6,I)=R 1770 I F PX>XN(0) THEN GOTO 1830 ELSE OQ=C<0)-PX 1780 IF (00>3.5 OR 00<2.5) GOTO 1S00 1790 OT=QT-l! : QD=OD-0 : 0Z=QZ-R*2 1800 I F (QO<PP) GOTO 1830 1810 IM=IM-1 : SX(IM,I)=QO : SYIIM,I)=0 1820 SZ1IM,I)=R : PP=PP*3! 1830 IF !EOF(3)=0) GOTO 1580 1840 CLOSE : PRINT • done 1" : PRINT 1850 IF (PT=0!) GOTO 1900 1860 SX(9,1)=XMV-C(8) : SY(9, I)=PD/PT : SZ(9,I)=PT 1878 I F (PZ<=0!) GOTO 1910 18S0 SZ(7,I)=SaR(PZ/PT) : SX(7,I)=PT 1890 SY(7,1)=2!*ABS(SZ(7, I ) ) : GOTO 1920 1900 SX(9,I)=0! : SY(9,I)=0! : SZ(9,I)=0! 1910 SX(7,I)=01 : SY(7,I)=0! : S Z ( 7 , I ) = 0 ! 1920 I F (QT=0!) GOTO 1940 1930 SZ(8,I)=QT : SY(8.1)=OD/QT : SX(8,I)=3! : GOTO 1950 1940 SX(8,I)=0! : SY(8,I)=0! : SZ(8,I)=0! 1950 NEXT I 1960 LPRINT CHHSU8) ;S?C(20) "Name : " ;AM;" [ SUBJECT DATA -•; KEY-2;"]" 1970 LPRINT : LPRINT SPC(10)"(";KEY-1;•) F o r ";CHRS(14); AF;CHRS(28); 1980 LPRINT " f i l e s :•;CHRS(27);CHRS(SS) : LPRINT 1990 LPRINT SPC(10)"The C h o r o i d - R e t i n a l Peak"; : J=0 : GOSUB 2590 121 2800 LPRINT "Mid-Vitreous"; : J=l : GOSUB 2590 2010 LPRINT "Vitreous Minimum"; : J = 2 : G0SU3 2590 2020 IF KEY=0 THEN LPRINT •(Autofluorescencel •; 2030 LPRINT "Lens Peak"; : J=3 : GOSUB 2590 2040 LPRINT "Aqueous Minimum"; : J=4 : GCSU3 2590 2050 LPRINT "Mid-Aqueous"; : J=5 : GOSUB 2590 2060 LPRINT "Corneal Peak"; : J=6 : GOSUB 2590 2070 LPRINT "values closest to the 3-, S-, 9-mm , etc., are 2083 GOSUB 2713 2090 FOR J=10 TO IM 2100 GOSUB 2603 2113 NEXT J 2123 LPRINT CHRS(27);CHRS(58); 2130 IF (KEY=3) GOTO 22S3 2143 LPRINT SPC(13)"The Axial Resolution-Ratio are :" 2153 G0SU3 2333 : LPRINT S?C(13)"At 3 mm"; 2163 LPRINT SPC(23)"About 3 mm* ;CHRS (27) ; CHRS< 45 ); CHRSO) 2173 FOB I = XZ TO .10 2180 LPRINT SFC(25); : PD=VAL(AN(I)) : LPRINT USING "»##.#";PD; 2193 PD=SY(11.I)/SY(0,I) : PT-SY(3.I)/SY(3.I) 2233 LPRINT SPC(24); : LPRINT USING "»##. ###";PD; 2213 LPRINT SPC121); : LPRINT USING "##*.*#*";??; 2223 IF KMO THEN LPRINT ELSE LPRINT CHRS (13); 2233 NEXT I 2243 LPRINT S?C(23)CHRS(27>;CHRS(45);CHRS( 1);S?C(72); 2253 LPRINT CHRS(27);CKRS(45);CHRS(3);CHRS(27);CHRS(53) 2263 LPRINT SPC(13)"The Lover Limit of Detection (or Sen s i t i v i t y ) ->" 2273 LPRINT S?C(15)"Average about Mid-Vitreou3 - 2SD"; 2283 LPRINT CHRS(27);CHRS(33);CHRS(1);"ras*;CKRS(27);CHRS(84); ", are :• 2293 GOSUB 2833 : LPRINT "Fluorescein Equivalent (ng/ml)"; 2330 LPRINT SPC(14)"Number•;CHRS(27);CHRS(45);CHRS(3) 2313 FOR I = KZ TO MQ 2323 LPRINT SPC(25); : PT=VAL(AN(I)) : LPRINT USING •#**.#";PT; 2333 LPRINT S?C(25); : ?T=SY (7, I)-SY U, XZ) : LPRINT USING •##•».*»#";?T; 2343 LPRINT SPC(23); : LPRINT USING '#*##.*»#";SX<7,I) ; 2353 IF KMO THEN LPRINT ELSE LPRINT CHRS(13); 2363 NEXT I 2373 LPRINT SPC(23)CHRS(27);CHRS(45);CHRS(1);SPC(75); 2283 LPRINT CHRS(27);CHRS(45);CHRS(31;CHRS(27);CHRS(58) 2393 LPRINT SPC(13)*The Reproducibility Percentages are :" 2433 GOSUB 2833 : LPRINT SPC(13)"Percent";SPC(13)"About"; 2413 LPRINT CHRS127)CHRS(45);CHRS(3) : JJ=3 2423 FOR J=8 TO 9 2433 FOR 1=2 TO HO 2440 IF (SYU, I-1)=0!) GOTO 2520 2450 PD=VAL(AN(I)) : PT=VAL(AN(1-1)) 2463 IF (PD-?T>>4! THEN GOTO 2520 ELSE JJ=JJ-1 2470 LPRINT SPC(21); : LPRINT USING "#.#.#";PT; 2480 LPRINT • -"; : LPRINT USING •###.*•• ;PD; 2493 PD=133!.(SY(J, D-SYIJ, 1-1))/SY (J, 1-1) 2533 LPRINT SPC122); : LPRINT USING •####.##•;?D; 2513 LPRINT S?C(9); : LPRINT USING •*##.#";SXIJ,1-1) 2523 NEXT I 2533 IF (J = 9 AND JJ=3) THEN LPRINT SPC(3S)*No Value can be calculated!" 2543 NEXT J 2553 KEY=KEY-1 : K2=l : GOTO 1473 2563 ERASE XN,3,C,3X,SY, 3Z 2573 LPRINT CHRS < 27 ) ; CHRS < 65) ; CHRS (12) ; CHRS (27); CHRS (53) ; CHRS (18) ; 2583 CLOSE : ON ERROR GOTO 3 : GOTO 153 2590 LPRINT " values are :" : KW=0 : GOSUB 2710 2630 FOR I = KZ TO MQ 2613 LPRINT S?C(25); : PD=VAL(AN(I)) : LPRINT USING "#*#.#";PD; 2623 LPRINT SPCI23); : LPRINT USING "####. ###";SY(J,I) ; 2633 LPRINT " -/-"; : LPRINT USING "#*».#*••;SZ(J,I); 2643 LPRINT S?C(22); : LPRINT USING "#»#.«##";SX(J, I ) ; 2653 IF I<>(I0 THEN LPRINT 2663 NEXT I 2673 LPRINT CHRS (13) ;S?C (23)CHRS(27 ) ;CHRS ( 45 ) ; CHRS (1) ; - 2633 LPRINT S?C(90)CHRS(27);CHRS(45);CHRS<3> 2693 IF J<13 THEN LPRINT CHRS ( 27) ; CHRS (58) ;SPC (13) "The •; 2733 RETURN 122 2713 G0SU3 2800 : LPRINT ' C o n c e n t r a t i o n -/- S. £. (ng / m l ) * ; 2723 QN KW GOTO 2743 2733 LPRINT S P C ( 1 3 ) " D i s t a n c e from R e t i n a (mm)'; : GOTO 2733 2743 LPRINT S?C(13)'Number o f p a i n t s e n t e r e d * ; 2753 LPRINT CHRS(27);CHHS<45);CHHS<0) 2763 RETURN 2773 LPRINT "INFO ERHGR Code »";ERR;"at L i n e #";ERL; 2783 IF NESR>13 THEN RESUME 2563 ELSE NERS = NEHR'l 2793 RESUME NEXT 2833 LPRINT CHRS (15); SPC (23) CHRS (27) ;CHRS (45) ;CHRS ( 1 ) ; 2813 LPRINT "P.I. Time (min)';SPC(13); 2823 RETURN A . 1 0 B L O O D . B A S 1333 REM Updated 233186 1013 A=" : PRINT A : PRINT *3LC0D" : PRIST A : PRINT 1323 DIM AF(6>,IXI135). IY(1351, T ( 56), Y ( 56), W(55), X(5) 1030 DIM Z I 5 ) , F R ( l l ) . C ( 4 , a > , H S ( 4 , i 3 ) , . - 3 ( 4 , i 3 ) 1040 DEF FNCGNC(X)=34.17«EX?<-139. 339-SORI11915. 9'X/.114441)) 1053 DEF FNI?U)=CINT(.318315.<L0G<X>'139.339) A2-232.586> 1863 ON ERROR GOTO 3523 : M=-l 1373 OPEN "I*,#3,"3:SU3JZCT. DAT* : INPUT #3. AF(5) 1380 PRIST TA3(13);"Name : ";AF(5) ; CLOSE : PRINT : PRINT 1393 FOR X«l TG !<C. 1133 HP=VAL(AN(K)) 1113 IF (HP=3!) GOTO 1133 1123 H=M-1 : FR(M)=KP 1133 NEXT K 1143 OPEN " I " , *2, "3:PLASMA.DAT" : NU.1=-1 : HP = 3! : INPUT *Z,GU 1153 INPUT »2.3U : 3=rNCCNC!GU) : INPUT #2, GV : D=rNCCNC;GV) 1163 NUM = ,W«-1 : INPUT #2, T!NUM) : INPUT *2. GU : INPUT #2, GV 1173 IF T(NUM)>K? THEN H?=T(NU.1) 1133 Y(NU«)=rNCGNC<GU)-3 : W ( SUM)=SSR((FNCCNCIGV))'3-D'2) 1133 I F i u r i 2 ) < > 3 THEN CL33E ELSE GOTO 1183 1233 IF (H?>100!) GOTO 1223 1213 GU=1! : IY(134)=i3 : KC = 99 : GOTO 1233 1222 GU=2!/3! : IY(134)=15 : KC=132 1233 PGR K=0 TO KC 1243 IX(X)=1-CINT(K«GU) 1233 NEXT X 1263 FOR K=3 TO 3 1270 C(X,S)=0! : C(X,7)=99 1283 NEXT X 1290 DEF FNFP(J,T)=C(J,3)»C(J,2).T-C(J,4).T*2 1333 DEF FNFa<J,T)=C(J,l).(T-.2S)*2 1310 IX(133)=CINT(FR(M) ) : XX=0 : KY=0 : J=0 : GOSU3 31S3 1328 AF(J) = *A ' B«t • C«t*2* : G0SU3 3333 : XX=0 : IY(3)=0 : GOSUB 3113 1333 FOR K=3 TO NUM 1343 GU=T(X) : GOSUB 3530 : XK=KK'l 1353 NEXT X 1363 GOSUB 2953 : E=l!/48! : D=C(J,4)*E-C( J , 0 ) / 4 I 1373 GV=C(J,2)/2! : GU=C(J,4)/3! : H P = C ( J , 3 ) / 4 I 1380 FOR X=0 TO M 1390 F=FS(K) : HS(J, K)=D-F« ( C ( J , 3)-GV'F-GU'r"2) 1400 FO( J , X)=FNFO( J, F) - H P ' T M - C I J , 5) « < E - F A 3 / 3 ! ) A2-<FNFP (J, F>) A2/4! 1413 GGSU3 3573 1423 IF HS(J,X)<=3! THEN C( J , 7)=KS(J,X) 1433 NEXT K 1443 LPRINT CHRSI12) 1450 FOR X=i TO XC 1463 H?=FNF?(J,K) : GOSU3 2370 1473 NEXT X 1483 GOSUB 2493 : KX=2 : KY=3 : J = l : AFIJM'A - B . l o g ( t ) • C * C l o g ( t ) J * 2 * 1493 GGSU3 31S3 : GU=C( J, 3)-C( J , 2) : GOSUB 3033 : KK = 0 : GOSUB 3110 1500 FOR K=0 TO NUM 1510 GU = L0G(T(K>> : GOSUB 3500 :. KK=KK-1 1523 NEXT K 1533 GOSUB 2960 : E=L0G(2!) : F=E-2! : H=F A2 : HP=F/4! : D=H/4! 1543 B=C(J,2).HP-C(J,4).O-C(J,0)/4! 1550 FOR K=0 TO a 1560 E=FR(K) : GU=LOG(E)-l! : HS(J, K)= 3*(FNF?(J,GU)-C(J.4))»E 1570 F0(J,X>=FNFQ(J,Z).C(J,3>«(H?-E.GU>'2-CtJ,5>.(E.<GU A2'l!>-D)'2 123 1530 GU=GU-1! : FQ(J,X)= FQ(J,X)•(FNF?(J, GU))A2/4! : GGSU3 3570 15S0 IF HS.'J,K)<=0! THEN C< J, 7) = nSi J, X) 1633 NEXT X 1613 LPRINT CHRSQ2) 1623 FGR X»l TO XC 1630 GU=LCG(K) : KP=FNF?<J,GU) : GGSU3 2373 1543 NEXT X 1650 GDSU3 2453 : KX=3 : KY=1 : J=2 : AF(j)="A-exp(3-t-C»tA2)" 1663 GOSU3 3163 : KK=3 : C(J,3)=EX?(C(J,a)) : C<i,1)»C(J,I)«C(J,3)"2 1673 3=.5'C(J,3> : GGSU3 2333 : GGSU3 3113 1633 OEF FNFE(i,T)=a-EX?(C(J,2>«T<-C(J,4)-TA2) i6sa FOR :<=a TO sua 1703 KK=XK-1 : DW=2!»FNFZ(J, T(K))-Y(X) : GOSUB 3513 1713 NEXT X 1723 GOSUB 2S63 : Z=FNFS(J,1!> : H=E/2! : G=FNFE(J,.5) : GU=G-H 1733 3Y=C(J,1)•(Gu/C;J,3))A2*C (J, 3)•v H-G/2! i A2*C(i,5 >• (K-G/4!)A 2 1748 KY = 2 : G=C(J, 1)/C<J, 3)A2 : H=2!*C(i,4) 1753 0=(G-C(J,3)-C(J, 5)-(C<J,2)-H)A2)'EA2 1763 FOR X=3 TO li 1773 L=CINT(FR(X) : HS(J,K)»GU : FQ,(J,X)=GV 1783 FOR N=KY TO L 1793 F=rNFE(J,N> : HS(J,X)= KS(J, X)-E-F : HP = NA2 1833 C0=<G-C(J,3>«KP-C!J, 5)«HPA2-(C ( J. 2)»H»N)A2/4!)-FA2 1313 FQ(J,X)=FQ(J,X)-D»CQ : D=CO : E=F 1323 NEXT N 1333 GOSUB 3573 : KY=L-1 : GU=HS(J,K) : GV=FQ(J,X> 1843 IF GU<=8! THEN C(J,7)=GU 1853 NEXT X 1863 LPRINT CHRSI12) 1873 FOR K=l TO KC 1883 HP=2!"FXF2(J,X> : GOSUS 2373 18S3 NEXT X 1S33 G0SU3 24S0 : XX=1 : KY=0 : J=3 : Ar(J)="A * 3/t • C/tA2" : GGS'u'3 3163 1913 a'u'=CIJ, 3)-2! »Ci J, 2)-3!«C( J, 4) : KX = 3 : GG3U3 3333 : GOS'u'3 3113 1923 FOR X=0 TO NUH 1933 KK=KX-1 : GU=1!/T(X) : GGSU3 3530 1943 NEXT X 1953 GGSU3 2963 : 3=LCG(21>-.5 : H?=CtJ, 2)"3-C(J,3)/4!-3!»C(J,4) : PRINT 1963 FOR X=3 TO M 1973 E=FR(X) : GU=LOG(E) : GV=1!/E : HS(J,X)=HP-C(J,3>"E-C(J, 2)-GU-C(J,4)«GV 1983 FO(J,X)=FNFO<J,£)-C(J,3)•(GU-3)A2-CtJ,5)•<31-GV)A2 1993 FQ(J,X>=Fa<J,X)-.25"<FNF?U,GV>)A2 : GCSU3 3573 2303 IF HS(J,K><=3! THEN C( J,7)=KS(J, X) 2013 NEXT X 2323 LPRINT CHRS(12) 2033 FOR K = l TO KC 2043 GU=1!/K : HP'FNFP(J,GU) : G0SU3 2373 2353 NEXT K 2363 GOSUB 2493 : GU=lD-22 : LPRINT CHRS(IS) : PRINT : PRINT 2373 PRINT "The Reduced CHIA2 of the FITs are;" : PRINT 23S3 FOR J=3 TO 3 23S3 PRINT : PRINT " ( ";J-l;"> \ C(J,S) ; " ";SGN(C(J,7)> 2138 IF (C(J,7)<=3! OR C(J,S)»>GU) GOTO 2123 2113 GU=C(J,S> : K=J 2123 NEXT J 2123 PRINT : PRINT "The 3EST fit va3 " : PRINT : PRINT " f = ";AF(X) : PRINT 2143 LPRINT SPC (23) "NAME : ";AF<51;" (Plasma Integration)" : LPRINT 2153 LPRINT : LPRINT : LPRINT S?C(131CHRS(27);CHRS(45);CHRS(1); CHRS(14); 2163 LPRINT "PLASMA FLUORESCEIN •;CHRS(23);"Results"; CHRS(27);CHSS(45);CHSS(9) 2170 LPRINT : LPRINT : LPRINT SPC(23)CHSS(27);CHRS(45);CHRS(i ) ; "Time (min)"; 2183 LPRINT 3?C(13)"Concentration (ng/nl)";CKRS(27!;CnRs(45);CHRS(3) 21S3 LPRINT CHRSI27);CHRS(S3); : XP'li 2233 FOR J=0 TO NUM 2213 DW=T(J1 ; 3W=Y(J) : CW=W(J) : LPRINT SPCI27); : GCSU3 2423 : LPRINT 2223 NEXT J 2233 LPRINT : LPRINT : LPRINT SPC(13)CHRS(14);"The BEST fit was <•>" 2243 LPRINT : LPRINT S?C(12)CHRS(14);"f = ";AF(K);"." 2253 OPEN "0",#3, "3:?LASflA. FIT" : PRINT #3, X 2263 FOR J=0 TO 5 STEP 2 2273 PRINT #3,CCK,j> : GU=SCR(A3S1CI K, J'l))) : PRINT #3,GU 2283 NEXT J 124 2298 FOR J=0 TO N 2380 PRIHT *3,FR(J) : PRINT #3,HS(K,J) : PRINT #3,FO(K,J) 2310 NEXT J 2320 ERASE AF. IX. IY.T.X, Y, Z. H.FR.C. HS, FQ 2330 ON ERROR GOTO 3 : LPRINT CHRS(27);CHRS(53); 2340 CLOSE 2358 KILL "3:PLASMA.DAT' 2360 GOTO 150 2370 IF HP>108! THEN GOTO 2400 ELSE IY(K>=3 2380 IF (K<IX(133) AND HP<0!) THEN C(J, 71-HP 2390 RETURN 2400 IY(K)=FNI?(H?) 2413 RETURN 2423 LPRINT USING •##*.#•;DW; 2430 LPRINT SPCIK?); 2453 LPRINT • •/-'; 2463 LPRINT USING '#». H*tt'---';Cil: 2473 RETURN 2433 REN Subroutine for printer-plotting. 24S3 G0SU3 2333 : LPRINT CHRSI15>;CHRS<27>;CKRS(49); 2533 LPRINT SPC123) 'Vertical (LOG) Scaie = X 13 ng/isl / 13 div. "; 2513 LPRINT S?C( 13)'Horizontal Scaie • ' ; IY( 134 ) ;'min / 13 div. • 2523 FOR K=0 TO 53 3 KY= 53-K : KP = XY/5! : N=8 : L--1 : LPRINT CHRS(13); CHRSU3) ;CHRS(9); 2543 IF (HP-FIX(HP)=3) GOTO 2563 2553 LPRINT CNRSI9);CHRS(124); : GOTO 2638 2563 IF (K=1S OR K-25 QR X»35 OR K = 45 OR X = 5) GOTO 2593 2573 IF K=3 THEN LPRINT 'ng/ml -'; ELSE LPRINT CHRSI9); • »•; 2533 GOTO 2533 2593 LPRINT ' 13';CHRS(27);CHRS<83>;CHRS(3);2-XYN13; CHRSI27);CHRS(34);' -"; 2633 FOR KZ=3 TG NUN 2613 IF FNI?(Y(KZ)=KY "KEN L=KZ 2623 NEXT K2 2633 IF !L=-i) GOTO 2713 2643 FOR XZ=3 TO NUN 2553 IF FNI?(Y(KZ) )<>KY THEN GOTO 2S93 ELSE XX=1-CINT(7(KZ) 2663 IF XX=9 THEN XX=1 ELSE XX=KX-N 2673 IF XX<>3 THEN LPRINT SPC ! XX-i ) "a"-; 2633 N=CINT(T(XZ)-l 2653 NEXT XZ 2733 N=9 : LPRINT CHRSI13);CHRS(9);CHRS(9) ; 2718 FOR XZ=3 TO KC 2723 IF IY1XZJOKY THEN GOTO 2763 ELSE XX=IX(KZ) 2739 IF XX=3 THEN XX=1 ELSE KX=KX-N 2743 IF XX<>8 THEN LPRINT SPC(XX-1)CHRS(249); 2753 N»IX(XZ) 2763 NEXT XZ 2773 NEXT X 2783 LPRINT CHRSI 10) ;CHRS< 13) ;CHRS(9) ;CHRSi9); • •; 2793 FOR K=8 TO 9 2803 LPRINT '•'; 2813 FOR Ns0 TO 8 2820 LPRINT 2833 NEXT X 2843 NEXT X 2853 LPRINT ••• : XY-2-IYI134) . LPRINT CHRS I 9) ;CKRS (9) ; 8; 2863 FOR X«l TO 4 2873 LPRINT SPCI16)KY-X; 2883 NEXT X 2893 LPRINT S?C( 17)'aiin" ;CHRS (27) ;CKR3 < 58) ; CHRS127);CHRS(S5);CHRSI12); 2930 LPRINT CHRS 127); CHRS (58) t LPRINT : LPRINT SPC (41) "CHI "2 = ' 2913 CW=C(J,6) : GOSUB 2463 : LPRINT 2920 IF XK>3 THEN CIJ, 6)-C(J,6)/(KX-3) 2933 LPRINT SPC (33) 'Reduced CHI "2 - "; : CW=C(J,S) 2940 GOSUB 2460 : LPRINT CHRSI12) 2950 RETURN 2963 LPRINT CHRSI 13) : LPRINT : LPRINT SPC(10)CHRS(27); CHRSI45);CHRS(1); 2973 LPRINT 'RESULTS of INTEGRATION';CHRS127);CHRSI 45);CHRS(0) 2980 LPRINT CHRS(27);CKRS(S8) : LPRINT SPCI151CHRSI27); CHRS145);CKRS(1); 2990 LPRINT "For 3 < t < T »in. AREA •/- S. E. Imin-ng/ml) *; • 3330 LPRINT SPCI6) 'Percent (X)";CKRS(27);CKRs<45);CHRS(3) 3318 RETURN 3823 REN Subroutine for printing curve-fitting coefficients. 3833 LPRINT CHRSI 18);CHRS(27);CHRs(65);CHRSI12);CKRS(27);CHRS(S3) 125 3040 LPRINT SPCIZaiVMAHE : •;AF(5);' (Plasma Integration)' : LPRINT 3053 LPRINT : LPRINT SPC(10)•(•;SIGHTS(STRS(J-l),1); •) Fitting to i = *;AF(J) 3363 LPRINT SPC(25)"A = "; : 3W=C(J,3) : CW=SCS(A3S(C(J,1))) : GGSUB 2440 3070 LPRINT : LPRINT S?C(25)'3 = •; : 3W=C(J,2) : CW=SCH(A3S(C(J,3)) 33S3 3CSU3 2440 : LPRINT : LPRINT 3?C(25)"C = : 3W=C(J,4) 3093 CW=SCniA3SlC<J,5)) : GGSU3 2443 : LPRINT : LPRINT : LPRINT 3133 RETURN 3113 LPRINT CKRS(27);CHRS(53) : LPRINT SPC(23)CH3S(27); CHRS(45);CHRS(1); 3123 LPRINT "Time (min) Data (ng/ml) CFit-Datai Percent IX)'; 3133 LPRINT CHRS(27);CHXS(45);CKRS(3>;CHRS<15> 3143 RETURN 3153 REM Subroutine for curve-fitting to oroer 2. 3163 FGR X=0 TO 4 3173 X(X)=3! : ZC<)=3! 3180 NEXT X 3193 FOR X=0 TQ NUM 3233 ON KX GOTO 3220.3230 3210 HP=T(K) : GOTO 3240 3223 HP=1!/T(K) : GOTO 3243 3223 HP=LOG(T(K>) 3240 ON KY GOTO 32S3 3253 ' GV=Y(X) : GU=I!/W(X)*2 : GOTO 3280 3263 IF (Y(X)<=3!) GOTO 3353 3273 GV*LCG(Y(K) : GU= ( Y( K > /W ( K) ) A2 3280 Z(4)=GU : Z(3)*GV.GU 3293 FOR N=3 TO 4 3333 X(N)*X(H)-Z<4> : ZI4)=Z(4)»HP 3313 NEXT N 3323 FOR .1=3 TO 2 3333 Z(N)=Z(N)-Z(3) : Z(3)=Z(3)'HP 3343 NEXT N 3353 NEXT X 3363 HP=X<1)/X(3> : 3U=X(2)/XI1) : GV=X(3)/X(2) : 3=X(4)/X(3) : G=GU-G7 3373 D = l!/GV-i!/GU : E=l!/HP-i!/GU : ZI 3)=G-D-(3-GV)•£ 3333 F»Z(0)/X<1)-Z(1)/X<2) : C(J, 4) = <F-D-(Z(2)/X(3)-Z(1)/X(2))*E)/Z(3) 3333 CIJ,3)=IF-C(J,4)'G)/E : C(J, 2>=Z(3)/X11)-C(J, 3)/HP-C(J,4>«GU : F=3-G'J 3400 H=(B»(GU-HP)/GU'HP-GV-G»HP/GV)"2 : C0=3-GV 3413 C(J. l)=C0A2/X(3)-2!•HPA2»C0'F/X(2) 3423 CIJ, i)=C(J, l)-(HP'F)A2/X(2)-2!'HP-G-CO/(X(3)-GV)•(HP'G)"2'3/X(3) 3433 C(J,l) = (C(J.l)-2!>H?A2«r"G/X(2))/H : CO = X(4)/X(2)-XI 2)/X(3) : G=GV-H? 3440 CIJ,3)=(F/GU>A2/X(0)-COA2/(X(3)'GV)-3«GA2/X(3)-2!-F-C0«HP/(GU«X(3>) 3450 C(J,3)=<C(J,3)-21'G-F'K?/X(3)-2!-Ca'G/X(3))/H : F=GV-GU : CO=GU-HP 3460 C!J,5)=(FAX(1)/X(3))A2/X(3)'G"2/(X(3)'GV)'3.(C0/GU)A2/X(3) 3473 C(J,5)=(C(J,5)-2!'(G-C0)-F'H?«X(1)/X(3)A2-2!>G«C0/(GU«X(3)))/H 3483 RETURN 3433 REM Subroutine for printing deviations. 3533 SW=FNF?(J.GU>-Y(X) 3513 LPRINT 3?C(33); : LPRINT USING •#«#.##•;T(K); : LPRINT 3?C(8); 3523 CW=Y(X) : GGSU3 2463 : LPRINT S?C(3); : CW = DV : GGSUB 2463 3533 LPRINT S?C(7>; : CW«100!«DW/Y(X) : GOSUB 2463 3543 LPRINT : C(J.5)=CiJ,S)-iDW/W(X))"2 3550 RETURN 3563 REM Subroutine for printing resuit of integration of fit. 3573 ?0(J,X)=SSR(A3S(rB(J,X))) : LPRINT S?C(18)"T = " ; : XP=8 : DW=F3(K) 3583 3«-KS<J,X> : CM=FSiJrXJ : GGSuo 2423 : LPRINT CHRSI15);" => •; 3590 CW=130!«FB(JfK)/KS(J,K) : G0SU3 2463 : LPRINT CHRS(27);CKSSI5S) 3633 RETURN 3613 REM Error subroutine 3623 IF NERR>13 THEN END ELSE NE3R=NERR-1 3633 LPRINT "3LCQD ERROR Code #";ERR;"at Line #";ERL; 3643 RESUME NEXT 126 1 1 C/VAZ.BAS 1000 REM Updated 123386 1818 DI!t A£(2),H(2),R<2S6),S(25S),U(256>,T(10>,a(6>,ZH(2) 1323 DIM X(2,256),Y<2,256),Z(2,255),W(2),TG(2),0(2),P(2),US(3> PRINT A : PRINT 'C/VAZ' PRINT A 1363 1873 1888 1898 1148 1158 1163 1173 1183 1193 1200- 1233 1243 1253 1263 1283 1293 1338 1313 1328 1338 1348 1353 1363 1373 1383 1398 1433 1418 1423 1438 1443 1453 1468 1523 1533 1543 1553 1563 INPUT #3, AW •(AW INPUT #3, A AEtl)=AN(NC) : INPUT #2,1 ?P = VAL(AEU) ) 1173 1833 A= ••***•• : PRINT 1848 QN ERROR GOTO 3463 1358 OPEN 'I',#3,•BiSUBJECT. DAT' : PRINT TAB(18);'Name ; INPUT #3,A : INPUT #3,A : INPUT #3,0(8) : INPUT #3,0(1) : INPUT #3,0(2) INPUT #3,US(3) : INPUT #3,USUI : INPUT #3,US(2) INPUT #3, W  : O(5)=250!«WW : SS=US (0)/Q( 0) 1138 CLOSE : NC=1 : Q( 3) 3 (US( 3) *US (1) *US(2)) / 2! 1113 KP=0 : NC=NC-1 : J=-l : KC=3 1123 AE(3)=AN(1) : OO=VAL(AE!0>> : 1138 OPEN 'I",#2,"3:PLASMA. FIT" FOR 1=3 TO 5 INPUT »2,CV NEXT I INPUT #2,YY : INPUT #2,CW : INPUT #2, CV IF (YY=aO OR YY=PP) THEN J = J-1 ELSE GOTO 0(J)=63!.CW : CV=60!-CV : P(J)=CVA2 IF J=l THEN CLOSE ELSE GOTO 1173 1218 LPRINT CHRS(27);CHRS(65);CHRS(12);CHRS<27>;CKRS(S31; 1223 GCSU3 1913 : ?P=0(3) : PRINT LPRINT SPC(8)"Centre of Retinal Curvature"; GOSUB 1953 : LPRINT : PRINT LPRINT SPC(19)'FLU0RESCEIN Used =';0(5);' mg ." LPRINT : LPRINT : LPRINT 1273 LPRINT SPC(5)CHRS(14);'!";RIGHTS(STRS<KC'1),1); •) ALIGNMENT by "; IF KC=3 THEN LPRINT "RETINA" ELSE LPRINT "C/R PEAK" IF XC=3 THEN AP=".RE7" ELSE AP=".CRP" OR 1=3 TO 1 LPRINT : LPRINT : A="3:"*AE(I)*AP : OPEN "I",#1,A L=-l : TT=3! : RR=1E*22 : XX=RR : PRINT TA3I13); •Reading ';A : PRINT L=L-1 : INPUT #1,R(L) : X(I,L)=R(L) : INPUT #1,S(L) Y(I,L)=S(L) : INPUT #1, YY : U(LXYYA2 : Z(I,L)=U(L) 00=SS-R(L) : QQ=ABS(00-3!) : PP=ABS100) IF (QO>RR) GOTO 1383 RR=00 : WW=S(L) IF (PP>XX) GOTO 1400 XX=PP : W(I)«S(L) : TG(I)=U(L) IF (00>31 OR S(LXTT) GOTO 1420 TT=S(L) : ZH(I)'RIL) IF (EOF(1)=0) GOTO 1333 CLOSE : LPRINT CHRS(18) : LPRINT SPC(20)'File : •; LPRINT CHRS(14);AE(IXAP;CHRS<20>;" has";L-1;"data-seta. LPRINT : N(I)=L : KEY=0 : GOSUB 2B10 NEXT I 1470 0=W(3)/WW : LPRINT CHRSI18) : LPRINT SPC(5)CHRS<14); 1480 LPRINT 'C/R CORRECTION CONDITION";CHRS(28);• *"; 1498 GOSUB 2020 : PP=11 : OQ=0! : WW=1! : GOTO 1573 1533 WW=W(1)/W<3) : XX=TG(1)/W(1)"2-TG<3)/W(3)A2 : TT=3! 1S13 FOR X=3 TO NO) YY=Y(8,X) : Y(0,X)=YY"WW IF (SS»X<3,X)>3! OR Y(8,XXTT) GOTO 1553 TT=Y(8,X) : ZH(0)=X(3,X) NEXT K SO=WW.SCR(XX) : P=WW 1573 GOSUB 1913 : LPRINT SPC!6)CHRSI14) ; 15ae LPRINT "For a C/R RATIO";CHRS<23); • ="; 1590 G0SU3 2300 : LPRINT CHRS(18) 1633 IF KC=0 THEN GOSUB 2100 ELSE GOSUB 2050 1613 LPRINT S?C(10)CHRS(14);AE(l);CHRS(20);"-minute minus •; 1620 LPRINT CHRS(14);AE(3) ;CHRS(28);"-minute scan leaves'; 1630 LPRINT L'l;"data-sets." : LPRINT 1648 1=1 : GOSUB 2210 : J5=0 1650 FOR K=0 TO L 1663 IF (R<K)>0(3)> GOTO 16S0 1670 IF (R(K)>0! AND S(X)>0!) THEN J5=J5-1 1680 NEXT K 1690 IF (J5>11) GOTO 1720 1700 LPRINT CHRS(18);SPC(16)"0nly';J5; •data-sets => no file was created.• 1710 GOTO 1840 127 1723 IF WW=1! THEN K = 3 ELSE X»l 1733 KEY=2*KC-1 : AA= *.CV*-RIGHTS(STRS<K'KEY>,1) 1743 A="3:"*AE(I)-AA 1753 OPEN '0\»3,A 1763 FOR K = 8 TO L 1773 IF (R<K>>8<3>> GOTO 1813 1783 IF (RIKX81 OR S(XX=8! OR U(KX=8I) GOTO 1833 1793 OO-SBRIUIK) : PRINT #3,R!K) : PRINT #3,S(K> : PRINT #3,00 1833 NEXT X 1813 CLOSE : LPRINT CHRSU8) ;S?C(13)J5; 'data-sets »ere saved in •; 1823 LPRINT CHRSI14);A : LPRINT : LPRINT : LPRINT 1833 LPRINT SPCI25)'Deleted : Yes [ 1 No C J* 1843 IF Ww>l! THEN GOTO 1533 ELSE KC=KC-1 1853 IF (KC=1) GOTO 1213 I860 IF <NC<MO) GOTO 1119 1873 ERASE AE, N, R, S, U, T, 3, ZH, X, Y, Z, W, TG, 0, P, US 1883 LPRINT CHRS (27); CHRS 158); CHRS (27); CHRS (65); CHRS (12); 1893 ON ERROR GOTO 3 : LPRINT CHRS( 27) ; CHRS( 53) 1930 CLOSE : GOTO 153 1913 XP=KP-1 1923 LPRINT CHRS (12); CHRS( 18) ;SPC ( 25) "NAME : *;AW; • I C/VAZ -';KP;']" 1933 LPRINT : LPRINT 1940 RETURN 1958 LPRINT • •;CHRS(247); • •; 1960 LPRINT USING •##.•#####• ;PP; 1970 LPRINT • na f ran. the RETINA. ' 1980 RETURN 1993 LPRINT SPC(40); : LPRINT USING "##.###'"•;00; : LPRINT SPC120); 2000 IF PP=-2.2E-22 THEN G0SU3 3360 ELSE LPRINT USING •##. t t t t l " " " ; ? ? ; 2013 LPRINT '•/-'; 2023 IF QQ--2.2E-22 THEN GOSUB 3360 ELSE LPRINT USING "»#.#####"A"";C0 2030 RETURN 2043 REN Subroutine ta (re-order and) subtract data-sets. 2050 FOR J*3 TO 1 2363 FOR X=0 TO N(J) 2373 X( J,XXX(J,X)-ZH!J> 2383 NEXT X 2393 NEXT J 2133 L=-l : «=8 2110 FOR K=0 TO N(l) 2123 IF (X(1,X)>X(0,N(0))) GOTO 2190 2130 FOR i=M TO N(0) 2140 IF X( 1, K) <>X(0, J) THEN GOTO 2170 ELSE L=L-1 2150 S(L)=Ytl,X)-Y(3, J) ; R(L)=X(1,K) : U(L)= Z(1, X)-Z(3,J) 2163 M=J-1 : GOTO 2183 2173 NEXT J 2183 NEXT X 2198 RETURN 2238 REM Subroutine for printer-plat. 2218 LPRINT CHRSI15) : LPRINT CHRS(27);CHRS<49); 2223 LPRINT SPC125)'Vertical (LOG) Scale * X 13 / 10 div.'; 2233 LPRINT SPC (131'Horizontal Scale > 1 u / 11 div.'; 2240 PJ=0t : RR*Q(0XO<1) : Q0=RR-a(2) 2250 FOR K=3 TO L 2263 IF <R(K>>0(3>> GOTO 2283 2273 R(X)=SS«R(X) : YY=R(X) : GOTO 2333 2280 IF (R(X)>RR) GOTO 2333 2298 R(K)»US(l)»(R(K)-a(3))/a(l)-YY : 00=R(K) : GOTO 2343 2333 IF (R(K)>flO) GOTO 2323 2313 R<K)=US(2)MR(K)-RR)/S(2)*00 : TT=R(X) : GOTO 2348 2323 R(K)=PSCALE»<R(K)-OQ)-TT 2333 IF <R(K)<0(3) AND S(K)>PJ) THEN PJ=S(K) 2343 NEXT X 2358 JY=3 2353 IF PJ>.8364 THEN JY=CINT(. 018315*(LOG(PJ)-109.339)*2-212.536) 2370 FOR K=0 TO 30 2388 J=33-K : YY»J/5! : M=3 : LPRINT CHRS(13);CHRS(13);CHRS(9); 2393. IF ((YY-FIX(YY))=0!) GOTO 2410 2433 LPRINT CHRS(9) ;CHRS(124); : GOTO 2463 2410 IF (X=5 OR K=15 OR X=25) GOTO 2448 2423 IF K*8 THEN LPRINT 'ng/inl •"; ELSE LPRINT CHRS< 9); * •" ; 2438 GOTO 2463 2443 LPRINT " 18*;CHRS(27);CHRS(83);CHRS(8);INT(J/13); 2453 LPRINT CHRS(27) ;CHRS(84); * <•; 2463 IF (J>JY) GOTO 2573 128 2473 FOR J5=3 TO L 2483 XW=0 2493 IF S(JS)>.3364 THEN KH-CINTl.313315' (LOG 1S(J5))'139.339)'2-212. 536) 2533 IF XWoJ THEN GOTO 2563 ELSE XL-CINTI13! «R( J5))-1 2513 IF (KL>U3 OR KL<0> GOTO 2563 2523 IF KL=0 THEN XR=1 ELSE XR=KL-H 2533 IF <KR=0> GOTO 2553 2543 LPRINT S?C(KR-1)CHRS<249>; 2553 M=KL 2563 NEXT J5 2573 NEXT X 2533 00=lE-22 : RR=3! : 00=00 : TT=RR : PP=111 : YY=0! : LPRINT 2593 FOR K=3 TO L 2633 IF !R(K)>PP) GOTO 2673 2613 'IF (R(KXYY) GOTO 2663 2623 IF S(X)>RR THEN SR = SIK) 2633 IF S(X)<00 THEN 00=S(X) 2643 IF R(K)>TT THEN TT=R(X) 2653 IF R(K)<00 THEN 00=R(K) 2663 NEXT X 2673 PP=3! : LPRINT CHRS(9);CHRS(9); • "; 2633 FOR K=3 TO 113 2693 IF (K/IBOPP) GOTO 2713 2733 • LPRINT : PP=PP*1! : GOTO 2723 2713 LPRINT 2723 NEXT X 2733 LPRINT CHRS(9);CHRS(9);" R";S?C(19)"2*;SPC(19)"4"; 2743 LPRINT SPC(19) "6" ;S?C( 19) "8" ;SPC(13)113';SPCI8) -M' : XEY=1 2753 LPRINT : PP=OB : LPRINT SPC(51) "LEFTMOST point*; : GOSUB 1953 2763 LPRINT : PP=TT : LPRINT SPC(53)"RIGHTMOST point'; : GOSUB 1953 2773 LPRINT : PP=RR : LPRINT 5?C(44)"MAXIMUM"; : GOSUB 3423 2783 LPRINT : PP = Q0 : LPRINT SPC(44)"MINIMUM" ; 2793 GOSUB 3423 : LPRINT : LPRINT : LPRINT 2833 LPRINT CHRSI27);CHRS(65);CHRS(12);CHRS(27);CHRS(S3);CHHS(13) 2813 XW=0 : XR=8 : TT=lE-22 : GOSUB 3333 2823 XX=-99 : XO=KK : J5=KX : KT=KK : KL=XT 2833 FOR K=0 TO L 2843 IF XEY=1 THEN 00=R(X) ELSE 00=SS"R(K) 2853 IF 00>6.5 THEN GOTO 3313 ELSE C0=ABS(31-00) 2863 IF (00>TT) GOTO 2883 2878 TT=00 : XT=K 2888 IF (S(KX=8!) GOTO 3388 2393 IF 00<2. 5 THEN GOTO 2978 ELSE KW = KW»1 2933 IF XW=1 THEN J5=K ELSE XL=X 2913 PP=(B(3)-00)*2 : T(3)=S(X)»PP : XX=1.59293E-33.00A2-. 3381543 2923 PP=41-(0(4)-XX)/PP : T(5)=(U(K)/S(K)A2-PP)*T(3)*2 2933 IF (XW = 1) GOTO 2963 2943 PP=T(2XT(3) : Q0=00-TI4) : RR=PP"0Q/2! : T(8)=T(8)*RR 2953 Tll>"T(l)-((T(5)-T(S))/PPA2-(T(9)-XX)/QQA2)«RRA2 2963 TI4X00 : T(6)»T(5) : T(2)=T(3) : T(9)=XX 2978 IF (00<21 OR 00>4!) GOTO 3888 2988 T(7)=T(7XS(K) : T(8)=T(8)-U(K) : KR=KR-1 2998 IF KR=1 THEN XK=K ELSE KQ=K 3338 NEXT K 3313 FOR X=2 TO 6 3323 T(X) = -9.9 3833 NEXT X 3348 IF XEY=1 THEN PP=1! ELSE ?P=SS 3350 IF J5*>3 THEN T(2) =R(J51 "PP 3863 IF KL=>3 THEN T(3)=R(XL)«PP 3873 IF XK=>8 THEN T( 4) =R (XK) "PP 3383 IF KO=>0 THEN T(5)SR(XO)*PP 3393 IF XT=>3 THEN T(6)=R(XT).PP 3188 LPRINT SPC( 18) CHRS (27) ; CHRS (43) ;CHRS( 1) ; "PERMEABILITY "; 3113 LPRINT "COEFFICIENT t PENETRATION RATIO"; 3123 LPRINT CHRS(27);CHRS(45) ;CHRS( 3);CHRS(27);CHRS(53) 3133 LPRINT ; YY=P(I)/0(I)A2 : LPRINT SPC(15)"Mean PERMEABILITY •; 3140 LPRINT "COEFFICIENT (cn"; : J5=l : GGSU3 3308 : SS=-2.2E*22 3158 LPRINT SPC(33)"is •; ; PP=. 32»T(3)/(0(I)-0(3) A2) 3163 IF (T(3X>0!) THEN OB=PP»SBR ( Tt1)/T<8>A2«YY»4!"B<4>/0(3)A2) 3178 GOSUB 2300 : LPRINT : LPRINT SPCIlSJ'Mean PENETRATION RATIO (•; 129 3180 G0SU3 3330 : LPRINT S?C(33)"i3 • ; 3193 IF <KR=3) GOTO 3213 3230 PP-Tt7)/(Q(I).KR) : C5=?P.SCR (T<8)/T( 7) *2-YY) : GOTO 3223 3213 PP = -2.2£-Z2 : S==PP 3223 GC3U3 2300 : LPRINT 3233 LPRINT S?C( 15) "PENETRATION RATIO </s), defined at"; 3240 IF (KT<0> GOTO 3270 3230 J5=6 : GCSU3 3330 : PP=S(KT)/0( I) : LPRINT S?C(30)*i3 •; 32S3 aB.=A3S<PP>.SSR(YY.U(KT)/S<XT)',2) : GGSU3 2300 : GOTO 3230 3270 GOSUB 3360 : GOSUB 3340 : GOSUB 3363 : LPRINT *•/-"; : G0SU3 3360 3230 LPRINT : LPRINT : LPRINT : LPRINT 3290 RETURN 3300 LPRINT "/a), defined from"; : J5=J5»1 3313 IF T1JSX3! THEN GGSUB 3360 ELSE LPRINT T(J5); 3320 LPRINT "to"; : J5=JS-1 3330 IF TCJ5X01 THEN GGSUB 3360 ELSE LPRINT T(J5); 3343 LPRINT "mm , ' 3350 RETURN 3363 LPRINT " ••. ; 3370 RETURN 3383 FOR X=3 TO 9 3390 TIKX0! 3400 NEXT X 3413 RETURN 3423 LPRINT " concentration *;CK3S<247>;* •; 3433 LPRINT USING "##.##### ;?P; : LPRINT " ng/ml ." 3443 RETURN 3453 REM Subroutine for error printing. 3463 IF (ESHool) GOTO 3493 3473 CLOSE : A=" A:"-AE( I)-AA 3483 RESUME 1753 3490 IF NERR>13 THEN RESUME 1873 ELSE NERR = NERR*1 3533 LPRINT "C/VAZ ER3GS Cade #";£3R;"in Line #";ERL 3513 RESUME NEXT A. 1 2 SLOPES. BAS 1000 REM Updated 090386 1010 DIM T(5),*<5),3(4),E(4),P<13>,S<13>,U<13),C(13,3), 0(13.3).0(13.4).H(13.3) 1023 DIM X(230),Y(200>,Z!200>,3(13,9),8(13.9).G(13,S). F(13.S),H(13,3),V(4,6) 1030 A-*..".." : PRINT : PRINT A : PRINT "SLOPES* : PRINT A : PRINT 1040 ON ERROR GOTO 2310 : NERR.0 1050 OPEN *I\#3."3:SU3JECT.DAT" : INPUT #3, AE 1060 PRINT TABUS); 'Name : ";AE : PRINT 1070 INPUT #3, A : INPUT #3,A : INPUT #3.A 1080 INPUT #3,SO : INPUT #3.SS : INPUT #3,?P 1090 INPUT #3,?P : INPUT #3,3S 1100 CLOSE : B(3)=RR/3!»PP : B(0XPP/3! : SC=PP/CO 1110 3<1)*2!«B(0) : BI2XPP-31 : SL=RR/SS 1120 FOR 1=9 TO MQ 1130 P(I)*VAL(AH(I>) 1140 NEXT I 1130 OPEN "I", #2.'3:PLASHA.FIT* : INPUT #2,1 : J»l 1160 FOR 1=0 TO 5 1170 INPUT #2. VV 1130 NEXT I 1190 INPUT #2,VV : INPUT #2. SS : INPUT #2. PP 1230 IF (VV<>P(J)  GOTO 1223 1210 SIJX631.SS : U(J)"63!«PP : J-J'l 1223 IF E0F(2X>9 THEN CLOSE ELSE GOTO 1190 1230 FOR 1=9 TO HQ 1240 A«'3:".AN(I).*.AVG* : OPEN "I*,#2,A : PRINT 1259 J—1 : BB'0I i CC=0t : SS-91 : SG-3t»BC3> : SD-SG/SC 1269 IF (EOF(2X>0 OR JJ«199 OS SS">B<3)> GOTO 1350 130 1270 INPUT #2, PP : INPUT #2,00 : INPUT #2, RR 128a IF PP>SD THEN SS=SG»(PP-SD>ASL ELSE SS-SC-PP 1290 IF <SS<"91) GOTO 1260 1300 JJ-JJ-1 : X(JJ)=SS : YUJ)=QO : Z(J)=RR 1310 IF (SS>B<0) OR QQ<BB) GOTO 1330 1320 E(0)=SS : BB'SQ 1330 IF (SS<B(1) OR 00<CC) GOTO 1260 1340 CC=00 : E(3)=SS : GOTO 1260 1350 CLOSE : PRINT •(•;I-l;•)•;JJ*1;'data-sets ure used from •; 1360 PP=E(3)-£(0> ': E(1)=PP/3I : E(2)>2I«E(1) : 0(I,3)"JJ»1 1370 E(l)*E<a)»E(l) : E<2)=E<0>-El 2) : (1=0 1380 FOR J=0 TO 2 1390 PRINT TAB! 10); ; 1400 IF J=l THEN PRINT •HID-"; 1410 IF J<=1 THEN PRINT 'POSTERIOR '; ELSE PRINT 'ANTERIOR '; 1420 PRINT 'VITREOUS •••• : VW=1E.12 : 11=0 1430 FOR XK=0 TO 3 1440 GOSUB 2540 : 22=01 : YY=0! 1450 FOR X=H TO JJ 1460 IF (X(K)>E(J-1)) GOTO 1500 1470 IF X(KXE(J) THEN GOTO 1490 ELSE PP-X(K) 1480 GOSUB 2930 : ZZ=ZZ-1! : YY=YY-((Y(K)-BB)/Z(K))A2 1490 NEXT X 1500 IF ZZ>21 THEN YY*YY/(ZZ-21) 1510 PRINT ,Fit';XK;'=> Reduced CHI'2 =';YY 1520 IF (YY>WW> GOTO 1540 1530 WV*=YY : KC=KK : XX=ZZ 1540 NEXT KX 1550 H=K : 11=1 ; 0(I,J)=XX : H(I,J)-*K : XK=KC : N(I,J)=KX 1560 FOR K=0 TO 2 1570 NN=3«J«K : L=2"X : 011, NN)=V(KC,L) : R(I, NN)=V(XC,L-l) 1580 NEXT K 1590 PRINT TAB(10);'9est for fit =';KC : PRINT 1600 IF J = 2 THEN PP=B(2), ELSE PP=6l«J<-31 1610 GOSUB 2930 : C(I.J)=BB : D(I,J) = DD : G(I,J)=GG 1620 F(I,J)=HH : l'J-3 : GU.L)=FF : F(I,L)=EE 1630 NEXT J 1640 NEXT I 1650 GOSUB 2240 : LPRINT SPCI25)CHRSI14);'REDUCED CHIA2' 1660 LPRINT CHRSI15) : GOSUB 2300 1670 FOR 1=0 TO HO 1680 PP=P(I) : GOSUB 2430 : LPRINT • •; 1690 FOR J=0 TO 2 1700 00=H(I,J) : GOSUB 2490 : LPRINT • -';INT!0<I,J)); •-';N(I,J);SPC(4); 1710 NEXT J 1720 LPRINT ' ';INT(0(I,3)) 1730 NEXT I 1740 LPRINT CHRSC18) : LPRINT SPC (25) CHRSI 14)"CONCENTRATION'; 17S0 LPRINT CHRS(20);" - C"; : A-'-l" 1760 GOSUB 2270 : LPRINT " 1" : 11=0 : GOSUB 2350 1770 FOR I>0 TO HQ 1780 PP«P(I) : GOSUB 2430 1790 FOR J»0 TO 2 1800 IF (1=0) GOTO 1820 1810 C(I,J)=C(I,J)-C(0,J) ; D(I,J)=D(I,J)*D(0,J) 1820 PP=C(I,J) : QQ=SQR(ABS(D(I,J)>> : GOSUB 2460 1830 NEXT J 1840 LPRINT 1850 NEXT I I860 11=0 : L=l : XX=0t ; YY=0t 1870 LPRINT CHRSI18) : LPRINT SPC(20)CHRS(14);'PENETRATION RATIO 1880 LPRINT CHRS(20);' - f s'; : A="-l' 1890 GOSUB 2270 : LPRINT • ]• : GOSUB 2350 1900 FOR I=L TO HQ 1910 PP=P(I) : GOSUB 2430 : BB=(U(I)/S11))'2 : K=0 1920 IF (11=0) GOTO 1940 1930 XX=C(1,X) : YY=D(1,X) 1940 00=C(I,K)-XX : PP=00/S(I) 1950 0Q=ABS(PP)"SQR(D<I,KXYY)/00A2-BB> : GOSUB 2460 1960 IF (X=2) GOTO 1980 1970 LPRINT SPC(28); : K=2 : GOTO 1940 1980 LPRINT 131 1999 HEXT I 2399 11=11-1 : L=2 2313 IF (11=I) GOTO 1373 2029 LPRINT CHR9(1S> : LPRINT S?C(27)CHR9<14);•GRADIENT*; 2333 LPRINT CHRŜ );" - t ng.»l*; : A»'-l' : GOSUB 2273 2040 LPRINT •.»••; : GOSUB 2273 : LPRINT • ]• : GOSUB 23S3 2050 FOR 1=3 TO RO 2360 PP=P(I) : GOSUB 2433 2973 FOR J=9 TO 2 2380 PP'G(I.J) : OQ=SOR(ABS(F(I,J)) : GOSUB 2469 2090 NEXT J 2199 LPRINT 2119 NEXT I 2123 LPRINT CHS9UB) : LPRINT S?C( 1S)CHR9( 14) ;'DIFFUSION CONSTANT * ; 2129 LPRINT CHR9I20);* - C cn'; : A="2" : GOSUB 2270 2140 LPRINT "a*; : A»*-l* : GOSU3 2270 : LPRINT • ]• : GOSUB 2280 2150 FOR 1=1 TO SO 2163 PP'P(I) : GGSUB 2430 : BO=G(1,1)-G(I. 3) 2170 LPRINT S?C(23); : PP=C(1,1>/(1000-PP-CC) 2180 00=0(1,i)/C(I, l)"2-.25/P(I)"2-(F(I.l)-F(I,3))/00"2 2190 00= PP-SCRUBS(CO) ) : GOSUB 2463 : LPRINT 2230 NEXT I 2210 ERASE T, W. 3, E. ?, S. U, C. D, 0. N. X, Y, 2, 3. R, G, F, 3, V 2223 QN ERROR GOTO 3 : LPRINT CHRSI12);CNRS<27);CHR9(Sai; "2239 CLGSZ : GOTO 153 2243 LPRINT Ch'RS(12) ;C:-i39(13) ;CHR9(27) ;CHR3(S5) ;CHRS(12); CHRSI27);C4RS<50); 2259 LPRINT S?C(23)*MAnE : •;A£;• C SLOPES - SETina 1' : LPSIHT : LPRINT 2263 RETURN 2273 LPRINT CHR9(27);CHRS(83);CHR9(3); 2283 LPRINT A;CHRS(27);CHR9(84); 2293 RETURN 2330 LPRINT S?C(23)CHRS(27);CKR9(45);CHR9(1);*TIME (am)"; 2213 LPRINT SPCC5)'POSTERIOR VITREOUS";S?C(9)"SID-VITREOUS'; 2323 LPSIHT 5?C(9)"ANTERIOR VITREOUS';S?C(7)'Points"; 2333 LPRINT CH3S(27);CHRS(45);CHRS(3) 2343 RETURN 2253 LPRINT CHR9I27);CH39(38>;SPC(35)'after •; 2369 IF 11=0 THEN LPRINT "BACKGROUND"; ELSE LPRINT "BOLUS"; 2373 LPRINT • subtraction" 2330 LPRINT CHRS(15) : LPSIHT S?C(201CHR9I27);CHR9(45);CHRs<1); •TISE (nin)"; 2390 LPRINT S?C(5)•POSTERIOR VITREOUS (3 nm/R)'; 2409 LPRINT S?C(5>'SID-VITREOUS (9 n/RI'; 2410 LPRINT 3?C(5)'ANTERIOR VITREOUS (3 an/L)'; CHRS(27);CHRS(45);CHRS(0) 2420 RETURN 2430 LPRINT S?C(22); 2440 LPRINT USING • f».##';PP( s LPRINT SPC(4); 2450 RETURN 2460 LPRINT S?C(S); 2470 LPSIHT USISG •$*.»**"—•;??; 2489 LPRINT • •/-•; 2490 LPSIHT USING •#«.»#»"*«• ;00; 2300 3ETURX 2510 IF HERR>10 THEX 3ESUHE 2213 ELSE NE3R*NE3H»1 2523 LPRINT "SLOPES E3S03 Code «*;£3R;"in Line #';E3L 2530 RESUME NEXT 2540 FOR K=0 TO 4 2550 T(K)=0l : ¥(K)*0t Z"iS0 NEXT X 2570 FOR K=« TO JJ 2580 IF (X(X)>E(J-1)) GOTO 2730 2390 IF (XIXXE(J) CR Y(KX=0l> GOTO 2729 2639 OX KK GOTO 2620, 2630,2640 2610 BB=X(K) : CC=Y(K) : DD=Z(K)"2 : GOTO 2559 2623 aa = LCG(X(X)l : CC=Y(K) : DD=ZCX>,2 : GOTO 2553 2633 3B»X(X) : CC=LCG(Y(X) : DD-( Z( K)/? (X)) "2' : GOTO 2559 2640 BB-X(!C) : CC=1I/Y(K) : DD=(Z(K)/Y(K)*2)*2 2559 »(4>»lt/DD : ¥(3)"W(4)'CC 2669 FOR L=« TO 4 2679 T(L)=T(L)-»(4> : V(4)*W(4)>BB 2680 NEXT L 2699 FOH L=0 TO 2 2799 »(L)»»(L)-»(3) : *<3)«¥(3>«8B 2710 HEXT L 2729 HEXT I 132 2739 BB=T(1)/TI0> : CC=TC2)/T(1> : DD=TI3)/T(2> : EE=T(4)/T(3> 2749 FF=CC-DD : GG*1!/DD-l1/CC : HH»1I/BB-1!/CC 2750 00=FF-GG-HH"<EE-DD) : PP=W(9)/T(1)-W(1)/T(2) 27S9 Q0=(PP«GG-<W<2)/T(3)-W<1)/T(2) )«HH)/W<3) : V(KK, 41=00, 2779 RR"<PP-0O.FF)/HH : V(KX,0)=RR : SS*W(0)/T( 1)-RR/BB-QQ*CC 2789 V(XK.2)=SS : PP=EE-CC : TT=(EE»<CC-BB)/CC<-8B-DD-FFABB/DD)A2 2799 UU=EE-DD : Q0=UUA2/T(9)-2!*UU»PP-BBA2/T(2) 2899 QO=aQ»<BB'PP)A2/T(2)-2I ABB»FF»UU/(T(9)ADD)*<BBAFF)A2-EE/T(3) 2819 00=<Q0A2!ABBA2»PP»FF/T<2))/TT : V(XK,I>"OQ 2829 UU=T(4)/T(2)-T(2)/T(0> : FF=DD-3B 2839 RR»<PP/CC)A2/T<0)-UUA2/(T(3>«OD)-EE"FFA2/T(3> 2849 RR=RR-2!«PP'UU"BB/(CC»T<3))«2I•FF»PP«BB/T(3) 28S9 RR»(RR-2!AUUAFF/T(3))/TT : V(KX,3)=RR 2869 PP*DD-CC : UU=CC-BB : SS= (PP»T( 1)/T(3) ) A2/T(8) 2879 SS=SS*FFA2/(T<3)"DD)->EE'(UU/CC) A2/T(3) 2888 SS=SS-2!A(FF-UU)APP"3B«Ttl)/T(3)A2 2899 SS=<SS-2!AFFAUU/<CC»T!3)))/TT : V<KK,5)=SS 2988 IF KK<>2 THEN RETURN 2919 V(XK,3)=EXP(V<KK,a) : V<KK, 1)=V<XX,1)»V<XX,9)A2 2928 RETURN 2939 ON XX GOTO 2958,3989,2949 2949 CC=PP : EE=1! : FF=2!ACC : GOTO 2968 2958 CC=L0G(PP) : EE=1!/PP : FF=2!»CC-EE 2968 BB=V(XX,9l*V<XX,2)ACC-y(XK, 4)»CC*2 2978 IF KX=3 THEN 3B=1!/BB 2988 , IF 11=9 THEN RETURN 2998 DD=V(KK,1) *Y ( KK, 3)»CCA2*V(KK, 5)»CCA4 3809 GG=V(KK,2)-EE*V(KK, 4)»FF : HH=V<KK,3)»EEA2*V<KK,5>»FFA2 3819 ON KK GOTO 3030,3070,3040 3020 EE=4IAV<KK. 5) : RETURN 3030 FF=EE»C-GG-FF) : EE = EEA2*(HH*41"V(KK, 5)»EEA2) : RETURN 3040 CC=BBA4 : DD=DD«CC : GG=-GG*BBA2 : HH=HHACC«DD-<2!-GG/BB)A2 3050 CC=GG/BB : FF=21•(GG»CC-V(KK, 4)«BBA2) : EE=16I»HH»CCA2 3069 EE = EEA0D<4!»CCA4«-(4t»BBAV<KK, 4))A2>*41»V<KK,5>»BBA4 3070 RETURN 3080 CC=PP : BB=V<KK,0)'EXP<V(KK,2)-CC»V(KK,4)»CCA2) 3898 IF 11=0 THEN RETURN 3100 DD = V(KK,1)•(BB/Y(KK, 0))A2*V(KK, 3)• ( CC'BB)A2-V(KK,5)•<BB'CCA2)A2 3110 FF=V(KK,2)»2I»V(KK,4)-CC : GG=BB»FF 3120 HH=DD-FFA2*<V(KX,3)-4I»V<KK,5)ACCA2>»BBA2 : EE=HH«FFA2 3130 FF=FF«G(WI-V<KK,4)ABB : EE=EE*Y(KK,3)"GGA2 3140 EE=EE*4I•V(KK,5)A((GG'CC)A2-BBA2>*41"DD"V(KK,4)A2 3150 RETURN A . 1 3 P L O T . B A S 1800 REN Updated 210186 1018 DIN L(256), .1(256), N<3), SXI15), SYU5), S2!15>, B(3) 1020 A»** — : PRINT : PRINT A : PRINT 'PLOT MENU." : PRINT A 1030 PRINT : PRIHT TAB(191;'The files in Drive 3 are:' : FILES'3:*.•' 1040 PRINT : PRINT : PRINT TAB(10):'FILE TYPE" : PRINT TAB(10);A 1059 PRINT "(0) .AVG files -> Averaged rav data" 1060 PRINT "(1) .RET files -> Retina-aligned, b/g-subtracted" 1070 PRINT "12) .CRP filea -> CRPeak-aligned, b/g-subtracted' 1080 PRINT : PRINT : PRINT 'Which type do you vish to plot 1090 INPUT I : PRINT : PRINT : PRINT "Choose froii:" 1180 IF (I<3) GOTO 1130 1110 PRINT : PRINT : PRINT "BAD ENTRY! Please try again.' 1120 PRINT : PRINT : PRINT : GOTO 1020 1130 ON I GOTO 1150,1160 1140 AF=".AVG" : K2=0 : FILES 'B:AVG' : GOTO 1170 1150 AF=".RET' : KZ=1 : FILES 'B:RET' : GOTO 1170 1160 AF='.CRP' : KZ'l : FILES 'B:«.CRP' 1170 PRINT : PRINT : N0--1 : PRINT "PRESS RETURN TO EXIT.• : PRINT 1180 LINE INPUT 'Filename -> ";A 1190 IF (A-"> GOTO 1210 1209 HO=MO-l : AN(«0)=A : GOTO 1130 133 1210 OPEN 'I',#l. 'BiSUBJECT. DAT" 1220 INPUT #1,A« : INPUT #1.A : INPUT #1.A : INPUT #1,A 1230 INPUT *1.N<0) : INPUT I I , N i l ) : INPUT #1,N(2> 1240 INPUT #1,3(0) : INPUT #1,3(1) : INPUT #1,3(2) 1250 CLOSE : N(1)=N(1)-N(0> : N(2)»N(2)-N(1) : HV»NO>\2 1260 NA»<HU)*M(2))\2 : PRINT : PRINT : PRINT 1270 FOR 1=0 TO 2 1280 B(I)=B(I)/N(I) 1290 NEXT I 1300 FOR I = KZ TO HQ 1310 SY(2)--13I : SY(3)=>1E*22 : SY(4)=-13I : SY(5)=SY(3) : SY(6)=-101 1320 JJ=-1 : PRINT TAB(10);"Reading ";AN(I> 1330 OPEN "I", #3, "3: "•ANdl'AF : PP=01 : 1)1*6 1340 INPUT #3,P : INPUT #3,0 : INPUT #3,R : K=CINT(P> : JJ=JJ*1 1350 IF (K<>«V) GOTO 1370 1360 SX(0)=P : SY(0)=O : SZ(0)=R 1373 IF (K<>HA) GOTO 1390 13S0 SX(1)=P : SY(l)=a : SZ(1)=R 1390 IF (X>HV OR 0<SY(2)> GOTO 1410 1400 SX(2)=P : SY(2)=0 : SZ(2)=R 1410 IF (K<0 OR X>N<0) OR 0>SY<3)> GOTO 1430 1423 SX(3)=P : SY(3)=0 : SZ(3)=R 1430 IF (K<HV OR X>HA OR Q<SY<4)> GOTO 1453 1443 SX(4)=P : SY(4)=Q : SZ(4)=R 1450 IF (K<N(1) OR K>N(2) OR 0>SY<5)> GOTO 1470 1463 SX(5)=P : SY(5)=0 : SZ(5)=R 1470 IF (K<HA OR 0<SY(6)  GOTO 1493 1483 SX(6)=P : SY(6)=0 : SZ(6)=R 1490 IF X>N(0) THEN GOTO 1530 ELSE SQ=B(0>-P 1533 IF (00<PP) GOTO 1523 1513 IN=I!1«1 : SX(IN)=QO : SY(I!l)=a : SZ(IN)=R : PP=PP-3! 1523 P'3.333-B<0)"P-11! : X=P : GOTO 1583 1533 IF (K>H(D) GOTO 1553 1540 P=3.333»3(1)»(P-N(3))*X : Y=P : GOTO 1583 1553 IF <X>N(2>> GOTO 1573 1563 P=3.333«B(2)-(P-H(11)»Y : Z=P : GOTO 1580 1573 P=.32"(P-N(2))-Z 1580 L(JJ)=CINT(P) 1593 IF Q<=.3364 THEN S(JJ>=3 ELSE N(JJ)=CINT(. 31831S»(L0G(0)«139.339)"2-212.586) 1630 IF EOF(3)<>0 THEN CLOSE ELSE GOTO 1343 1613 LPRINT CHRS(18);SPC(23)'NAHE : ";Ai!;" C PLOT ]• : LPRINT 1623 LPRINT : LPRINT : LPRINT SPCI23)CHRS(14);AN(I);CHRSI23); 1633 LPRINT "-minute PROFILE has";JJ-1;"data-seta,• 1643 LPRINT SPC(30)'stored in ";AN(I)»AF;" ." : LPRINT 1650 LPRINT CHRS(151 : LPRINT CHRS(27);CHRS(49) 1660 LPRINT SPC(25)'Vertical (LOG) Scale = X 10 ng/ml per 10 div.'; 1670 LPRINT SPCI7)"Horizontal Scale = 3 mm per 10 div.' 1680 FOR J=0 TO 40 1690 KK=40-J : P«XK/5I : LL=3 : LPRINT CHRSI13);CHRS(13);CHRS(9); 1739 IF ((P-FIX(P))=31) GOTO 1723 1710 LPRINT CHRS19);CHRS1124); : GOTO 1770 1720 IF (KX*5 OR KX=15 OR XX=25 OR KK-35) GOTO 1750 1730 IF J=0 THEN LPRINT 'ng/ml ••; ELSE LPRINT CHRS<9); '*"; 1740 . GOTO 1770 1750 LPRINT" 10" ;CHRS<27) ;CHRS(83) ;CHRS(0) ;XK\10; 1760 LPRINT CHRSI27);CHRS(34);" ••; 1770 FOR X=0 TO JJ 1780 IF (H(XX>KK OR L(X)>110) GOTO 1820 1790 IF L(K)=« THEN Mfl'l ELSE NH=L(K!-LL 1800 IF N«=3 THEN GOTO 1813 ELSE LPRINT SPC(MH-1)CHRS(249); 1813 LL=L(X) 1823 NEXT X 1833 NEXT J 1840 LPRINT : LPRINT CHRSI9);CHRSI9);• •; : LL=0 1850 «M=CINTI.3333»B(0)»N(0))»11 1863 FOR J=0 TO 10 1S70 IF (10»J)=«H THEN GOSUB 2380 ELSE LPRINT 1880 FOR X=0 TO 8 1890 XK=10-J*K-1 1900 IF XX = HI THEN GOSUB 2380 ELSE LPRINT 1910 NEXT K 1920 NEXT J 1933 IF (l««110 THEN LPRINT •'• ELSE LPRINT ••• 1940 LPRINT CHRS(9);CHRS(9);'-3';SPC(9)'R';SPC(19)'6";SPC(18);'12'; 1950 LPRINT SPC(18)"18';SPC(ia)"24';SPC(ia)'mm' : LPRINT : LPRINT 134 1968 LPRINT CHRS(27);CHRS(65);CHRS(12);CHRS(27);CHRS(50> 1973 LPRINT SPC(20)CHRS(27>;CHRS(45);CHRS(l);'Landmark'; 1983 LPRINT SPC(29)"Concentration -/- S. E. (ng/ml)"; 1993 LPRINT SPC(5)"Distance from RETINA (mm)"; CHRS(27>;CHRS<45);CHRS(3) 233a LPRINT SPC(2a>"Choroid-Retinal Peak";SPC(9); : X=B(0)"SX<2) : Y=SY(2) 2013 Z=SZ(2) : GOSUB 2253 : LPRINT "Mid-Vitreous";SPC(17); : X=B(0)'SX(0) 2823 Y»SY(0) : Z=SZ(0> : GOSUB 2258 : X=B(0)»SX(3> : Y=SY<3) : Z-SZI3) 283a LPRINT "Vitreous Lo»";SPC(17); : GOSUB 225a : P»B(0)»N(0> : Y=SY(4) 2a48 LPRINT "Lens Peak";SPC(28); : X=B( 1) MSXM)-N<0) >-P : Z=SZ(4> 2353 GOSUB 2258 : LPRINT "Anterior Lo»";SPC(17); : P'P-B(1)»(N(1)-N(8)) 286a X=B(2)'(SX<5)-K(1))-P : .Y«SY(3) : Z=SZ(5) : GOSUB 225a : Y=SY(1) 2878 LPRINT "Mid-Aqueous";SPC(18); : X=B(2)•<SX(1)-N<1))-P : Z'SZ(l) 2888 GOSUB 2258 : LPRINT "Anterior Peak";SPC(16); : X»B(2)"(SX(6)-N(1))*P 2893 Y=SY(S) : Z=SZ(6) : GOSUB 2253 : LPRINT 213a LPRINT SPC(33)CHRS(27);CHRS(45);CHRS(1); •DISTANCE from RETINA (mm)"; 2113 LPRINT SPCl18)"READING -7- S. E. (ng/ml)"; CHR9(27);CHRS<45);CHR3(8) 2128 FOR J=7 TO IM 2133 X=SX(J) : Y 'SYU) : Z=SZ(J) : GOSUB 2263 2148 NEXT J 2158 LPRINT CHRS(27);CHRS(58);CHRS(12) 2163 NEXT I 2173 PRINT : PRIHT "Enter ESC to exit from plot , • : PRINT 218a PRINT " ANY OTHER key to continue.'; : A=INPUTS(1) : PRINT 2198 IF (A<>CHRS(27)  GOTO 1B28 2238 ERASE L, M, N. SX. SY, SZ, 3 2218 LPRINT CHRS(27);CHRS(65);CHRS(12);CHRS(27);CHRS(53);CHRS(18); 2228 CLOSE : GOTO 158 2233 LPRINT SPC (63) "The data are stored in •; 224a RETURN 225a KX=0 : GOTO 2333 2268 KX*1 : LPRINT SPC (4a); : GOTO 228a 2278 LPRINT SPCl18); 2288 LPRINT USING •#*.#»*#»•;X; 2293 IF (XK=1) GOTO 2328 2330 LPRINT : LPRINT SPC(28); 231a RETURN 2328 LPRINT SPC(2a); 2338 LPRINT USING •##.#»###«*«*• ; Y ; 234a LPRINT ' •/-•; 2358 LPRINT USING * t # . #####****•;Z; 236a IF KK=1 THEN LPRINT ELSE GOTO 227a 2373 RETURN 2388 LPRINT : LL=LL-1 2390 MM«CINT(.3333»B(LL)«(H(LL)-N(LL-1) ) >»HH 2400 RETURN A - 1 4 D R A W . B A S 1000 REM Updated 253186 1313 DIM X(256),Y(256),N(4),Z(4) 1323 A"'-*--' : PRINT A : PRINT 'DRAW : PRINT A : 11=0 1030 ON ERROR GOTO 1970 1040 LPRINT CHR9<15);CHRS(27);CHRS(49);CHRS(27);CHRS(79); : MO*U 1060 FOR J=0 TO HQ 1070 OPEN •I'.fl."BtSUBJECT. DAT' : INPUT #1,AAA 1080 PRIHT : PRINT TAB(15);'Name : 1 ;AAA : PRINT : PRINT 1090 INPUT »1,A : INPUT #1,A : INPUT #1,A llOa INPUT #1,KX : INPUT #1,HS : INPUT fl.LL : INPUT #1,EE 1110 CLOSE : PRINT 'The files in Drive 3 are:' : PRINT 1 3 5 112a IF u=e> GOTO 115a 113a zia>»N(a>/KX : Z(i>»<N<i>-N<a)>/HH 1148 Z(2)'(H<2)-NU))/LL : GOTO 1173 1158 NI8XKX : N(1)'MM-KX : DD=KK/EE : JF=2!«DD-20 11S8 N<2XLL'N(1> : H<3XKK\2-15 : JG*4!«DD-20 117a LY'-l : MM=HH*XK : LL=LL*MM 118a FILES •Bf.'" : PRINT : PRINT 1198 PRINT TAB(5);"Enter . to change Diskette.' : PRINT 12aa PRINT TABI5);"Enter RETURN to exit." : PRINT : PRINT 1218 LINE INPUT 'Enter the COMPLETE Filename. ';AN(J> 1228 A-RIGHTS(AN(J),4) : PRINT : PRINT 1238 IF (AN(JX""> GOTO 1753 124a IF (AN(J)o'.*) GOTO 12S8 1253 J»J-1 t PRINT : GOTO 1728 126a IF <A=".AVG" OR A*'.RET" OR A»".CRP") GOTO 12B8 1278 PRINT "INCOMPLETE Entry. Please try again.' : GOTO 1188 128a OPEN "I",#2."3:"*AN(J) : NN=-1 1293 NN-NN'l : INPUT #2.X(NN) : INPUT #2, Y(NN) : INPUT #2,ZZ 138a IF XINNX-20! THEN NN=NN-1 1313 IF (NN<0 OR J=0) THEN GOTO 1358 ELSE ZZ'X(NN) 132a IF (ZZ<=LL AND ZZ>MM) THEN X(NN)=(ZZ-MM)'Z(2!'N(1) 1338 IF (ZZ<=MM AND ZZ>KK) THEN X(HH)=(ZZ-KX)'Z(1)'N<3) 134a IF (ZZ<=KX> THEN X<NNXZZ"Z(3) 1358 IF EOF(2)<>8 THEN CLOSE ELSE GOTO 1298 1363 IF (J<>9) GOTO 1388 1373 LPRINT : GOSUB 186a : LPRINT 1388 MC=8 : L=0 : H=0 : A*RIGHTS<STRS(J),1) 1398 IF N(2)>NH THEN XK*N(2>.10 ELSE KK=NN-18 1438 IF (XX>218) THEN KX=XK«-13 ELSE KK=210 1413 FOR K«0 TO XX 1423 IF (J<>3) GOTO 1538 1433 IF < <N<M)-20X>K> GOTO 1463 1448 IF (M=0 OR H»l) THEN LPRINT "L"; ELSE LPRINT "C; 1453 M'H-1 : GOTO 1518 1463 IF (XOJF) GOTO 1483 1473 LPRINT "2"; : GOTO 1513 14S3 IF (XoJG) GOTO 1538 1493 LPRINT '4'; : GOTO 1513 1538 IF X<>23 THEN GOTO 1528 ELSE LPRINT "R'; 1513 GOSUB 1863 : GOTO 1533 1523 IF RIGHTS(STRSCK), 1X"0* THEN LPRINT •<••; ELSE LPRINT CHRS1124); 1538 LPRINT CHRSI13);" "; 1548 IF <K<XN(3)-II) OR MC=1) GOTO 1568 1558 GOSUB 1838 : 11=11.2 156a FOR JJ=L TO NN 157a LL=CINT(X<J).23I)-X 1588 IF (LL<3) GOTO 1643 1598 IF (LL>8) GOTO 1658 1633 IF Y(JJ)>0! THEN ZZ=L0G( Y( JJ) ) ELSE ZZ=8I 1 6 i a HM=CINTC4.47-25«ZZ». 114»ZZ*2) 1628 IF MM<8 THEN HM=9 1638 IF MM<13  THEN LPRINT SPC<MN)A; 1543 L=L»l : LPRINT CHRSU3);" •; 165a NEXT JJ 1668 L«JJ : LPRINT 1678 NEXT X 1683 IF J'MO THEN GOTO 1758 ELSE PRINT 169a PRINT TAB(18);"PAUSE";J;"for PAPER Adjustment' 1738 PRINT TAB!12);'(Enter ESC to exit.) •; : A=INPUTS(1) 1718 IF A=CHRS(27) THEN GOTO 1753 ELSE RESET 1723 PRINT : PRINT TAB! 15);•INSERT NEW Diskette."; : A*INPUTS(1) 1738 PRINT : PRINT : PRIHT : PRINT : PRINT : LPRINT : RESET 1748 NEXT J 1753 LPRINT CHRSI27) ;CHRS( 65) ;CHRS( 12) ;CHRS< 27); CHRS( 53) ; 1763 LPRINT CHRSI27);CHRS(53) : PRINT : PRINT ; A=INK£YS 177a PRINT TAB(13);"Start a NEW plot? (N) •; : A»INPUTS(1) 1788 IF <A='Y' OR A="y"> THEN PRINT ELSE GOTO 1838 1798 PRINT : PRINT : PRINT : J=8 : GOTO 1848 1883 ON ERROR GOTO 8 1813 ERASE X.Y.N.Z 1828 CLOSE : GOTO 158 1838 LPRINT SPC188)"[";RIGHTS(STRS(J),1);•] =•> »;AN(J); 1848 LPRINT • Profile for ";AAA;CHRS<131 "j : MC»1 1853 RETURN issa FOR LV»I TO 12a 187a IF RIGHTS(STRS(LV),l)»"a" THEN LPRINT ELSE LPRINT iaaa NEXT LV 1890 LY-LY'l 136 1909 ON LY GOTO 1920,1930,1940.1950,1950,1960 1910 LPRINT • LOG*; : RETURN 1929 LPRINT • RETINA"; : RETURN 1930 LPRINT • 2 «•'! : RETURN 1940 LPRINT ' 4 mm'; : RETURN 1950 LPRINT • LENS"; : RETURN 1960 LPRINT " CORNEA"; : RETURN 1979 IF (ERL<>1280) GOTO 2019 1980 PRINT : PRINT CHRS(7);TA8(13);"This file does not exist." 1990 PRINT : PRINT TAB!12);"Try again. " : PRINT 2000 RESUME 1183 313 PINT CHRS(7>;"ERROR Code #";ERR;"in Line #";ERL 2323 RESUME NEXT A . 1 5 L U N D . B A S 1339 REM Updated 183236 1910 OIM X(130),YI130), 2(133), 0(100), P(2), 0(2), R(2) 1023 DIM S(2,2>,T(2>,U(2>,V(2,2),'rf(2>,P3(3),PL0(3) 1030 A="--~" : PRINT A : PRINT "LUND" : PRINT A 1340 XP=0 : ASO=CHRS<27>«CHRS(83)<-CHRS(0).'2'*CHRS(27)-CHRS(84) 1050 ON ERROR GOTO 3270 : NERRa0 1063 OPEN "I",#1."3:SUBJECT.DAT" : INPUT #1.AA 1073 INPUT #1.A : INPUT *1,A ; INPUT #1,A 1383 INPUT #1,X(0) : INPUT #1,Y(0) : INPUT #1,X(2) 1399 INPUT #1,Y(0) : INPUT #1,Y(1) : INPUT #1,Y(2) 1103 CLOSE : RAD=(Y(9)*Y(1)*Y(2))/2! : SC=Y(9)/X(9) 1119 OPEN "I", #2, "9:PLASMA. FIT" : INPUT #2,XX 1123 FOR 1=0 TO 5 1130 INPUT t2,P0 : J-I\2 1140 IF (1=9 OR 1=2 OR 1=4) THEN PB(J)=P9 1150 NEXT I 1169 CLOSE 1170 PRINT "The files in Drive B are ;• 1130 FILES "B:-.»" : PRINT : PRINT 1190 PRINT TAB(5);"Enter RETURN to exit." : PRINT 1200 PRINT TAB(5);"Enter . to CHANGE diskette." : PRINT : PRINT 1213 LINE INPUT "Enter the COMPLETE ????????.??? : ";AE : PRINT 1220 IF AE<>".• THEN GOTO 1250 ELSE PRINT 1230 RESET : PRINT 'Replace diskette NOW; enter ANY key vhen ready.'; 1240 A-INPUTS(l) : PRINT : RESET : GOTO 1360 1253 IF (AE<>"> GOTO 1290 1260 ERASE 0, P, 0, R, S, T, U, V, W, X, Y, 2, PB, PLO 1270 ON ERROR GOTO 0 : LPRINT CHRSI 12) ;CHRSI27);CHRS(58); 1280 CLOSE : GOTO 150 1290 GOSUB 2340 : PRINT : PRINT 'Enter the LEFT limit •; : INPUT PL 1300 PRINT 'Enter the RIGHT limit •; : INPUT RP : PRINT : PRINT 1310 IF (RP< = PL OR RP>RAD) THEN RP=RAD 1320 LPRINT SPC(19)'Analyzing file : •;CHR9(14);AE;CHRS(20); 1330 LPRINT • vithin t•;PL;',";RP;"1 mm' : LPRINT 1340 A=*INITIAL ESTIMATES of : PRINT 'Enter the ';A 1350 LPRINT SPC(18)A : A"'Permeability Coefficient P (cm/s> =• 1360 PRINT A; : INPUT P(0) : LPRINT SPC(20)A;P(0) 1370 A='Increment of P to be used (cm/a) =' : PRINT A; 1380 INPUT R(0) ; LPRINT SPCI21)A;R(0) : LPRINT SPC(22)"and' 1390 A='Diffusion coefficient D (cm'-ASO-'/s) =• 1403 PRINT 'Diffusion coefiicient D (cmA2/s) ="; 1413 INPUT P(l) : LPRINT SPC(22)A;PU) 1423 A»'Increment of D to be used (cm'-ASO-'/s) =' 1430 PRINT 'Increment of D to be used <cmA2/a> =•; 1440 INPUT R(l) : LPRINT SPCI20)A;R(1) : LPRINT : PRINT 1450 FOR 1=0 TO 1 1460 IF 1=0 THEN P9»600I ELSE P0-6309! 1479 P(I)=P(I)»P9 : R(I)=R(I)-P0 1483 NEXT I 137 1490 PRIST : II = -1 : PP=.001 : AAA=RIGHTS(AE,3 > 15B0 IF LEFTS<AAA.2X>"CV THEN P0=SC ELSE P0=1! l S i a OPEN "I",#3,"3:"*AE 1523 IF II>99 THEM GOTO 15S3 ELSE 11=11*1 153a INPUT »3,X(II) : X(II)=P0*X<II) : INPUT *3,Y<II) : INPUT #3, Z(II) 1543 IF X d l X a i THEN 11 = 11-1 1553 IF (E0F(3)=3) GOTO 1523 1563 CLOSE : J=II : II = -1 1573 FOR X=3 TO J 1583 IF (X<X)=>RP) GOTO 1623 1593 IF (Y(K)<=0! OR X(XXPL) GOTO 162a 1633 11=11*1 : X(I)=RAD-X(K) : Y(II)=Y(X) : ZUI)=Z<K)A2 1613 PRINT II*1;"> •;X(II);TAB(13);Y(II);"*/-•;Z(X) 1623 NEXT X 1633 PRINT : PRINT 7AB(10);II*1; 'points vere entered." : PRINT 1643 PRINT : LPRINT SPC(33)11*1;'points vere read. " 1653 FOR J=8 TO 1 1663 T(J)»3I 1673 FOR K=3 TO J 1683 S(J,K)=ai 1698 NEXT X 1730 NEXT J 1718 00=0! : J=LEN(AE)-4 : TT=VAL(LEFTS!AE,J)  1723 FOR 1=0 TO II 1730 GOSUB 2450 : 0(I)=WW : Z1 = Y(I)-WW : 00=QQ*Z1 A2/Z(I) 1740 FOR 1=0 TO 1 1753 U1 = P!J) : P(J)'U1*R!J) : GOSUB 2458 : U2=WW : PUXUl-RU) 1763 GOSUB 2453 : U( J) = . 5»(U2-WW)/R( J) : P(J)=U1 1773 NEXT J 1738 FOR J=8 TO 1 1790 TU)=T(J)*Z1"U(J)/ZCI) i a e a FOR X=0 TO J 1310 S(J,X)=S(J,X)*U<J)*U(X)/Z(I) 1823 NEXT X 1830 NEXT J 1840 NEXT I 1850 00=00/(11-2) : LPRINT SPC(30)" = > FIRST CHI";ASO.;" =';00 1860 FOR J=0 TO 1 1873 FOR X=3 TO J 1883 S(K,J)"S(J,K) 1893 NEXT X 1900 NEXT J 1918 FOR J=8 TO 1 1923 FOR X=3 TO 1 1938 V(J,X)=S(J,X)/SOR(S(J, J)»S(X,K)) 1943 NEXT X 1958 V(J.J)=11*PP 196a NEXT J 1978 GOSUB 2798 1988 FOR J=a TO 1 1998 W(J)=P(J) 2808 FOR K=a TO 1 2818 W(J)=W(J)*T(K)'V<J,K)/SOR(S(J, J)'S(K.X)) 2B2a NEXT K 2838 HEXT J 2843 GOSUB 2430 : SS=0l : LPRINT 2050 LPRINT SPC(14)CHRS(27);CHRS(45);CHRS(1);"S/No."; 2060 LPRINT SPC<8)"Reading*;SPC(ll)"Fit #1•;SPC(11)"Fit #2'; 2870 LPRINT CHRS(27);CHRS(45);CHRS(0) 2080 FOR 1=0 TO II 2090 LPRINT SPC115); : LPRINT USING "**";I; 2100 LPRINT SPCI10); : LPRINT USING "f#.#•##";Y(I) ; 2110 LPRINT SPC(18); : LPRINT USING •##.####•;0(I); 2123 GOSUB 2450 : LPRINT SPC110); : LPRINT USING "»#.####";W¥ 2130 0(I)=WW : SS=SS*(Y(I)-W*)'2/Z!I) 2140 NEXT I 2150 SS=SS/<II-2) : GOSUB 2400 : LPRINT : PLO(KP)=SS 2160 LPRINT SPC(30)"=> SECOND CHI";ASQ;" =";SS 2170 IF (00=>SS OR KP=2) GOTO 2190 2180 PP=10!*PP : XP=KP*1 : GOTO 1910 2190 FOR J=0 TO 1 2200 P<J)=W<J) : fl(J)»SOR(V(J, J)/S(J,J)) 2210 NEXT J 2220 PP=PP/131 : Pl=P(3)/630! : P2=O(0)/600! : LPRINT 2230 LPRINT SPCl12)'The BEST Fit Values are:' 2240 LPRINT SPC(13) 'PERHEA3ILITY COEFFICIENT ='; 2250 LPRINT USING •##.##*"A"A';P1; : LPRINT ' */-'; 2260 LPRINT USING •##.###AAAA';P2; : LPRINT 'cm/a .' 138 2270 Pl=P(l)/6000! : P2=Q(1)/6300! 223a LPRINT SPC(IS)'DIFFUSION COEFFICIENT = •; 2290 LPRINT USING •#•.###""• ;P1; : LPRINT • •/-•; 2300 LPRINT USING •#•.###' P2; : LPRINT 'cm';ASQ; Vs .' 2310 LPRINT SPCC20)"with Reduced CHI';ASQ;' =•;SS;CHRS<12) 2320 IF (SS>2!) GOTO 1910 ELSE GOTO 1170 2330 REN Subroutine to print page title. 2340 LPRINT CHRSU8) ;CHR9(27) ;CHRS(65) ;CHR9<12) ;CHRS(27) ;CHR9<53>; 23S0 LPRINT SPC(25)'NAHE : •; AA; • C LUND ]• 2360 LPRINT : LPRINT SPCI15)'Radius of retinal curvature CHR9I247);' 2370 LPRINT USING •»#.###«• ;RAD; : LPRINT • m  .• : LPRINT : LPRINT 2380 RETURN 2390 REM Subroutine far exchanging curve-fitting parameters, P(I). 2400 FOR J = 8 TO 1 2410 SWAP P(J),W(J> 2420 NEXT J 2430 RETURN 2440 REM Subroutine for calculating function, 0(1). 2450 P0=SQR(P(1>) : P1=P(0)/P0-P0/RAD : P6=P(0)"RAD/(X(I) *P0) 2460 P2=.56419»PS : P3=P6-P1 : P5 = . 5/P3 : P4=P5-(RAD-X(I) ) 2470 P5=P5-(RAD-X(I)  : P6=P3«EXP(2!•P1'P4) : P7=P3-£XP(2! »P1«P5) 2480 WW=8! : T2 = 3! : N=INT(TT) 2490 FOR JI*1 TO N 2500 TU=JI : GOSUB 2590 : Tl = l!/SCR < TU) 2510 0F=P2»T1-<EXP(-P4A2/TU)-EXP<-P5'2/TU) 2520 0D=EXP(TU«P1*2) : P3=P1/T1 : 01=P3-T1»P4 2530 GOSUB 2710 : 0F=aF-?6'QE»aD : 01=P3-T1»P5 2540 GOSUB 2713 : 0F=0F*?7»QE»0D : QF=TP'QF 2550 WW=WW.(T2*0F)/2! : T2-QF 2563 NEXT II 2570 RETURN 2583 REM Subroutine for calculating plasma value, TP. 2593 T1=TT-TU 2638 IF Tl=3! THEN TP=3! ELSE GOTO 2623 2513 RETURN 2623 ON XX GOTO 2643,2638,2650 2533 GOTO 2653 2643 T1=L0G(T1> : GOTO 2653 2653 T1=1!/T1 2663 TP=PB(3)'P3(1)»T1*PB(2)*T1*2 2673 RETURN 2683 TP=PB(8)AEXP<P9(1)AT1-PB(2)AT1A2) 2698 RETURN 2738 REM Subroutine for calculating error function, OE. 2713 01=01/58! : QE=31 : 02=1! : 04=01/2! 2720 FOR K=l TO 58 2733 SK=X : P0=<SK»Q1)A2 : P0=EXP(-P8) 2740 QE=QE»04M02*P3) : 02=P3 2758 NEXT X 2753 0E=l!-1.1283a»0E 2778 RETURN 2733 REM Subroutine to inverting matrix, V(I,J), and find det., 00 2798 DIM IA(2),JA(2) 2338 30=1! 2813 FOR X=8 TO 1 2820 01=0! 2830 FOR I=X TO 1 2840 FOR J=K TO 1 2850 IF (A3S(01)>ABS(V(I, J)>) GOTO 2870 2360 01=V(I,J) : IA(X)=I : JA(K)=J 2873 NEXT J 2880 NEXT I 2893 IF (Ol<>0!) GOTO 2910 2900 00=01 : GOTO 3250 2910 IF (lA(XXK) GOTO 2830 2923 IF (IA(K)=K) GOTO 2960 2930 FOR J=0 TO 1 2943 S1=V(K,J> : V(K,J)=V(IA(K),J) : V(IA(K),1)=-Sl 2950 NEXT J 2963 IF (JA(KXK) GOTO 2330 2970 IF (JA(X)=K) GOTO 3010 2980 FOR 1=0 TO 1 2990 S1 = V(I,K) : V(I,X)=V(I, JA(K)) : V( I, JA(X))«-Sl 3003 NEXT I 3018 FOR 1=0 TO 1 139 / IF I<>K THEN V(I,X)=-V(I, K)/Q1 HEXT I FOR 1=0 TO 1 FOR J=0 TO 1 IF (K>K AND J<>8) THEN V(I, J)=V(I, J)*V(I, K)"V<K,J) NEXT J HEXT I FOR J=0 TO 1 IF K>K THEN V(K. J)=V(K, Jl/01 NEXT J Y(K,X)=1!/01 : 00=Q0'Q1 NEXT X 3140 FOR J=0 TO 1 3150 K=l-J IF (IA(KX = K) GOTO 3200 FOR 1=0 TO 1 S1=V(I,X) : V(I,X)=-V(I, IA(K)) : V(I, IA(X))»S1 NEXT I IF (JA(KX = K) GOTO 3240 FOR 1=0 TO 1 S1=V(K,I) : V(K,I)=-V(JA(K),I) : V(JA(X),I)=S1 NEXT I NEXT J 3250 ERASE IA.JA 3260 RETURN 3270 IF NERR>10 THEN RESUME 1260 ELSE NERR=NERR*1 3230 PRINT CHRS(71;"ERROR Code #";EHR;"in Line #";ERL 3290 RESUME NEXT 3020 3030 3040 3050 3060 3070 3080 3090 3100 3110 3120 3130 3160 3170 3180 3190 3200 3210 3220 3220 3240 A . I S F i l - F o r m a t s a) Filenames Created by ??.DAT VITSCAN. BAS Format : (1) Pod "zero" position at beep (in DAS units). (2) Pod position (in DAS units). (3) R/H-Log Amp reading (in DAS units). (4) Repeat (2), (3) b) Filenames Created by ??.AVG ; ??.RET ; ??.CRP REDUCE. BAS Format : (1) Retina-zeroed Pod position (in DAS units). (2) Averaged Log Amp output (in*'). (3) Standard deviation ot (2) (in"). (4) Repeat (1),(2),(3) Filename SUBJECT.DAT Created by SCANMENU. BAS (1) Subject's name. (2) Subject's age. (3) Eye scanned. (4) Date oi scan. (5) Vitreal length (in DAS units). (6) Lens thickness (in DAS units). (7) Aqueous depth (in DAS units). (8) Ultra-sound vitreal length (in mm). (9) Ultra-sound lens thickness (in mm). (10) Ultra-sound aqueous depth (in ma). (11) Fluorescein injected (in ml). (12) Coaaents and observations. (13) Repeat ii necessary, (12). 140 d) Filename PLASMA. DAT Created by : PLASCAN. 3AS Format : (1) Background sanple time (aet » 0). (2) Background average (in DAS units). (3) Background standard deviation (in DAS unita). (4) Blood sampling time (in minutes p.i.). (5) Average fluorescence reading (in DAS units). (6) Standard deviation (in DAS units). (7) Repeat (4),(5),(6) e) Filename PLASMA. FIT Created by BLOOD.3AS Format : (1) Best-fit function's code number. (2) A coefficient. (3) Error of A. (4) 3 coefficient. (5) Error of 3. (6) C coefficient. (7) Error of C. (8) Measurement p.i. time (in minutes (9) Area up to (8) (in"1.min). (10) Error of (9) (in'. min). (11) Repeat (8),(9),(10) p.i.). f) Filenames Created by Format (1) (2) (3) (4) : ??.CV? : LUND.BAS Position from Retina (in mm). Average concentration (in"1). Standard deviation of (2) (in Repeat (1),(2),(3) 141 •7 S a m p l e C O N S E N T F O R M T H E UNIVERSITY OF BRITISH COLUMBIA VANCOUVER. B.C.. CANADA V5Z 3N9 F A C U L T Y O F M E D I C I N E DEPARTMENT OF OPHTHALMOLOCY I1S0 WILLOW STREET TELEPHONE >7S-M*l LOCAL 1431 CGNSEMT_FgRM lb? D?^ei2Bn!?Di of a Vitreous EIy9r9phgtgmeter Studying Alterations in the 31ggd Retinal Barrier Dr. I.S. Begg ; Dr. T. 'Cox; Dr. D.A. Balzarini; Mr Pang K. T. A study is being carried out to measure the abnormal leakage from retinal blood vessels which indicate cell damage in the early stages of disease prior to visible retinal changes. In most patients and normal sub jects, the measurements will be carried out following fluorescein angiography (photography) which is a customary diagnostic . procedure frequently used in clinical practice to visualize damage to the retinal structure, and to indicate disease severity which is useful in prognosis and in planning treatment. The measurement of fluorescein leakage is made in a follow-up procedure called VITREOUS FLUOROPHOTOMETRY at several time intervals fallowing the angiography. The vitreous cavity of the eye is scanned by the light beam <of an adapted clinical microscope,) which is directed into the eye through a contact lens. The scan of an eye takes about 30 seconds with minimal discomfort. After each scan, a finger-prick blood sample is taken to measure plasma fluorescein. There are no side effects related to these measurements. The injected fluorescein dye colours the skin slightly yellow for about 4 hours, and urine for about 24 hours. About 47. of patients experience brief spells of nausea. Serious allergic reactions are rare and have not been encountered in over some 20000 fluorescein angiograms done in the Department of Ophthalmology, UBC. The results of the tests and the personal medical records will be kept confidential by using a code number for each patient. The entire procedure, with repeated scanning, takes about 90 ainutes. The test may only be carried out with your signed consent and the understanding that you may decline to participate, or withdraw at any time during measurement without jeopardizing any routine medical treatment. CONSENT for the procedure and acknowledgement of receigt of § 992Y. of the consent fgrm̂ Signature of. Patient Date Signature of Witness 142 A P P E N D I X J3 M A T E R I A L - U S E D B . 1 E l e c t r o n i c s TYPE QTY REMARKS Capacitors (in nF) (Code Colour & Shape) 1000 2 CI - red box 30 3 C2 - yellow cylinder 10 2 C3 - grey/yellow box 0.15 1 C4 - violet-white 10000 3 C5 - blue/yellow bulb 100 1 C6 - orange Analogue-to-Digital Converter ADC0804LCN 1 ADC - black 20-pin Analogue Multiplexer IH6108 1 MUX - black 16-pin Operational Amplifiers TLD81ACP 2 OP - black 8-pin Sample-and-Hold IH5111IDE 2 S/H - violet-gold 16-pin Voltage Regulators LM336 2. 5 V 1 PI - black half-cylinder LM340T5 5.0 V 1 P2 - black w/a heat sink Resistors (in kOhms) 10 2 Rl - grey trim-pot 20 3 R2 - grey trim-pot 50 1 R3 - grey trim-pot 100 1 R4 - grey trim-pot 10 3 R5 - cylinder 1.8 1 R6 - cylinder 1 1 R7 - cylinder 0.18 1 R8 - cylinder 0.1 2 R9 cylinder Table 23. Electronic components of blue c i r c u i t board of the DAS. (See Figure 12.) 143 B. 2 Equipment 1. LOGARITHMIC AMPLIFIER Model E97 (built by the Electronic Shop, Department of Physics). 2. 15-V 100-mA POWER SUPPLY (designed by the Electronic Shop, Department of PhyBics). 3 . GAMMA SCIENTIFIC DIGITAL RADIOMETER Model DR-2. 4. GAMMA SCIENTIFIC PHOTOHULTIPLIER DETECTOR Model D-47A. 5. GAMMA SCIENTIFIC SCANNING PHOTOMETRIC.MICROSCOPE EYEPIECE Model 700-10-30X (Left ocular). 6. GAMMA SCIENTIFIC SCANNING PHOTOMETRIC MICROSCOPE EYEPIECE Model 700-10-34A (Right ocular vith fibre optic). 7. Modified NIKON ZOOM-PHOTO SLIT LAMP MICROSCOPE. 8. 2-V VOLTMETER (built by Stephen CLARK to monitor the lamp intensity). 9. A MODEL, EYE (built to required specifications by the Machine Shop, Department of Physics). 19. SPECTROTECH FILTERS: a. SE4 - excitor filter. b. SBS - barrier filter. 11. KEPCO POWER SUPPLY Model RMK 09-S (for slit lamp). 12. OSBORNE 1 64K Microcomputer. 13. OKIDATA MICROLINE 192 Dot Matrix Printer. 14. HAYES SHARTMODEfl 300 (for communicating with UBCHet). 15. FISHER ACCUMET Expanded Scale Research pH METER Model 320 (for preparing and measuring buffer pH). IS. IEC CENTRIFUGE Model CENTRA-4 (for spinning blood samples). 17. HAMILTON MICROLITER #702 Hicropipette (for measuring out plasma samples). 18. SONOMETRIC Ultrasonic Digital Biometric Ruler Model DBR 400 (for measuring intra-ocular lengths). 19. CONTACT LENSES: a. Piano PERHALENSR Hydrophilic Contact Lens (soft lens). b. COOPER VISION Plano-concave Hard Plastic Lena. 20. RED TIP HEPARINIZED MICRO-HEMATOCRIT CAPILLARY TUBES (for collecting blood saaples). 21. FUNDUSCEIN Fluorescein Sodiua 25X Aapoules (for intravenous injections). 22. BETHOCEL 2X Sterile (vater-based, highly viscous methyl-cellulose for holding the hard lens in place, and, to provide the optical continuity betveen interfaces). 23. CYCOLGYL' IX (Cyclopentolate Hydrochloride for dilating the pupil, and anaesthetizing the cornea). 24. M 4 3 Fluorescein Sodiua povder (for aaking calibration solutions). 25. MONOPAN 200 g Balance (for preparing calibration solutions). 144 B . 3 M o d e l E y e The cross-section of the model eye i s shovn below. The middle component i s made of lucite (or plexiglass). A l l other components are of aluminium. A l l dimensions are in mm. The assembly i s held together by four bolts <horizontal dash lines). 145 A P P E N D I X CZ C A L I B R A T I O M R E S U L - T S C . 1 P o d - D A S The least-squares f i t was to the straight line, Y = A + B * X by setting X = Osborne/DAS units, and Y - translation in mm . The gradient of the f i t was found to be 0.095919 0.000070 mm per DAS unit. The correlation coefficient was 0.999. C A L I B R A T I O N S O F S L I T L A M P T R A N S L A T I O N 40- 0SB0RNE/D.A.S. OUTPUT UNITS Figure 31. Calibration curve of the Pod-DAS. 146 DISPLACEMENT OUTPUT/DIFFERENCE 0 8 1 16 8 2 23 7 3 31 8 4 38 7 5 45 7 6 52 7 7 59 7 8 66 7 9 73 7 10 80 7 11 86 6 12 93 7 13 100 7 14 107 7 15 113 6 16 120 7 17 127 7 18 133 6 19 139 6 20 146 7 21 152 6 22 159 7 23 165 6 24 172 7 25 178 6 26 185 7 27 191 6 29 197 6 30 203 6 31 209 6 32 215 6 33 221 6 34 228 7 Table 24. Results of Pod calibrations. The data under the DISPLACEMENT column are multiples of 0.635 mm, i.e. the pod vas advanced 1/40'h of an inch at a time. 147 C . 2 L o g a r i t h m i c A m p l i f e r The least-squares f i t was to a logarithmic function, Y = A + B • log X where input voltages, X were in mV and output voltages, Y were in V. The correlation coefficient of this f i t was 0.997. INPUT VOLTAGE <mV) OUTPUT VOLTAGE (V) 1.2 0.40 1.5 0.50 2.1 0.65 2.4 0.70 2.5 0.73 4.5 1.00 5.7 1.11 7.9 1.25 11.2 1.40 17.5 1.61 23.4 1.73 32.2 1.87 46.8 2.04 57.5 2.12 70.0 2.22 99.8 2.37 109.0 2.42 180.0 2.63 224.0 2.73 343.0 2.79 400. 0 3.99 489.0 3.08 590.0 3.16 778.0 3.28 878.0 3.33 1000.0 3.38 1245.0 3.50 1574.0 3.60 1923.0 3.68 2230.0 3.75 3142.0 3.91 4085.0 4.02 5091.0 4.12 6000.0 4.20 7000.0 4.27 9000.0 4.38 Table 25. Results of Log Amp test. 148 0-j : : i • , ; l . . : • . . i l ... : : : : r i . : ; 1 10 1 0 0 1 0 0 0 1 0 0 0 0 INPUT VOLTAGE in mV Figure 32. Performance of the Log Amp. C.3 p H D e p e n d e n c e Water Buffer X Y X Y 6. 7 35.5 12.4 70 47 82 39 100 67 97 47.5 103 78 112 - 69.5 115 105 105 84 118 115 105 103 126 195 120 142 143 630 152 410 154 1830 172 870 179 3250 202 1220 178 3400 193 4850 219 8900 205 9258 228 Table 26. Water and buffer sample differences. 149 To test the pH dependence mentioned in Section 3.1, tvo sets of sample solutions vere prepared. The f i r s t set vas made vith deminera- lized d i s t i l l e d vater; the second set vas made vith the pH 7.4 Soren- sen's buffer solution. (See Section 4.1.) Note that the buffer set vas not the fi n a l calibration set given belov in Appendix C.4. The results of linear least-squares f i t , of concentrations, X ( to DAS outputs, Y (Osborne/DAS units), vere: Y = -7.354 • 24.449 • In X for vater samples, Y = 13.347 + 23.897 * In X for buffer sample. The correlation coefficients vere both 0.99. The slopes of the tvo f i t s vere not significantly different for a t(12-2) test (P = 5%). The "intercepts" vere significantly different (P = 0.5X) Hence, buffer samples produced higher outputs than vater. 3 0 0 - i 00 2 5 0 3 2 0 0 Q _ I— O CO 150 < S 1 0 0 ̂ cc O CD in O 50 X X A A A X Legend A BUFFER X WATER 0H 1— i 1— i i 1 1 1H I 1 — i i 1 1 1 M I 1 i i 1111M i ' . io 100 1000 10000 100000 SAMPLE CONCENTRATIONS in ng/ml Figure 33. pH dependence. 150 C . 4 R / M - L o g A m p - D A S F i t s to two functions were done. They were: a) X = A + B » l o g Y + C » <log Y)* , and, b) Y = A » exp< B • X • C • X" ) , where X = Osborne/DAS averaged r e s u l t , and, Y = sample concentration i n"1 . F i t <a) was inverted, and then compared to f i t (b) at every Qsborne/DAS output. F i t (a) was chosen because i t produced smaller deviations ( e s p e c i a l l y at the lower concentrations). The reduced chi-square was 2.765. The c o e f f i c i e n t s were A = 4.47039 +/- 1.65113 , B = 25.0257 +/- 0.73117 , and, C = 0.11444 +/- 0.06722 . The r e s u l t i n g equation on in v e r t i n g f i t (a), was Y = exp( AA + J X/CC * I F ) , where AA = -109.339 , BB = 11915.9 , and, CC = 0.11444 . RADIOMETER/LOG AMP CALIBRATIONS 30 100 130 200 230 0SH0RNE/0.A.S. OUTPUT UNITS Figure 34. Ca l i b r a t i o n curve f o r R/M-Log Amp-DAS. 151 X Y Fit-Y X 6357.47 231.34 1.07 0. 46 5487.33 227.51 0.92 0. 40 2143.54 204.83 -1.67 -0.82 719. 50 174.66 -0.61 -0. 35 313.79 152.80 -0.68 -0.45 143.01 134.50 -3.01 -2. 24 30.41 90. 26 1.00 1.11 30.41 90. 58 0.68 0.75 29.87 87.95 2.85 3.24 29.87 90.09 0. 71 0. 79 20.42 80.17 0.84 1.05 20. 42 80.25 0.76 0.94 13.78 70.08 0.81 1.16 13.78 70.37 0. 53 0.75 7.52 58.24 -2.81 -4.82 7. 52 57. 32 -1.89 -3. 30 Table 27. Calibration results of the R/M-Log Amp-DAS. Note that several scans were made at the lower concentrations. This was to "weight" the lower part of the calibration curve during curve-fitting. The reason for doing this was to "improve" the c a l i - bration curve in this region so that PMT and R/M noise during the calibration procedure would not adversely affect the curve-fitting process. 152 C . 5 A -fc. -h e n u a -fc. dL o n The effects of the attenuation was studied during the c a l i b r a - t ions of the R/M-Log Amp-DAS. Scans through the sample c e l l s holding various sample concentrations were made. Figure 36 shows two samples where attenuation effects were seen. The peak was the pos i t ion when the ent i re diamond was i n the so lu t ion . Attenuation effects were studied by re-zeroing the t rans l a t ion axis of each scan at the peak; then estimating the (negative) slope ( i f i t existed) . Refraction caused the displacement of the s l i t lamp to be d i f ferent from the diamond, l i k e the F-numbers ( in Figure 3). The conversion i s given below. Consider a p r o f i l e where s i g n i f i c a n t attenuation was observed. The negative, concentration-dependent slope, B(c), i n the semi-log p lot i s found (from Eq. 5, Section 2.2) to be 153 Figure 36. Attenuation in sample solutions. 154 • i c = A(c) - B(c) * X , where c and X are the concentration and translation respectively. A(c) i s a "constant". B(c) represents the decrease in log-concentra- tion per unit increase in the distance the probe focus "penetrates". When B(c) i s small, there i s l i t t l e attenuation. Hence, an estimate of the concentration at which B(c) i s small approximates the lower limit above which attenuation should be taken into consideration. The table below gives the results of the straight-line f i t s to the profiles and their respective correlation coefficients. c"1 B(c) mm"1 r 9493 0.289 0.994 6357 0.240 0. 999 5487 0.182 0.999 2143 0. 055 0.982 Table 28. Concentration and gradients of attenuated samples. A least-squares f i t to B<c) = AA + BB * c was made. The results were AA = 1.08*10-3 and BB = 3.25*10"5 ; r = 0.97 . Hence, for B(c) = 0, c = -33* ! From Eq. 5, for a mean aqueous chamber depth of d(=X) = 3.5mm, at cft = 1000"1, B(c") = 0.034, attenuation < 12% ; and, at c" = 100"1, B(c*) = 0.0043, attenuation < IV. . 155 C . £ 5 P e r f o r m a n c e D a t a CHARACTERISTIC/PARAMETER UNITS S l i t width 0. 1 mm S l i t height 2 mm Probe diameter 0. 45 mm Beam-Probe angle 16 o Lamp intensity monitor 141(1) mV F i l t e r overlap at 502.9nm 0.5 7. Max. s l i t lamp displacement 24. 5 mm Max. in vitro concentration 6000"1 AR in vitro 3 mm AR ratio in vivo on Normals 0.032 LLoD in vivo 4. 4"1 Sensitivity at 20"1 0. 77 ng. ml"* Osb"1 Sensitivity at 2000"1 85. 16 ng. ml"1Osb"1 EoM at 20"1 -0.03 EoM at 2000"1 0. 06 R in vivo 19 % No. of data-points in 10 s 800 pairs Table 29. Performance characteristics of the VF system. Sensitivity and EoM are in vitro estimates using the calibrations in Appendix C.4. Note: Sensitivity i s defined as gradient of calibration curve at given concentration. 156 . "7 o <>•> M o d e l E y e ? S«=san P r o f i l e C O N C E N F R f i r I O N P R O F I L E I N T H E M O D E L ETE C / R V I T R E O U S 9 . 0 3 E - 6 4=. 0 9 7 E - 7 AQUEOUS 1 . 1 3 E - 6 1 . 0 1 E - 7 I , , , , j _ 0 . 0 4 . 0 8 . 0 1 2 . 0 1 6 . 0 2 0 . 0 2 4 . 0 TRANSLATION IN MM Figure 37. Model eye sample profile. 157 A P P E N D I X O G L O S S A R Y BLOOD-RETINAL BARRIER - The barrier that separates the retinal neural tissues from the blood. BOLUS - A concentrated mass of pharmaceutical preparartion given intravenously for diagnostic purposes; a mass of scattering material. CHOROID - The network of small blood vessels immediately behind the retina. DIABETIC RETINOPATHY - A non-inflammatory disease of the retina due to diabetes vhich can lead to blindness. ENDOTHELIUM - The layer of epithelial c e l l s that lines the cavities of the heart and of the blood and lymph vessels, and the serous cavities of the body. EPITHELIUM - The covering of the internal and external surfaces of the body, including the lining of vessels and other small cavities. It consists of c e l l s joined by small amounts of cementing substances. EMMETROPIC - When rays entering the eye parallel to the optic axis are brought to focus exactly at the retina. FENESTRATED - Pierced with one or more openings. FUNDUS - That part of the back of the eye furthest from the pupil. HAEMOGLOBIN - The red pigment of the blood carried by the red blood c e l l s ; i s composed of globin (a protein) and haem <an iron compound); i s the means of oxygen transport. HAEMACYTOMETER - Instrument for counting blood corpuscles. HEMATOCRIT - Volume X of erythrocytes in whole blood. Originally applied to the apparatus of measurement. Many of the above descriptions were summarized from "The Penguin Medical Encyclopedia" by Peter Wingate (U.K. 1976), or, from the "Dorland's Illustrated Medical Dictionary", 26th ed., published by W.B. Saunders Co., (Toronto, 1981). 158 HOMEOSTASIS - The primary function of most organs; the processes of maintaining constant physical and chemical conditions vithin the body despite external changes. HYPERTENSION - High blood pressure secondary to specific disease, or, essential. INFILTRATE - To penetrate the interstices of a tissue. Material deposited by i n f i l t r a t i o n . JUNCTIONAL COMPLEX - The intercellular arrangement between the adjacent columnar epithelial cells. LESION - Any pathological or traumatic discontinuity of tissue or loss of function of a part. LUMINAL SURFACES - Surfaces of the cavity or channel within a tube or tubular organ. MACULA - Usually the retinal macula. In general, any area that i s distinguishable from i t s surrounding by colour, etc. MICROANEURYSM - Bulge at the weak point in the wall of an artery. MULTIPLE SCLEROSIS - A chronic central nervous system disease in which the nerve fibres lose their protective myelin sheaths and their a b i l i t y to conduct impulses. OPTIC DISC - The intraocular portion of the optic nerve formed by fibres converging from the retina and appearing as a pink to white disc. PERIPHLEBITIS - Inflammation of the tissues around a vein, or of the external coat of a vein. PLASMA - Sticky, pale amber liquid with faint, sickly smell; the medium in which v i t a l substances are transported to a l l body tissues; a solution in water of salts, proteins, glucose, etc. SEQUELA - Any lesion or affection following or caused by an attack of disease. ZONULAE OCCLUDENS - That portion of the junctional complex of columnar epithelial cells, just beneath the free surface, where the intercellular space i s obliterated. It extends completely around the c e l l perimeter. Also called •tight junctions". 159 A P P E N D I X E A B B R E V I A T I O N S U S E D A Operational Amplifier; Figure 12. ADC Analogue-to-Digital Converter; Figure 12. AR Axial Resolution. AVG Averaged data (by RET); filetype. BAB Blood-Aqueous Barrier. BRB Blood-Retinal Barrier. b attenuation or extinction coefficient; Eq. 5. CR Choroid-Retina(l). CRP Alignment by CR peaks; filetype. c concentration (; superscripts for various uses. D Diffusion constant; superscript. W means in water. DAS Data Acquisition System. DAT (Raw) Data (filetype). DRP Diabetic Retinopathy. DVM Digital VoltMeter; Figure 8. d Usually depth, or distance, or displacement. EoM Error of Measurement. H« Null Hypothesis. H" Alternative Hypothesis. hex hexadecimal (base-16) number(ing). LLoD Lower Limit of Detection. Log Amp Logarithmic Amplifier. MS Multiple Sclerosis. MUX MUltipleXer (analogue electronic switch); Figure 12. P Probability of Type I error - rejecting a true H'; Permeability constant; permeability index i f with superscript I. PMT PhotoMultiplier Tube. PR3 Penetration Ratio averaged about 3mm from retina. PR3» Penetration Ratio at 3mm from retina, p. i . post-injection. 160 R Reproducibility; Eq. 27. RET Alignment by visually pin-pointed RETinae; filetype. RPE Retinal Pigment Epithelium. R/M RadioMeter. S.D. Standard Deviation. S/H Sample-and-Hold electronic chip; Figure 12. t ( 1> First blood sampling time. VF Vitreous Fluorophotometry (Fluorophotometric). 161 R E F E R E N C E S 1. D.M. Maurice: "A New Objective Fluorophotometer", Exp. Eye Res., 2, 33 (1963). 2. S.R. Waltman, H.E. Kaufman: "A New Objective S l i t Lamp Fluorophotometer", Inves. Ophth., 9, 247 (1970). 3. J.G. Cunha-Vaz: "Sites and Functions of the Blood-Retinal Barriers", in The_Blggd2Retinal_Barriers, J.G. Cunha-Vaz, ed., Plenum Press, N.Y. (1980). 4. J.G. Cunha-Vaz, J.R. Faria de Abreu, A.J. Campos, G. M. Figo: "Early Breakdown of the Blood-Retinal Barrier in Diabetes", Brit. J. Ophth., 59, 649 (1975). 5. T.C. Prager, H.H. Chu, C. A. Garcia, R.E. Anderson, J.B. Field, E.A. Orzeck, J.P. Comstock: "The Use of Vitreous Fluorophotometry to Distinguish between Diabetics with and without Observable Retinopathy: Effect of Vitreous Abnormalities on the Measurement", Inves. Ophth. Vis. Sci., 24, 57 (1983). 6. J.G. Cunha-Vaz, R. Zeimer, P. Mahlberg, H. Tessler: "Vitreous Fluorophotometry Studies in Pars Planitis", in Retinal_Diseases, Grune & Stratton, Inc., N.Y. (1985), pp. 53-7. 7. J.G. Cunha-Vaz, CC. Mota, E.C. Leite, J.R. Abreu, M. A. Ruas: "Effect of Sulindac on the Permeability of the Blood- Retinal Barrier in Early Diabetic Retinopathy", Arch. Ophth., 103, 1307 (1985). 8. T. Engell, 0.A. Jensen, L. Klinken: "Periphlebitis retinae in multiple sclerosis. A histopathological study of two cases", Acta Ophthal., 63, 83 (1985). 9. A.C. Arnold, J.S. Pepose, R.S. Hepler, R. Y. Foos: "Retinal Periphlebitis and Retinitis in Multiple Sclerosis: 1. Pathologic Characteristics", Ophthal., 91, 255 (1984). 10. T. Engell, P.K. Andersen: "The Frequency of Periphlebitis Retinae in Multiple Sclerosis", Acta Neurol. Scand., 65, 601 (1982). 11. B.R. Younge: "Fluorescein Angiograpghy and Retinal Venous Sheathing in Multiple Sclerosis", Can. J. Ophth., 11, 31 (1976). 162 12. C.R. Bamford, J.P. Ganley, W.A. Sibley, J.F. Laguna: "Uveitis, Perivenous Sheathing and Multiple Sclerosis", Neurology, 28, 119 (1978). 13. T. Engell, A. Hudberg, A. Uhrenholdt: "Multiple Sclerosis: Periphlebitis Retinalis et Cerebro-Spinalis - A Correlation between Periphlebitis Retinalis and Abnormal Technetium Brain Scintigraphy", Acta Neurol. Scand., 69, 293 (1984). 14. R.C. Zeimer, N.P. Blair, J.G. Cunha-Vaz: "Vitreous Fluorophotometry for C l i n i c a l Research: I. Description and Evaluation of a New Fluorophotometer", Arch. Ophth., 101, 1753 (1983). 15. J.R. Gray, M.A. Hosier, B.M. Ishimoto: "Optimized Protocol for Fluorotron" Master", presented at the International Meeting on Ocular Fluorophotometry, Paris, France (1982). 16. T.C. Prager, D.J. Wilson, G.D. Avery, J.H. Merritt, CA. Garcia, G. Hopen, R. E. Anderson: "Vitreous Fluorophotometry: Identification of Sources of Variability", Inves. Ophth. Vis. Sci., 21, 854 (1981). 17. J.G. Cunha-Vaz: "Vitreous Fluorophotometry: Techniques, Methodology and Clinical Applications", handout from Amer. Acad. Ophth. instruction course. 18. R.C. Zeimer, J.G. Cunha-Vaz, M.E. Johnson: "Studies on the Technique of Vitreous Fluorophotometry", Inves. Ophth. Vis. Sci., 22, 668 (1982). 19. J.A. van Best, L. Vrij, J.A. Oosterhuis: "Lens Transmission of Blue-Green Light in Diabetic Patients as Measured by Autofluorophotometry", Inves. Ophth. Vis. Sci., 26, 532 (1985). 20. J.G. Cunha-Vaz, D.M. Maurice: "The Active Transport of Fluorescein by the Retinal Vessels and the Retina", J. Physiol., 191, 467 (1967). 21. N.P. Blair, R.C. Zeimer, M.M. Rusin, J.G. Cunha-Vaz: "Outward Transport of Fluorescein from the Vitreous in Normal Human Subjects", Arch. Ophth., 101, 1117 (1983). 22. B. Krogsaa, H. Lund-Andersen, J. Mehlsen, L. Sestoft, J. Larsen: "The Blood-Retinal Barrier Permeability in Diabetic Patients", Acta Ophth., 59, 689 (1981). 163 23. P.S. Chahal, P.J. Chovienczyk, E. M. Kohner: "Measurement of Blood-Retinal Permeability: A Reproducibility Study in Normal Eyes", Inves. Ophth. Vis. Sci., 26, 977 (1985). 24. J.Z. Winkelman, R.J. Zappia, A.J. Gay: "Human Arm to Retina Circulation Time", Arch. Ophth., 86, 626 (1971). 25. G.N. Wise, C.T. Dollery, P. Henkind: Ihe_Retinal QilS5=i§ii°Q- Harper & Row, N. Y. (1971), pp. 146-9. 26. H. Lund-Andersen, B. Krogsaa, J. Larsen, E. Scherfig, L. Sestoft: "Fluorophotometric Evaluation of the Vitreous Body in the Development of Diabetic Retinopathy", in Retinal_Diseases, J.G. Cunha-Vaz, ed., Grune & Stratton, Inc., N.Y. (1985), pp. 9-13. 27. J.G. Cunha-Vaz, R.C. Zeimer, N.P. Blair: "Ocular Fuorophotometry", I.S.O.F. Quarterly, 1, 6 (April 1984). 28. B.P. Conway: "An Analysis of Vitreous Fluorophotometry", i n i§iiQ§i_Di§§§§§§, J. G. Cunha-Vaz, ed., Grune & Stratton, Inc., N.Y. (1985), pp. 59-66. 29. R.C. Zeimer, N.P. Blair, J.G. Cunha-Vaz: "Vitreous Fluorophotometry for Cl i n i c a l Research: II. Method of Data Acquisition and Processing", Arch. Ophth., 101, 1757 (1983). 30. B. Krogsaa, H. Fledelius, J. Larsen, H. Lund-Andersen: "Photometric Oculometry. I.: An Analysis of the Optical Principles in Slit-Lamp Fluorophotometry", Acta Ophth., 62, 274 (1984). 31. R.C. Zeimer, J.G. Cunha-Vaz: "Evaluation and Comparison of Commercial Vitreous Fluorophotometers", Inves. Ophth. Vis. Sci., 21, 865 (1981). 32. B.M. Ishimoto, R.C. Zeimer, J.G. Cunha-Vaz, M.M. Rusin: "Data Analysis and Presentation in Vitreous Fluorophotometry", I.S.O.F. Quarterly, 2, 1 (Sept. 1985). 33. J. Larsen. H. Lund-Andersen, B. Krogsaa: "Transient Transport Across the Blood-Retina Barrier", Bull. Math. Biol., 45, 749 (1983). 34. P.R. Bevington: Data_Reduction_and_Errgr_An £he_Physical_Sciences, MacGraw-Hill, N.Y. (1969), pp. 235-42. 164 35. S-E. Bursell, F. C. Delori, A. Yoshida, J.S. Parker, G.D. Collas, J.W. McMeel: "Vitreous Fluorophotometric Evaluation of Diabetics",' Inves. Ophth. Vis. Sci., 25, 703 (1984). 36. G. T. Feke, R. Zuckerman, G.J. Green, J.J. Weiter: "Response of Human Retinal Blood Flow to Light and Dark", Inves. Ophth. Vis. Sci., 24, 136 (1983). 37. R.F. Brubaker, R.L. Coakes: "Use of a Xenon Flash Tube as the Excitation Source in a New Slit-Lamp Fluorophotometer", Amer. J. Ophth., 86, 474 (1978). 38. E.M. Kohner, A.R. Alderson: "Vitreous Fluorophotometry", Trans. Ophth. Soc. U.K., 101, 446 (1981). 39. D.M. Maurice: "The Use of Fluorescein in Ophthalmological Research", Inves. Ophth., 6, 464 (1967). 40. D.W. H i l l : "The Techniques of Fluorescence Retinography", in Ih§_yiiii§!D_d§9^§Q2i§_5§Qi§Q§IZ_5yn!E2§iy!D_2D_Ih§ Qsyl§E_Circulatign_in_Health_an J.S. Cant, ed., Kimpton, London, U.K. (1969), p. 53. 41. F. Delori, I. Ben-Sira, C. Trempe: "Fluorescein Angiography with an Optimized F i l t e r Combination", Am. J. Ophth., 82, 559 (1976). 42. N.P. Blair, M.M. Rusin, E. Shakin: "The Effect of pH on the Transfer of Fluorescein across the Blood- Retinal Barrier", Inves. Ophth. Vis. Sci., 26, 1133 (1985). 43. P. Horowitz, W. H i l l : Ihe_Art_of.Electronics, Cambridge U. Press, N.Y. (1980). 44. T. Hogan, M. Iannamico: QSBORNE_l_Technical_Manual, Osborne Computer Corp., CA. (1982), Part No. 2F00153-01. 45. T. Hogan, M. Iannamico: QSBORNE_l_User^s_Reference_Guide, Osborne Computer Corp., CA. (1981). 46. A.R. Miller: §080/Z280_Assembly_Language2_Te Imprgved_Prggrarnming, Wiley, N. Y. (1981). 47. R.C. Zeimer, N.P. Blair, J.G. Cunha-Vaz: "Pharmacokinetic Interpretation of Vitreous Fluorophotometry", Inves. Ophth. Vis. Sci., 24, 1374 (1983). 165 48. G.F. Carrier, C.E. Pearson: P§rtial_Differential_Eguatigns Jhfgr^_and_Technigue, Academic Press, Inc. , N.Y. (1976), pp.257-81. 49. W.A. Ray, D.M. O'Day: "Statistical Analysis of Multi-Eye Data in Ophthalmic Research", Inves. Ophth. Vis. Sci., 26, 1186 (1985). 50. S.N. Markowitz, J.D. Morin: "The Ratio of Lens Thickness to Axial Length for Biometric Standardization in Angle-Closure Glaucoma", Amer. J. Ophth., 99, 400 (1985). 51. G.W. Snedecor, W.G. Cochran: Statistigal_Methods, 6th ed., The Iowa State U. Press, IA. (1967), 52. R.E. Walpole: Intrgductign_to_Statistics, 3rd ed., MacMillan, N.Y. (1982). 53. Y. Ogura, Y. Tsukahara, I. Saito, T. Kondo: "Estimation of the Permeability of the Blood-Retinal Barrier in Normal Individuals", Inves. Ophth. Vis. Sci., 26, 969 (1985). 54. H. Lund-Andersen, B. Krogsaa, M. la Cour, J. Larsen: "Quantitative Vitreous Fluorophotometry Applying a Mathematical Model of the Eye", Inves. Ophth. Vis. Sci., 26, 698 (1985). 166


Citation Scheme:


Usage Statistics

Country Views Downloads
China 3 0
Japan 1 0
City Views Downloads
Beijing 3 0
Tokyo 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}


Share to:


Related Items