Photogrammetry of the optic disc in glaucoma and ocular hypertension with simultaneous stereo photography*

Transcription

1 Photogrammetry of the optic disc in glaucoma and ocular hypertension with simultaneous stereo photography* Chris A. Johnson, John L. Keltner, Marijane A. Krohn, and Gerald L. Portney Stereophotogrammetric evaluations of the optic cup were performed for normal, ocular hypertensive, and glaucomatous eyes. Average volume, area, and depth measurements were progressively larger from normal to ocular hypertensive to glaucomatous eyes, although the distributions of individual values exhibited considerable overlap among the three groups. Similar results were obtained for volume, area, and depth asymmetry between each pair of eyes. None of these measurements was able to distinguish accurately between normal and glaucomatous optic cups. However, normal eyes showed a high correlation (r = +0.85) between area and depth of the optic cup, whereas this area/depth relationship was reduced in ocular hypertensives (r = +0.63) and completely broke down for glaucomatous eyes (r = +0.04). Approximately 89% of the glaucomatous eyes and 47% of the ocular hypertensive eyes were beyond the range of normal area/depth correlation values. These findings represent an improvement over most previous attempts to quantitatively differentiate between normal and glaucomatous eyes on the basis of optic disc measurements alone, and support the hypothesis that optic disc damage usually precedes visual field loss in glaucoma. With further technical refinements such as computer image processing, stereophotogrammetry of the optic cup may become a valuable differential diagnostic technique for glaucoma. Key words: glaucoma, ocular hypertension, stereophotogrammetry, optic disc evaluation, glaucomatous optic disc cupping R ecent studies have shown that both ste reo and nonstereo 16 " 20 photogrammetric From the Department of Ophthalmology (C. A. J., M. A. K., and G. L. P.) and the Departments of Ophthalmology, Neurology, and Neurological Surgery (J. L. K.), School of Medicine, University of California, Davis. Supported in part by National Eye Institute Research grant EY (JLK) and National Eye Institute Academic Investigator Award K07-EY00095 (CAJ). Submitted for publication July 24, Reprint requests: Chris A. Johnson, Ph.D., Department of Ophthalmology, University of California School of Medicine, Davis, Calif This paper is dedicated to the memory of Gerald L. Portney, M.D., who died on May 23, The present investigation represents a continuation of his work on photogrammetric analysis of the optic disc as an objective diagnostic procedure for glaucoma. processing of fundus photographs can produce accurate measurements of optic cup topography. Since this technique provides an objective, quantitative evaluation of the optic nerve head, it represents a potentially important diagnostic tool for glaucoma and other optic nerve diseases. Several investigators 1 " 3 ' 17 have reported volume, area, and depth characteristics of the optic cup in normal populations, based upon photogrammetric analysis. Their results indicate that in normal subjects, these optic cup measurements are highly symmetrical between the two eyes. Portney 2 ' 3 and Holm et al. 17 found that the average volume, area, and depth of the optic cup are larger in glaucomatous eyes than in normal eyes. However, the distributions of individual values exhibit /79/ $01.20/ Assoc. for Res. in Vis. and Ophthal., Inc.

2 Volume 18 Number 12 Stereophotogrammetry of optic disc 1253 Fig. 1. Schematic representation of the revised static-kinetic visual field test procedure, consisting of 2 kinetic isopters beyond 30 and static perimetry along 10 meridians within the central 30. considerable overlap between the normal and glaucoma populations, thereby limiting the diagnostic efficacy of these measurements. Comparison of optic cup volumes in unilateral and bilateral glaucoma patients revealed significant asymmetry between eyes, in contrast to the results for normal subjects. Portney 2 ' 3 concluded that optic cup volume asymmetry between eyes was the most sensitive photogrammetric measurement for differentiating normal subjects from glaucoma patients. The purpose of the present investigation was to refine existing photogrammetric methods of evaluating the optic cup and thereby to increase the diagnostic utility of this technique for glaucoma. To enhance the early detection of glaucomatous damage, particular attention was directed toward patients at high risk (elevated intraocular pressure with no demonstrable visual field loss) and patients with early glaucomatous visual field defects. Materials and methods All participants in the study received a complete eye examination, including refraction, Goldmann applanarion tonometry, ophthalmoscopy, biomicroscopy, stereo disc photography, and kinetic and static perimetry. Patients were excluded if good quality stereo disc photographs could not be obtained or if they were unable to perform reliably on visual field testing. One hundred sixtyfour eyes of 84 patients were selected for photogrammetric evaluation according to the following criteria: (1) average intraocular pressures less than 20 mm Hg with normal optic discs and visual fields (normals, N = 40 eyes), (2) average intraocular pressures greater than 25 mm Hg with normal visual fields (ocular hypertensives, N = 106 eyes), (3) early glaucomatous optic disc cupping with early visualfielddefects (glaucoma, N = 18 eyes). Following this initial classification, all stereo

3 1254 Johnson et al. Invest. Ophthalmol. Visual Sci. December T RXrS FOB 90 DEGREE PLRNf MO HOD Fig. 2. Contour map and cross-sectional plots of a representative glaucomatous optic disc.

4 Volume 18 Nutnlxr 12 Stereophotogrammetry of optic disc 1255 disc photographs were manually processed by a photogrammetric engineer according to standard techniques. 21 The photogrammetry data were converted to digital form and analyzed by computer to determine depth, area, volume, and other pertinent geometric measurements of the optic cup. Visual field testing. A combination of extensive static and kinetic perimetry was used to evaluate visual fields. Kinetic perimetry (with either the Goldmann or Tubingen perimeters) consisted of at least 2 isopters beyond 30 radius, 3 isopters within 30 radius, and numerous spot checks between isopters within the central 40 radius, as previously described by Portney and Krohn. 22 Static perimetry (with the Tubingen perimeter) consisted of threshold determinations along the 45, 135, 225, and 315 meridians in 1 intervals out to 20 radius and 2 intervals between 20 and 30 radius. Additional static perimetry was performed along both radial (meridian) and circular paths which intersected the length and width of visual field defects plotted by kinetic perimetry. Approximately two thirds of the eyes were evaluated with this procedure, which required about 1 hr per eye. The remaining one third of the eyes were tested according to the method described below. In view of recent findings 22 which show that static perimetry is more sensitive and reliable than kinetic perimetry for detection of early glaucomatous visual field defects, 42 eyes were tested with the revised procedure illustrated in Fig. 1 (shown for a right eye). This visual field examination consisted of at least 2 isopters beyond 30 radius (kinetic testing) and static perimetry along four meridians (45, 135, 225, 315 ) in 2 intervals and six intermediate meridians in 2.5 intervals across the central 30 radius of the visual field. Additional circular and radial (meridian) static perimetry was performed when it was necessary to further define specific portions of the visual field. All determinations were conducted on the Tubingen perimeter and required approximately 45 to 50 min per eye. This revised procedure thus provided the dual advantages of greater time efficiency and more effective detection of early glaucomatous visual field defects. Specific criteria were established to define the presence or absence of visual field defects, based upon extensive previous experience and existing guidelines developed by other investigators. 23 ' 24 According to our standards, areas of visual field loss had to be at least (1) 5 by 5 in size and 0.5 log unit of luminance (apostilbs) deep or (2) 3 by 3 in NORMAL n = 40 eyes n = 106 eyes GLAUCOMA OPTIC CUP VOLUME (mm 3 ) n= 18 eyes Fig. 3. Frequency distributions of optic cup volume for each of the three patient groups. size and 0.7 log unit of luminance (apostilbs) deep. These values are generally beyond the range of response variability in normal subjects. Visual field defects were also verified on at least two static profiles from adjacent meridians, or a combination of one meridian and one circular static profile. Additional testing with the' Fieldmaster automated perimeter was conducted as a confirmatory procedure. 25 Patients with questionable or borderline results were tested a second time. To minimize the influence of blur on perimetric determinations within the central 30 radius, patients were given an appropriate refractive correction for distance plus an added near correction for age. Accuracy of this correction was checked at the perimeter bowl by subjective refraction techniques. 26 Fixation was carefully monitored throughout each testing session, and response variability was determined at several visual field locations to ensure that threshold variations were less than 0.4 log unit of luminance (apostilbs) during perimetric testing. Classification of optic discs. Optic discs were classified according to whether they appeared to be normal or exhibited early glaucomatous damage. Portney's cone-cylinder-hemisphere category system 2 " 6 (based upon Elschnigs types I to IV categories) was used to classify normal optic discs. Judgments of glaucomatous damage to the optic

5 1256 Johnson et al. Invest. Ophthalmol. Visual Sci. December 1979 NORMAL n = 40eyes NORMAL n = 40 eyes n= 106 eyes n = IO6 eyes GLAUCOMA n= 18 eyes GLAUCOMA n= 18 eyes OPTIC CUP ORIFICE AREA (mm 2 ) Fig. 4. Frequency distributions of optic cup area for each of the three patient groups. disc were based upon Kronfeld's criteria of central deep atrophy, mottled or "moth-eaten" appearance of the base of the optic cup, upward or downward extension, marginal excavation and narrow nerve rims. 2 " 6 ' 27> 28 Other indicators of glaucomatous optic cupping 29 " 36 (e.g., vertical ovality of the cup, notching of the nervefiberrim, asymmetry of the cup between eyes, pallor of the optic cup) were also employed to distinguish between normal and glaucomatous optic cups. As with the evaluation of visual field results, conservative criteria were used for classifying normal vs. glaucomatous optic cups. The investigators agreed upon nearly all optic disc classifications, and repeated evaluations at periodic intervals showed a high degree of consistency. Stereophotography and photogrammetric analysis. Simultaneous stereo disc photographs were obtained with the Donaldson stereo fundus camera. 37 Since Kodak photomicrography film (ASA 16) produced fundus photographs which were underexposed for most eyes, it became necessary to use Kodachrome 25 film (ASA 25). Frisen 38 has reported that the resolution characteristics of Kodak photomicrography film and Kodachrome 25 are approximately equivalent for luminance conditions similar to those employed in this study OPTIC CUP DEPTH (mm) Fig. 5. Frequency distributions of optic cup depth for each of the three patient groups. Preliminary investigations revealed that photogrammetric measurements performed on both types offilmproduced comparable results. 39 Several of the optical specifications provided with our Donaldson stereo fundus camera were in error. The stereo base (reported as 5.75 mm in the specifications provided with the camera) was found to be 2.87 mm by dividing the center-tocenter distance of the photographic plate by the magnification of the camera's objective lens (2x). In addition the 3.25X and 4.25X magnification settings were calculated to be 2.69X and 3.22X, respectively. This was determined by photographing a model eye containing a calibrated line grating and calculating the ratio of image size to object size. With these revised values, photogrammetric measurements of Donaldson stereo photographs were consistent with similar determinations performed on Zeiss stereo photographs (with the Allen rotary prism) of the same optic disc. Each pair of stereo disc photographs was processed by a photogrammetric engineer using a Wild A-10 photogrammetric plotting instrument (magnification, 6.55X; stereo base, 200 mm). Previous studies 6 have reported that the variability of this procedure (using simultaneous stereo photographs) is approximately 10%. Digital x, y,

6 Volume 18 Number 12 Stereophotogrammetry of optic disc 1257 Table I. Optic cup volume, area, and depth measurements in normal, ocular hypertensive, and glaucomatous eyes Volume (mm 3 ): Mean S.D. Range Area (mm 2 ): Mean S.D. Range Depth (mm): Mean S.D. Range Normal (N = 40 eyes) Ocular hypertensive (N = 106 eyes) Glaucoma (N = 18 eyes) and z coordinates were established for 400 to 700 data points on each optic disc (according to an arbitrary "count" scale used by the photogrammetric plotting device) and then transferred to punch cards for subsequent computer processing. Horizontal (x) and vertical (y) counts were converted to millimeters according to the photogrammetric plotter-to-camera and camera-to-eye magnification ratios, and depth (z) counts (in arbitrary units) were converted to millimeters by the following equation: AZ = R b,) AP where AZ is the change in depth (in mm); Mi is the magnification of the Donaldson fundus camera (2.69x or 3.22x) ; b! is the stereo base of the Donaldson fundus camera (2.87 mm); M 2 is the magnification of the Wild A-10 photogrammetric plotter (6.55X); b 2 is the stereo base of the Wild A-10 photogrammetric plotter (200 mm); R is the ratio of the magnification and stereo base of the Donaldson stereo fundus camera to the magnification and stereo base of the Wild A-10 photogrammetric plotter, i.e., R = (M, b,)/(m 2 b 2 ); f e is the distance between the posterior principal point of the eye and the fundus (17.2 mm); and AP is the horizontal parallaxes measured on the photographs. Volume, area, and depth measurements were then calculated by the computer. A significant problem in determining optic cup volume is specifying the top of the optic cup. Since therimof the optic cup has considerable topographic variation about its circumference, volume measurements may show substantial variation as a result of different criteria employed to establish the top of the optic cup. Comparisons among different optic cups exacerbate the problem because of marked variation in the overall configuration of the cup from one eye to another. In the present study, the top of the optic cup was defined by the lowest point on the rim of the cup orifice. This point was determined by a computer algorithm which began at the bottom of the optic cup (greatest depth, or largest z value) and checked the x, y coordinates around the cup for successively smaller depth values. A "spill-over" at the lowest point of the optic cup rim was thereby defined by an abrupt, extremely large change in x, y coordinates at some position along the cup orifice. This permitted a standard, objective criterion to be applied to all optic cups undergoing photogrammetric analysis. An illustrative example of a contour map and cross-sectional graphical plots derived from photogrammetric processing of stereo disc photographs is shown in Fig. 2. Results Figs. 3 to 5 present the frequency distributions of volume, area, and depth of the optic cup, respectively, for normal, ocular hypertensive, and glaucomatous eyes. Each of these optic cup measurements exhibited differences in the distribution of values among various patient groups. Small optic cup volumes, areas, and depths occurred more frequently in normals, followed by ocular hypertensives. Patients with glaucomatous optic disc cupping and visual field loss showed a higher percentage of optic cups with large volume, area, and depth values. A chi-square analysis of these data revealed

7 1258 Johnson et al. Invest. Ophthalmol. Visual Sci. December I ii NORMAL (Bilateral) n= 20 Patients 50-i NORMAL (Bilateral) n = 20 Patients l i i i r o (Bilateral) n= 45 Patients 50- GLAUCOMA (Unilateral or Bilateral) 40- n = 15 Patients OPTIC CUP VOLUME ASYMMETRY BETWEEN EYES (mm 3 ) Fig. 6. Frequency distributions of volume asymmetry between fellow eyes for each of the three patient groups. NORMAL (Bilateral) n= 20 Patients (Bilateral) n = 45 Patients GLAUCOMA (Unilateral or Bilateral) n=l5 Patients OPTIC CUP AREA ASYMMETRY BETWEEN EYES (mm 2 ) Fig. 7. Frequency distributions for area asymmetry between fellow eyes for each of the three patient groups. (Bilateral) n= 45 Patients GLAUCOMA (Unilateral or Bilateral) n= 15 Patients OPTIC CUP DEPTH ASYMMETRY BETWEEN EYES (mm) Fig. 8. Frequency distributions for depth asymmetry between fellow eyes for each of the three patient groups. statistically significant differences in the distribution of volume, area, and depth of the optic cup among the three patient groups, (X 2 =115.9, p < for volume; x 2 = 158.0, p < for area; x 2 = 110.3, p < for depth). Despite these highly significant differences, the extensive overlap among the three groups greatly restricts the differential diagnostic utility of these measurements for individual patients. Means, standard deviations, and ranges of optic cup volume, area, and depth for the three patient groups are presented in Table I. The asymmetry of optic cup volume area, and depth between eyes was determined for patients with (1) bilaterally normal eyes (N = 20 patients), (2) bilaterally ocular hypertensive eyes (N = 45 patients), and (3) early glaucomatous optic disc cupping and visual field defects in either one or both eyes (N = 15 patients). Frequency distributions of volume, area, and depth asymmetry between fellow eyes are presented in Figs. 6 to 8 for the three patient groups. Each of the optic cup measurements in glaucoma patients showed (on the average) greater asymmetry

8 )lld( Volume 18 Number 12 Stereophotogrammetry of optic disc 15 NORMAL GLAUCOMA 2.8- n=40 eyes r = +.854(p<.OOOI) 2.8- n =!06 eyes r = +.629(p<.OOOI) 2.8-, n=l8 eyes r=+.o4(p>.io) CJ E E ORIFICI: ARE Q. ^D CJ t i , ' > s.. V. * * '.. ' ^1.6- y s '0- s ':-. '* y 16 '''+/'' ' ' " * *...8- / ^. :: * ' '.' '.4- /!* OPTIC CUP DEPTH (mm) Fig. 9. Relationship between optic cup area and depth for normal, ocular hypertensive, and glaucomatous eyes. Table II. Optic cup volume, area, and depth asymmetry between fellow eyes in normal, ocular hypertension, and glaucoma patients Volume asymmetry (mm 3 ): Mean S.D. Range Area asymmetry (mm 2 ): Mean S.D. Range Depth asymmetry (mm): Mean S.D. Range Bilateral normal (N = 20 patients) Bilateral ocular hypertension (N = 45 patients) Unilateral or bilateral glaucoma (N 15 patients) between eyes. Ocular hypertensive patients tended to exhibit slightly greater asymmetry than normals. A chi-square analysis of these results showed statistically significant differences among the three patient groups for volume asymmetry (x 2 = 66.0, p < 0.001), area asymmetry (x 2 = 95.6, p < 0.001), and depth asymmetry (x 2 = 29.2, p < 0.005) between fellow eyes. However, these values also exhibited considerable overlap among the three groups, thereby limiting the clinical usefulness of these measurements for an individual patient. Table II presents means, standard deviations, and ranges for volume, area, and depth asymmetry between fellow eyes in the three groups of patients. The most important finding was obtained for the relationship between area and depth of the optic cup. As shown in Fig. 9 (left graph), there was a high positive correlation (r = +0.85, p < ) between optic cup depth and area in normal eyes. That is, normal optic cups which were large in area also tended to be quite deep, and vice versa. This strong relationship between area and depth was not present for optic cups which had sustained glaucomatous damage, as indicated by the right graph in Fig. 9. Here, the correla-

9 1260 Johnson et al. Invest. Ophthalmol. Visual Sci. December Normal Optic Disc n=70 eyes r=+.65l (p<.oooi) 2.4 Optic Disc Cupping n=36 eyes r = (p<.05) s' 1.6-, * v * " / ' 1.2- *'. >» O ' * OPTIC CUP DEPTH (mm) 1.0 Fig. 10. Relationship between optic cup area and depth for ocular hypertensive eyes with normal optic cups and ocular hypertensive eyes with clinical evidence of early glaucomatous optic cupping. Table III. Percentage of ocular hypertensive and glaucomatous eyes that exceeded the range of normal optic cup measurements Volume Area Depth Volume asymmetry Area asymmetry Depth asymmetry Area/depth relationship Ocular hypertension (N =106 eyes) Glaucoma (N = 18 eyes) tion of optic cup area and depth has completely broken down (r = +0.04, p > 0.10) indicating no reliable area/depth relationship for glaucomatous optic discs. In most cases, this breakdown in the area/depth relationship occurred as a result of increasing area of the optic cup relative to its depth. Only one eye with glaucoma showed a greatly increased depth of the optic cup relative to its area. The area/depth correlation results for ocular hypertensive eyes (r = +0.63, p < 0.001) are between the values for normal and glaucomatous eyes, as shown in the middle graph of Fig. 9. Correlations between optic cup area and depth in normal eyes have also been described by Armaly 40 and Jonsas. 1 To our knowledge, however, the breakdown in optic cup area/depth relationships in glaucomatous eyes has not been previously reported. Additional optic cup area/depth relationships were determined for ocular hypertensive eyes with normal optic discs (Fig. 10, left graph) and ocular hypertensive eyes that appeared to have early glaucomatous optic cupping (Fig. 10, right graph). Although the absence of a visual field defect made the distinction of glaucomatous optic cupping somewhat questionable for the latter group, the results in Fig. 10 provided tentative support for these subjective observations of the optic disc. Area/depth correlations for the two ocular hypertension groups represented intermediate values between normal and glaucomatous eyes (see Fig. 9). Discussion The present findings indicate that in proceeding from normal to ocular hypertensive to glaucomatous eyes, there are increases in the average volume, area, and depth of the optic cup, the average optic cup volume, area, and depth asymmetry between fellow eyes becomes larger, and the optic cup area/depth relationship progressively breaks down. Do these measurements provide a

10 Volume 18 Number 12 Stereophotogrammetry of optic disc 1261 means of accurately differentiating between normal and glaucomatous optic cups for potential clinical diagnostic use? To evaluate this question, we determined the percentage of ocular hypertensive and glaucomatous eyes that exceeded the range of values found in normal eyes for each of the seven optic cup measures (volume, area, depth, volume asymmetry, area asymmetry, depth asymmetry, and the area/depth relationship). The findings are presented in Table III. Depth and depth asymmetry between fellow eyes showed rather poor results, whereas volume, area, and volume and area asymmetry between fellow eyes produced somewhat better separation between the three groups. However, the area/depth relationship revealed the most impressive findings; 89% of the glaucomatous eyes and 47% of the ocular hypertensive eyes were beyond the range of area/depth values for normal optic cups. It should be noted that the two glaucomatous optic cups (11%) that fell within the range of normal area/depth values were from the same individual, who happened to be a high myope. Therefore our photogrammetric magnification factors, which are based upon Gullstrand's reduced schematic eye, 45 introduced errors that may have appreciably underestimated the overall dimensions of the optic cup. Future evaluations that provide corrections for large refractive errors might produce even greater separation among normal, ocular hypertensive, and glaucomatous eyes. There are at least two hypotheses that may be proposed to account for the relationship between glaucomatous damage and optic cup characteristics. (1) Large optic cups may generally be more susceptible to sustaining glaucomatous damage. (2) Glaucomatous damage to the optic disc produced changes in the size and shape of the optic cup. The consistent, graduated increases in optic cup measurements across the three patient groups in this study appear to support the second hypothesis. In particular, the area/depth relationships presented in Figs. 9 and 10 are highly consistent with the second hypothesis and would be difficult to explain solely on the basis of large optic cups. Many previous investigators have reported that glaucomatous damage to the optic cup usually precedes visual field loss. 8 ' 40 ' 41 Our current photogrammetric data generally support this interpretation. The results of this study represent a significant improvement over most previous attempts to differentiate between normal and glaucomatous eyes on the basis of optic disc evaluation alone, 2 " 6 17< 33j ' and at the same time to avoid the problems and variability introduced by methodologic differences in the measurement of cup/disc ratios. 42 Recent multivariate analysis studies by Susanna and Drance 43 and Drance et al. 44 exhibit detection rates which are comparable to our current photogrammetric results, although many of the statistical evaluations were performed on subjective clinical observations. Stereophotogrammetry offers the advantage of objective measurements of the optic cup characteristics and may produce a higher degree of uniformity and standardization among different examiners. We feel that both stereophotogrammetric evaluations of the optic cup and multivariate statistical analysis may be of potential value for clinical diagnostic purposes in glaucoma. We are indebted to Dr. V. Ralph Algazi and Mr. James Stewart from the Department of Electrical Engineering, University of California, Davis, for performing the computer processing of photogrammetric data. REFERENCES 1. Jonsas, C: Stereophotogrammetric techniques for measurements of the eye ground, Acta Ophthalmol. Suppl. 117, Portney, G.: Stereometric analysis of normal and glaucomatous optic discs. In Herron, R.E., and Karara, H.M., editors: Proceedings of the Symposium of Commission V of the International Society for Photogrammetry, Falls Church, Va., 1974, American Society for Photogrammetry, pp Portney, G.: Photogrammetric analysis of volume asymmetry of the optic nerve head cup on normal, hypertensive and glaucomatous eyes, Am. J. Ophthalmol. 80:51, Portney, G.: Qualitative parameters of the normal optic nerve head, Am. J. Ophthalmol. 76:655, 1973.

11 1262 Johnson et al. Invest. Ophthalmol. Visual Sci. December Portney, G.: Photogrammetric categorical analysis of the optic nerve head, Trans. Am. Acad. Ophthalmol. Otolaryngol. 78:275, Portney, G.: Photogrammetric analysis of the three-dimensionel geometry of normal and glaucomatous optic cups, Trans. Am. Acad. Ophthalmol. Otolaryngol. 81:239, Dowman, I., and Elkington, A.: Photogrammetric measurements of the retina of the eye. In Herron, R.E., and Karara, H.M., editors: Proceedings of the Symposium of Commission V of the International Society for Photogrammetry, Falls Church, Va., 1974, American Society for Photogrammetry, pp Saheb, N., Drance, S., and Nelson, A.: The use of photogrammetry in evaluating the cup of the optic nervehead for a study in chronic simple glaucoma, Can. J. Ophthalmol. 7:466, Schirmer, K., and Kratky, V.: Photogrammetry of the optic disc, Can. J. Ophthalmol. 8:78, Schirmer, K.: Instamatic photogrammetry, Can. J. Ophthalmol. 9:81, Kottler, M., Rosenthal, R., and Falconer, D.: Analog vs. digital photogrammetry for optic cup analysis, INVEST. OPHTHALMOL. 15:651, Kottler, M., Rosenthal, R., and Falconer, D.: Digital photogrammetry of the optic nerve head, INVEST. OPHTHALMOL. 13:116, Rosenthal, R., Kottler, M., Donaldson, D., and Falconer, D.: Comparative reproducibility of digital photogrammetric procedure utilizing three methods of stereophotography, INVEST. OPHTHALMOL. VI- SUAL SCI. 16:54, Falconer, D.: Digital stereophotogrammetry of the ocular fundus, Appl. Optics 12:1388, Crock, G., and Parel, J.: Stereophotogrammetry of fluorescein angiographs in ocular biometrics, Med. J. Aust. 2:586, Krakau, C, and Torlegard, A.: Comparison between stero and slit image photogrammetric measurements of the optic disc. In Herron, R.E., and Karara, H. editors: Proceedings of the Symposium of Commission V of the International Society for Photogrammetry, Falls Church, Va., 1974, American Society for Photogrammetry, pp Holm, O.C., Becker, B., Asseff, C, and Podos, S.: Volume of the optic disc cup, Am. J. Ophthalmol. 73:876, Holm, O., and Krakau, C.: A photographic method for measuring the volume of papillary excavations, Ann. Ophthalmol. 1:327, Schirmer, K.: Simplified photogrammetry of the optic disc, Arch. Ophthalmol. 94:1997, Goldmann, H., and Lotmar, W.: Rapid detection of changes in the optic disc: Stereo-chronoscopy, Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 202:87, Thompson, M.: Manual of Photogrammetry, ed 3, Falls Church, Va., 1966, American Society for Photogrammetry. 22. Portney, G., and Krohn, M.: The limitations of kinetic perimetry in early scotoma detection, Trans. Am. Acad. Ophthalmol. Otolaryngol. 85:287, Armaly, M.: Ocular pressure and visual fields: a ten-year follow-up study, Arch. Ophthalmol. 81:25, Drance, S., Fairclough, M., Butler, D., and Kottler, M.: The importance of disc hemorrhage in the prognosis of chronic open angle glaucoma, Arch. Ophthalmol. 95:226, Keltner, J.L., Johnson, C.A., and Balestrery, F.G.: Suprathreshold static perimetry: initial clinical trials with the Fieldmaster automated perimeter, Arch. Ophthalmol. 97:260, Enoch, J.M., Lazarus, J., and Johnson, C.A.: Human psychophysical analysis of receptive fieldlike properties. I. A new transient-like visual response using a moving windmill (Werblin-type) target, Sens. Processes 1:14, Kronfeld, P.: The early ophthalmoscopic diagnosis of glaucoma, J. Am. Optom. Assoc. 23:156, Kronfeld, P.: The optic nerve. In Symposium on Glaucoma, Transactions of the New Orleans Academy of Ophthalmology, St. Louis, 1967, The C. V. Mosby Co., pp Kirsch, R., and Anderson, D.: Identification of the glaucomatous disc, Trans. Am. Acad. Ophthalmol. Otolaryngol. 77:143, Kirsch, R., and Anderson, D.: Clinical recognition of glaucomatous cupping, Am. J. Ophthalmol. 75:442, Weisman, R., Asseff, C, Phelps, C, Podos, S., and Becker, B.: Vertical elongation of the optic cup in glaucoma, Trans. Am. Acad. Ophthalmol. Otolaryngol. 77:157, Hitchings, R., and Spaeth, G.: The optic disc in glaucoma. I. Classification, Br. J. Ophthalmol. 60:778, Fishman, R.: Optic disc asymmetry: a sign of ocular hypertension, Arch. Ophthalmol. 84:590, Shaffer, R., and Heatherington, J.: The glaucomatous disc in infants: a suggested hypothesis for disc cupping, Trans. Am. Acad. Ophthalmol. Otolaryngol. 73:929, Schwartz, B.: Cupping and pallor of the optic disc, Arch. Ophthalmol. 89:272, Schwartz, B., Reinstein, N., and Lieberman, D.: Pallor of the optic disc, Arch. Ophthalmol. 89:278, Donaldson, D.: A new camera for stereoscopic fundus photography, Arch. Ophthalmol. 73:253, Frisen, L.: Resolution at low contrast with a fundus camera. Comparison of various photographic films, INVEST. OPHTHALMOL. 12:865, Krohn, M.A., Keltner, J.L., and Johnson, C.A.: Comparison of photographic techniques and films used in stereophotogrammetry of the optic disc. Am. J. Ophthalmol. (in press). 40. Armaly, M.: The optic cup in the normal eye. I. Cup width, depth, vessel displacement, ocular tension

12 Volume 18 Number 12 Stereophotogrammetry of optic disc 1263 and outflow facility, Am. J. Ophthalmol. 68:401, Armaly, M.: Optic cup in normal and glaucomatous 44. eyes, INVEST. OPHTHALMOL. 9:425, Schwartz, J.T.: Methodologic differences and measurement of cup-disc ratio, Arch. Ophthalmol. 94:1101, Susanna, R., and Drance, S.M.: Use of discriminant analysis. I. Prediction of visual field defects from features of the glaucoma disc, Arch. Ophthalmol. 96:1568, Drance, S.M., Schulzer, M., Douglas, G.R., and Sweeney, V.P.: Use of discriminant analysis. II. Identification of persons with glaucomatous visual field defects, Arch. Ophthalmol. 96:1571, Helmholtz, H.: Treatise on Physiological Optics (J. P. C. Southall, editor), New York, 1962, Dover Publications, vol. I. Information for authors Most of the provisions of the Copyright Act of 1976 became effective on January 1, Therefore, all manuscripts must be accompanied by the following written statement, signed by one author: "The undersigned author transfers all copyright ownership of the manuscript (title of article) to The Association for Research in Vision and Ophthalmology, Inc., in the event the work is published. The undersigned author warrants that the article is original, is not under consideration by another journal, and has not been previously published. I sign for and accept responsibility for releasing this material on behalf of any and all co-authors." Authors will be consulted, when possible, regarding republication of their material.