Pallor of the optic disc. Quantitative photographic evaluation

Automatic computer-aided analysis of optic disc pallor in fundus photographs

PURPOSE

Assessment of optic disc pallor in fundus photographs may be frequently misinterpreted due to the subjective nature of interpretation. We developed a fully automatic computer-aided detection (CAD) system for optic disc pallor using colour fundus photographs and evaluated the accuracy of the system.

METHODS

A newly proposed CAD system was developed for automated segmentation and image analysis of optic disc pallor, and a logistic regression model was developed for risk analysis. A total of 230 photographs with variable degree of optic disc pallor, and 123 normal optic discs confirmed by optical coherence tomography were tested for validation of the software. Sensitivity and specificity of the CAD system in automatic detection of optic disc pallor using colour fundus photographs were evaluated. The results of manual detection of optic disc pallor on fundus photographs by two independent ophthalmologists were compared with the efficacy of the CAD system.

RESULTS

The fully automated CAD system achieved a sensitivity of 95.3% and a specificity of 96.7% for detecting optic disc pallor in colour fundus images. The overall accuracy of the CAD system was 96.1%, which was superior to the results of manual detection by individual examiners.

CONCLUSIONS

We developed an automated CAD system that successfully detected optic disc pallor in fundus photographs. The proposed algorithm can assist the clinical judgement of ophthalmologists for detecting optic disc pallor in fundus photographs.

Agreement in assessing cup-to-disc ratio measurement among stereoscopic optic nerve head photographs, HRT II, and Stratus OCT

PURPOSE

To compare the level of agreement between subjective and objective methods in estimating horizontal and vertical cup-to-disc ratios (HCDR and VCDR, respectively) to determine if objective techniques may be used as surrogates for subjective cup-to-disc (CDR) estimation.

METHODS

Fifty-one glaucoma patients and 49 control subjects underwent full ophthalmic examination, stereoscopic optic nerve head photographs (ONHPs), confocal scanning laser ophthalmoscopy (Heidelberg Retina Tomography II [HRT II]), and optical coherence tomography (Stratus OCT). The intraclass correlation coefficient (ICC) and Bland Altman plots were used to assess the agreement across the three methods. Repeated measures analysis of variance (ANOVA), Tukey adjustment, and pairwise P values were used to compare the HCDR and VCDR estimates between three clinicians who reviewed photos, Stratus OCT, and HRT II.

RESULTS

For the clinicians, the agreement in subjectively assessed HCDR and VCDR was substantial (ICC=0.84 and 0.85, respectively), and for all three methods, overall agreement was good (ICC=0.75 and 0.77 for the HCDR and VCDR, respectively). Stratus OCT provided the largest overall mean+/-SD HCDR (0.68+/-0.14) and VCDR (0.62+/-0.13). The smallest overall mean+/-SD HCDR was provided by ONHP (0.32+/-0.16), and the smallest overall mean+/-SD VCDR was provided by HRT II (0.26+/-0.20). Repeated measures ANOVA test demonstrated significant differences across the three methods for glaucomatous (P=0.0017 and 0.0016, HCDR and VCDR, respectively) and normal (P=0.0001 for both HCDR and VCDR) eyes. Tukey adjustment demonstrated specific statistical differences between pairs of methods.

CONCLUSIONS

Although the overall agreement between various methods was good, the mean estimates were statistically different. Additional studies are needed to evaluate the sources of variability, their level of significance, and longitudinal agreement between various methods of the CDR estimation.

Neuroretinal rim width ratios in morphological glaucoma diagnosis

AIMS

To evaluate the inferior to temporal neuroretinal rim width ratio and superior to temporal rim width ratio as measures of rim shape for diagnosis of glaucoma.

METHODS

Colour stereo optic disc photographs of 527 normal subjects, 100 ocular hypertensive individuals with normal visual fields, and 202 open angle glaucoma patients with a mean perimetric defect of less than 10 dB were morphometrically evaluated. Eyes with an optic cup area of < 0.2 mm2 were excluded.

RESULTS

In the normal subjects, inferior to temporal rim width ratio (1.67 (SD 0.53)) was significantly (p < 0.0001) higher than superior to temporal rim width ratio (1.56 (0.49)). Both ratios were significantly (p < 0.0001) higher the more vertically the optic disc was configured. In the normal eyes, both ratios were statistically independent of disc size, rim area, refractive error, age, and sex. With the differences being more marked for the inferior to temporal ratio than for the superior to temporal ratio, both rim width ratios were significantly (p < 0.005) lower in the ocular hypertensive group than in the normal group. Despite the high significance of the differences, diagnostic power of the inferior ratio and the superior ratio was 59% and 58%, respectively, indicating a marked overlap between the groups.

CONCLUSIONS

Abnormally low inferior to temporal and superior to temporal rim width ratios can indicate glaucomatous optic nerve damage in some ocular hypertensive eyes. Being independent of optic disc size and ocular magnification, the rim width ratios may be taken as one among other variables for the ophthalmoscopic optic disc evaluation, taking into account, however, a pronounced overlap between normal eyes and ocular hypertensive eyes.

Three-dimensional evaluation of optic disc pallor in open angle glaucoma

We evaluated the extent of pallor on the walls of the optic cup three-dimensionally using simultaneous stereophotographs of the optic discs of 29 normals, 29 ocular hypertensives and 28 primary open angle glaucomas. Pallor was located at the bottom of the optic cup. Pallor ascends the walls of the cup as its extent increases. Statistically significant differences in the extent of pallor in all four quadrants of the optic cup were observed, with the glaucomas having greater extents of pallor than ocular hypertensives, and the ocular hypertensives greater than the normals. The greatest percentage increase in mean pallor from normals to open angle glaucomas occurred on the nasal and inferior walls. This differential extent of pallor could be useful in characterizing ocular hypertension and chronic open angle glaucoma.

Increased preferential optic disc asymmetry in ocular hypertensive patients compared with control subjects

The authors determined asymmetrical differences in optic disc cupping and pallor between the right and left eyes of 57 control subjects and 75 patients with ocular hypertension. Photogrammetry was used to measure the optic disc cup, and computerized image analysis was used to measure the optic disc pallor for the total disc and its quadrants. Generally, for control and ocular hypertensive eyes, the median values of cupping and pallor were significantly larger in the left eye than in the right, except for depth in the inferior quadrant (right eye minus left eye = 9.3%, P = 0.000) and pallor in the nasal quadrant (right eye minus left eye = 21.3%, P = 0.01), which were preferentially larger in the right eye than in the left in ocular hypertensive eyes. Ocular hypertensive eyes also showed a greater difference between the right and left eyes compared with control eyes for inferior depth (P = 0.06) and nasal pallor (P = 0.02). Detection of optic disc differences as a sign of ocular hypertension between the right eye and the left should evaluate the inferior and nasal quadrants of the disc.

Static threshold asymmetry in early glaucomatous visual field loss

Ten normal subjects underwent static threshold visual field testing of both eyes with the Humphrey perimeter, with one eye tested twice. The mean sensitivity of the field seemed virtually identical in the two eyes, with the average difference between the right and left eyes (0.65 decibels [dB]) being no greater than the testing error as reflected in the difference between the same eye tested twice (0.7 dB). The authors provide the mathematical basis for recognizing that a right eye-left eye difference in mean sensitivity might be abnormal. Additional information is needed about the variance of the right eye-left eye difference in the population at large, but present information suggests that a 2-dB difference may be meaningful on a single examination. A 1.5-dB difference is statistically significant if confirmed on a second test, and a difference as small as 1 dB may be meaningful if shown consistently in a series of four examinations. In all cases, nonglaucomatous causes of field abnormality needs to be ruled out, and the generalized asymmetry is most meaningful if it is consistent with asymmetry of cupping or intraocular pressure. Several cases are reported in which a mild (1 dB) generalized depression of the visual field is the only recognizable abnormality in the visual field in eyes with early glaucoma.

Reproducibility of computerized pallor measurements obtained with the Rodenstock Disk Analyzer

Computerized pallor measurements of the optic nervehead were performed by global analysis with the Rodenstock Disk Analyzer. Total intervideographic and intravideographic variability was calculated. The coefficients of variation ranged from 8.3% to 20.08% for intervideographic variability of different pallor histogram values. The intravideographic variability for pallor histogram values ranged from 0.82% to 2.94%.

The reproducibility of computerized boundary analysis for measuring optic disc pallor in the normal optic disc

The optic discs of eight normal subjects were photographed on three separate visits under widely varying conditions. Each photograph was scanned by a television camera, converted to digital form, and then analyzed using computer image analysis techniques. The computer was programmed to delineate the boundary around the optic disc and the area of pallor. The percentage of disc pallor was then computed across the entire disc and for each disc quadrant. The reproducibility of the procedure was determined by evaluating several photographic parameters: camera flash intensity, film type, magnification, and centering of the disc in the photograph. The focus adjustment on the television scanner was also varied. Under carefully controlled conditions, results showed that the reproducibility of measuring the total percent of disc pallor was 2%. Results for reproducibility are also given for less than optimal conditions and for manual planimetry on the same set of photographs.

Detection of increased pallor over time. Computerized image analysis in untreated ocular hypertension

Computerized image analysis was used to measure changes in percent area of pallor of the optic disc in a retrospective study of a group of 16 normals and 39 untreated ocular hypertensives. The optic discs of the subjects were photographed on at least two visits at least one year apart (mean, 2 years). No significant changes in area of pallor for the normal group were observed between visits, except in the inferior quadrant of the left eye. The ocular hypertensive group showed a significant increase in area of pallor between visits for the total disc and for most quadrants of both eyes. Multiple regression analysis revealed that, in addition to ocular hypertension, the significant factors associated with a change in optic disc pallor were change of ocular pressure, standard deviation of the ocular pressures, presence of vascular hypertension, and standard deviation of vascular pulse pressures. Clinical implications of early detection of disc changes for follow-up and treatment of ocular hypertensives are discussed.

Glaucoma screening in the health care setting

Glaucoma surveys are an important means of detecting early cases of glaucoma. Traditionally, most such surveys have been conducted in the community, while the health care setting has been underutilized as a screening location. We reviewed studies on surveys conducted in community and health care settings, using tonometry and ophthalmoscopy, and presented data from a study we carried out in the outpatient department of a hospital. Results of our review indicate that over 11 times as many people were screened in the community studies as in the health care studies, but the latter had the higher percentage of referrals. Data from the literature review, as well as from our own study, suggest that the high number of referrals is related to the characteristics of the population available for screening in a health care facility. In such a population there is a higher percentage of individuals at risk for development of glaucoma. Risk factors for the disease include age, sex (males more at risk than females), race, (blacks more at risk than whites), family history of glaucoma, and presence of diabetes and vascular diseases. It was also found that ophthalmoscopy is an effective, but underutilized, method of screening. We recommend an increase in use of health care facilities as sites of glaucoma screening surveys. We also suggest that physicians employ ophthalmoscopy to screen for glaucoma as part of the routine funduscopic examination.

Glaucoma screening. A cost-effectiveness analysis

We designed an intricate model of the process of glaucoma screening, diagnosis, and treatment in order to evaluate the medical care costs of such efforts in relation to the benefits in terms of quality-adjusted years of vision saved, utilizing the economic principles of cost-effectiveness analysis. Although a relatively limited data base and numerous assumptions concerning the accuracy of diagnostic tests, the natural history of ocular hypertension and glaucoma, and the effectiveness of available treatment modalities, limit our ability to draw definitive conclusions concerning the cost-effectiveness of various glaucoma screening options, our analysis indicates that glaucoma screening is probably cost-effective when targeted at certain subgroups of the population. Our analysis suggests that changes in several aspects of existing screening policies may be appropriate if cost-effectiveness is to be used as one of the criteria for the efficient allocation of resources to and within screening programs. Specifically, combinations of screening tests and screening targeted at high risk populations such as blacks, diabetics, and relatives of glaucoma patients are probably more cost-effective than screening of the general population with a single test. In younger populations, the importance of detecting ocular hypertension argues for the use of tonometry. Ophthalmoscopy may be more cost-effective in older age groups in whom the higher prevalence of glaucoma outweighs the need for identifying ocular hypertensives. In the very elderly, automated perimetry becomes cost-effective since the detection of established field loss will have greater yield. In addition, it is evident that diversion of resources away from actual screening efforts and towards efforts aimed at improving follow-up and compliance would be an additional cost-effective strategy.

What happens to the optic disc and retina in glaucoma?

In glaucoma there is a spectrum of disc damage ranging from highly localized tissue loss (a notch in the neuroretinal rim), vertically oval cupping, or simple concentric enlargement of the cup. The accompanying visual loss may likewise range from a well-localized nerve fiber bundle detect to a generalized depression. Cases with diffuse loss of axons, a concentric enlargement of the cup, and a generalized reduction in visual function are more difficult to recognize unless the previous state of the disc and vision is well documented or there is asymmetry between the two eyes.

The normal cup-disk ratio

We determined the cup-disk ratios of 580 individuals, ranging in age from 4 to 91 years, with a Hruby lens; we also determined the cup-disk ratios of 289 of the 580 with a direct ophthalmoscope. All had applanation pressures of 18 mm Hg or less. There was no linear relationship (covariation) between cup-disk ratio and refractive error or intraocular pressure, but there was a tendency toward increasing cup-disk ratio with increasing age. The mean cup-disk ratio in the group examined with the Hruby lens was 0.38 and that in the group examined with the direct ophthalmoscope was 0.25. With either method almost all eyes had cup-disk ratios of 0.7 or less. In 9% of the eyes the cup was ovoid. In only 10% of these cases was the vertical cup-disk ratio greater than the horizontal cup-disk ratio, and in only one case was this difference 0.2 or more. The Hruby lens method consistently gave a slightly larger cup-disk ratio than the direct ophthalmoscope did. However, there was seldom a disparity of more than 0.2. The Hruby lens or biomicroscopic lens should be used to determine the fullest extent of the contour of the cup. Any cup-disk ratio of 0.7 or more, any vertical cup-disk ratio larger than the horizontal cup-disk ratio, and any disparity between the direct ophthalmoscope estimation and Hruby lens estimation of more than 0.2 should be viewed with suspicion.

The variation and covariation of cup and disc diameters

The variation and covariation of cup and disc diameters were studied in a material derived from a population survey and consisting of 2,334 fundus photographs from as many eyes in 1,322 subjects. A simple device was used to facilitate focusing of the camera. The colour slides were projected on to a screen at a fixed distance and measured on ruled paper. The effect of refraction on the magnification in the eye-camera system was compensated by the use of a simple correcting factor. Some apparently quite normal discs, nevertheless, had an area more than four times larger than that of other equally normal ones. The sizes of discs and cups covaried, however, to a suprisingly great extent (r = 0.8) and changes in disc diameter were in general parallelled by similar changes in cup diameter. The amount of tissue in the optic nerve head therefore varied somewhat less than the disc size. Cup diameters were widely dispersed, unevenly distributed and heavily dependent on disc size. The "average rim breadths", on the other hand, were much less dispersed, normally distributed and independent on the disc diameter. By taking the covariation of cup and disc diameters into account the detection of any enlargement or diminution of the optic cup ought to be facilitated.

Optic disk topography and visual field defects in patients with increased intraocular pressure

One hundred one eyes of 54 ocular hypertensive patients were studied to determine the value of optic disk evaluation in predicting the presence or absence and the location and type of visual field defects by means of binocular slit-lamp examination and projected 35-mm monocular color slides. We suspected visual field loss erroneously in only one patient, and failed to detect ten (16%) of 61 eyes with field defects. An examiner experienced in slit-lamp evaluation of the optic disk can accurately predict the presence, absence, and location of field defects in most cases by stereoscopic evaluation of disk topography.

Acquired cupping of the optic nerve head in normotensive eyes

Clinical measurements of a horizontal cup/disc ratio were obtained among a sample of 160 normal twin subjects by stereoscopic biomicroscopy. Measurements of horizontal cup/disc ratio were normally distributed and theere was no significant association with sex, laterally, race, or refractive error in the mid-range. However, there was a significant association between size of the cup/disc ratio and age as well as intraocular pressure. After taking into account the correlation between age an intraocular pressure in normal eyes, the association between cup/disc ratio and each of these variable remained statistically significant. It is not always recoginized that the size of the normal physiological cup increases with age, and it was not previously observed that "normal" cupping occurs with age apart from the effect of pressure. The average effect of age on cup/disc ratio, amounting to an icrease of 0-0026 per year for this sample, is small. But, allowing for individual variation, some persons could show a recognizable acquired increase in cup/disc ration with age, even in the absence of glaucoma. The possibility of "normal"cupping occurring in association with age, pressure, or high myopia, alone or in combination, merits cosideration in the clinical evaluation of glaucoma suspects and serves to re-emphasize the importance of an adequate central visual field examination in the diagnosis of glaucoma. Studies are needed on the differentiatin between "normal" and "morbid" moulding of the optic nerve head.

Use of photographs for measuring cupping in the optic disc

Images