RELIABILITY OF OPTIC DISK TOPOGRAPHIC MEASUREMENTS RECORDED WITH A VIDEO-OPHTHALMOGRAPH

The video-ophthalmograph records the topography of the optic disk via simultaneous stereoscopic images which are stored and analyzed with the help of a microcomputer. This information is used to generate the vertical cup-disk ratio, the vertical optic disk diameter, the cup volume, and the neuroretinal rim area. To determine the reliability of the data, we recorded information for one eye of each of five patients ten times to determine the interphotographic error variance. We also analyzed one photograph for each of five patients ten times to deter mine the intraphotographic variance attributable to repeated analysis of the same photograph. The interphotographic and intraphotographic coefficients of variation were 2% to 18% and 2% to 7% respectively for these measurements.


RELIABILITY OF OPTIC DISK TOPOGRAPHIC MEASURE MENTS RECORDED WITH
The video-ophthalmograph records the topography of the optic disk via simultaneous stereoscopic images which are stored and analyzed with the help of a microcomputer . This information is used to generate the vertical cup-disk ratio , the vertical optic disk diameter, the cup volume , and the neuroretinal rim area. To determine the reliability of the data, we recorded information for one eye of each of five patients ten times to deter mine the interphotographic error variance. We also analyzed one photogr aph for each of five patients ten times to determine the intraphotographic variance attr ibutable to repeated analysis of the same photograph. The interphotographic and intr aphotographic coefficients of variation were 2% to 18% and 2% to 7% respectively for these measurements.
The topography of the optic nerve head is of the utmost importance in the diagnosis and monitoring of the progression of glaucoma. Pederson and Anderson 1 showed that progressive disk cupping can precede measurable visual field defects in glaucoma. Serial stereoscopic photographs are useful in monitoring changes in optic disk topography, but a rapid, reliable tech nique of quantitating these changes is lacking.
Previous investigators have used photogrammetry to characterize the topography of the optic disk. These techniques have the drawback of needing the extraction of data from stereoscopic pairs of photographs by a skilled photogrammetrist. 2-t 7 A technique of digital photogrammetry of the optic nerve head involving electronic scanning and digitizing of a pair of stereoscopic photographs 1 s. 21 makes it unnecessary for a skilled photogrammetrist to perform the labor-intensive translation from the stereoscopic pairs.
The Rodenstock video-ophthalmograph records the topography of the optic disk by recording simultaneous stereoscopic video images which are stored and analyzed with the help of a microcomputer. 22 This technique does not require skilled operators and also has the advantage of bypassing the photographic process. Additionally, the results of optic disk measurements are available rapidly and sequential comparisons can be extracted from the digitized information.

MATERIAL AND METHODS
The instrument-The video-ophthalmograph consists of seven major compo-nents: an optical head, containing all of the optical system and some of the electronics along with the support for the patient's head and chin; the operator's desk, the pedestal of which contains most of the electronics; a televideo CRT moni-. tor, on which the operator's instructions are displayed; a televideo keyboard on which the operator can type information and instructions; another CRT monitor on which the results are displayed; a light pen which the operator uses to control the instrument; and a floppy disc drive to store the results.
The optical system of the videoophthalmograph performs three general functions. First, it illuminates the eye with infrared light and forms an image of the iris and pupil on an infrared-sensitive television camera, permitting the operator to perform coarse manual alignment of the instrument with respect to the pupil. This subsystem also permits automatic fine continuous alignment.
A second subsection of the optical system provides a fixation target for the patient. This target is a flashing white spot bright enough to be seen even when the eye is brightly illuminated with the measuring light. The position of the target is controlled by the operator to center the optic disk on the video screen.
The third optical subsystem illuminates the fundus with light and forms a simultaneous stereoscopic pair of images of the fundus on a sensitive television camera. The lights, motors , and cameras in the optical head are controlled by two computers inside the desk. These computers are controlled by the keyboard and the light pen. While the instrument is on all of the pictures and data are stored on a hard disc. At the end of a data-taking session the operator can store the pictures and data on 8-inch floppy discs.
Methods-Once the pictures and data are stored, the computer calculates the depths of points on the optic disk by measuring the distance between corresponding points of the simultaneous stereoscopic pairs, giving depth values with a theoretical sensitivity of 12 µm. This is performed for 1,600 points on the optic disk. The computer then calculates the vertical cup-disk ratio, neuroretinal rim area, cup volume, and vertical diameter of the optic disk (Figure). The operator indicates the optic disk margin by choosing four points at the horizontal and vertical margins to which the computer fits an ellipse that defines the edge. To calculate the vertical cup-disk ratio the computer defines the cup margin by subtracting 150 µ.m from the level at the optic disk margin along one meridian and then measuring the distance from the center of the optic disk along that meridian to that defined depth. This is repeated 51 times superiorly each degree along radii from 65 to 115 degrees and inferiorly each degree along radii from 245 to 295 degrees. The mean of these 102 measurements is the vertical cup-disk ratio. To determine the reliability of the data obtained with the video-ophthalmograph, we recorded the data for one eye of each of five patients ten times at the same sitting. We then analyzed the ten records to generate the vertical cup-disk ratio, vertical optic disk diameter, cup volume , and neuroretinal rim area. These data were then analyzed statistically to determine the interphotographic error variance. We also analyzed one record for each of the five patients ten times. These data were analyzed statistically to determine the intraphoto-graphic error variance attributable to repeated analysis of the same record. RESULTS We calculated the percent error as a proportion of the mean (coefficient of variation) as follows: C. V. = standard deviation + mean of sample. The coefficient of variation in the interphotographic and intraphotographic studies respectively was 11. 2% and 3. 9% for the cup-disk ratio, 2.4% and .l.9% for the optic disk diameter, 17. 7% and 6. 7% for the neuroretinal rim area, and 18.6% and 2.5% for the cup volume. The standard deviations for the cup-disk ratio were 0.067 and 0.024 for interphotographic and intraphotographic studies respectively (Table).

DISCUSSION
The video-ophthalmograph provides a quick and reliable method of determining the vertical cup-disk ratio and recording any changes in this measurement. Current clinical techniques produce interobserver and intraobserver differences in the cup-disk ratio. [23][24][25] The frequen cy of intraobserver differences of 0. 2 or greater was 28% in an experienced group of ophthalmologists. 25 The frequency of inter- observer differences of 0. 2 or greater was 25% with contact lens biomicroscopy. 25 The video-ophthalmograph provides an objective means of recording this measurement with a standard deviation of 0. 067. Therefore, with this instrument only 5% of repeat cup-disk measurements will differ by more than : 0.13 (2 S. D.) and only 0.3% differ by :t 0.2 or more (3 S.D. ).
Cup volume and neuroretinal rim area are quantities that cannot be easily estimated in a routine clinical examination. It has been shown that neuroretinal rim area may be more valuable than the cupdisk ratio in determining the presence of glaucomatous damage. 7 ,2.6 Now that the reproducibility of this technique has been determined, we intend to correlate the cup-disk ratio recorded clinically with that obtained on the video-ophthalmograph. We plan to compare the neuroretinal rim area measured by planimetry with the videoophthalmographic measurements. These measurements will then be compared to the findings on psychophysical tests in an attempt to determine the usefulness of this machine in differentiating normal subjects from patients with any stage of glaucoma.
Although the intraphotographic error is small enough to be acceptable, the interphotographic error is significant enough that it must be taken into account when measuring change in the optic disk as a feature of glaucoma. Attempts to reduce interphotographic errors should be made.