Computerized stereo photo analysis may be superior to standard fundus photography

Study suggests technology may be advancing past standard color stereo disc photography.

MIAMI – Digital optic nerve photos are less prone to interpretative variability when compared to standard fundus photos, according to a study performed here at the Bascom Palmer Eye Institute.

The study, by Donald Budenz, MD, Michele Lim, MD, and Steven Gedde, MD, sought to measure the variability difference between a digital stereo imaging system (Discam; Marcher Enterprises Ltd., Hereford, England) and standard color stereo disc photography.

Background of study

The study was an unexpected consequence of a move to fully computerized patient medical records, with all photos and visual fields to be stored digitally. As the authors were pursuing this goal they were contacted by the makers of the Discam system.

“Our primary interest in this was actually because we were switching to an electronic medical record,” Dr. Budenz said. “I didn’t realize when we agreed to test this system out for our electronic medical record purposes that it actually may offer a better way to judge cup:disc ratios.”

Gold standard is challenged

---Intraobserver variability with the Discam was as good as intraobserver variability with stereo disc photography.

Dr. Budenz explained that new technologies may allow for better optic nerve analysis than can be realized with standard photography alone.

“We’re really living in the dark ages in terms of looking at the optic nerve. The standard, according to the American Academy of Ophthalmology’s (AAO) Preferred Practice Pattern, is still stereo disc photographs. We’re ignoring the tremendous developments in computer technology in helping us judge cup:disc ratios.”

The problem, he explained, is that some of the newer technologies may not be offering levels of reproducibility comparable to photography. That has been one of the main criticisms, he said.

“In glaucoma, we’re looking for very small changes, so that if a system has a margin of error of 10%, you’d actually have to see a change much greater than 10% to actually be able to call it glaucomatous change,” Dr Budenz said.

The study by Dr. Budenz and colleagues examined whether there is more or less variability with computer-assisted digital photography as compared to standard stereo disc photography. He pointed out this was not a comparison with other new technologies, but with the prevailing standard.

“Since the current recommendation from the AAO, from the people who lecture on progressive glaucoma and in the Ocular Hypertension Treatment Study (OHTTS) is really stereo disc photographs, we’re going to target that and see if we can’t do better,” Dr. Budenz said. “We try to focus on the gold standard and determine if this is better. Maybe this will become the gold standard against which other things will be compared.”

Concise preliminary study

---Computer screen appearance of a glaucoma suspect with disc and cup outlines.

The study group consisted of 40 patients, all with a diagnosis of glaucoma or suspected glaucoma. Standard and computer-assisted pictures were obtained on the same day. The images were later evaluated for variability between observations by the same observer (intraobserver variability) and then for variability between two observers (interobserver variability) in describing cup:disc ratio.

Intraobserver variability is the comparison of disc readings on two separate occasions by one observer, Dr. Budenz explained. “That means if I look at an optic nerve photo at one time and say it’s 0.4 horizontally and 0.5 vertically, then read it again at a completely different time on a completely different day, will I get a 0.5 and a 0.6, or a 0.4 and 0.5 again? The closer my two measurements, the lower the intraobserver variability of that measurement.”

Interobserver variability was measured by taking the average of two separate readings recorded by Dr. Budenz and comparing them to averages of two separate readings recorded by Dr. Gedde.

Interobserver readings may be particularly significant, said Dr. Budenz: “Historically, our gold standard, stereo disc photographs, has a relatively low intraobserver variability but a relatively high interobserver variability. This means that if you compare my readings with another glaucoma specialist, they are likely to vary quite a bit.”

According to Dr. Budenz, intraobserver variability with the Discam was as good as intraobserver variability with stereo disc photography. Interobserver variability was better with the Discam, he said. He declined to give further details pending publication of a peer-reviewed version of the study.

Further studies initiated

The study, which began in fall of 1999, has been concluded. Dr Budenz indicates that this was a preliminary investigational study with a small study size. But the results have warranted two other longer-term studies, one of which will seek to determine which of the two methods can detect glaucomatous change first. The other will look for correlation between Discam data and visual field data in glaucoma patients.

Discam simultaneous stereo images of glaucoma suspect showing disc and cup outlines.

Same photograph with disc rim thickness at 16 locations using the Discam.

For Your Information:

• Donald Budenz, MD, is an associate professor of ophthalmology at Bascom Palmer Eye Institute, 900 NW 17th St., Miami, FL 33136. He is also assistant director of the optic disc reading center of the Ocular Hypertension Treatment Study. He can be reached at (305) 326-6310; fax: (305) 326-6337; e-mail: dbudenz@med.miami.edu. Dr. Budenz has no direct financial interest in the products mentioned in this article, nor is he a paid consultant for any companies mentioned.
• Steven Gedde, MD, is also at Bascom Palmer Eye Institute. He can be reached at (305) 326-6435; fax: (305) 326-6337; e-mail: sgedde@med.miami.edu.
• Michele Lim, MD, is in practice at the University of California–Davis Medical Center, Department of Ophthalmology, 4860 Y St., Ste. 2400, Sacramento, CA 95817; (916) 734-6961; fax: (916) 734-6992; e-mail: michele.lim@ucdmc.ucdavis.edu.

• For more information about Discam, contact Marcher Enterprises Ltd. at Clearview Court, Rotherwas Industrial Estate, Hereford HR2 6RJ, England; (44) 1432 278400; fax: (44) 1432 354218; e-mail: enqs@marcher.co.uk; Web site: www.marcher.co.uk/frames.htm.