Stanford University School of Medicine scientists have created a new, smarter method for predicting if and when patients with age-related macular degeneration (AMD) will contract the most damaging form of the disease.

The technique, which relies on computer analysis of retinal scans already commonly collected by ophthalmologists and optometrists, will enable doctors to accurately calculate whether any given patient is likely to progress to “wet” AMD — the later stage of the illness that causes blindness if not treated in time.

AMD, the leading cause of vision loss in adults over 65, results from deposits called drusen in the retina. In dry AMD, drusen buildup can damage the macula, the area of the retina most key to sight. Wet AMD, a far more debilitating version of the disease, occurs when abnormal blood vessels also accumulate behind the retina and leak. Blindness can quickly ensue.

There is no feasible way to treat wet AMD preemptively. And until now, there was no way to tell with certainty whether dry AMD would become wet AMD. Patients just had to hope their next doctor’s visit would come at the right time to detect and head off any further deterioration before it was too late.

Thanks to the new computer algorithm, ophthalmologists can now predict whether a particular person’s AMD will progress within one, three or five years, enabling doctors to tell their patients much more accurately when their next visit should be.

“A larger follow-up study is needed,” said the study’s senior author, Daniel L. Rubin, assistant professor of radiology and of biomedical informatics in a Medical School report.