In the harsh hinterlands of the country, where medical facilities and most modern technology are hard to come by (except the ubiquitous cell phones, of course), eye care may perhaps be available at the click of a button (or the tap of a screen).
Imagine an app or a piece of software on the computer that analyses the image of an eye, even if it is taken through a smart phone, and then, with an accuracy that matches medical professionals, detects the earliest symptoms of ‘glaucoma’.
A team of researchers from the Indian Institute of Science (IISc), Bengaluru, have developed a Java-based software that, by seeing the images of the interior surface of the eye (called the fundus), can detect ‘glaucoma’ to an accuracy of 90 per cent.
Glaucoma is an eye disorder that leads to the damage of the optic nerves and irreversible blindness. However, the condition often sneaks on to the patient, with no symptoms until it is too late to treat it. With an estimated 12.3 per cent suffering from glaucoma in India, it is the second leading cause of blindness after cataract.
Professor Chandra Sekhar Seelamantula and his team from the Department of Electrical Engineering, IISc, have coded a tool that can identify potentially glaucomatous conditions from the fundus image.
Low-cost cameras and addendums to smartphones are now available that would allow for the clicking of the fundus image. The images are compared with symptoms of the conditions. For instance, one of the parameters that the software calculates is the ‘cup-to-disc ratio’ (CDR) from images of the posterior chamber of the eye. The optic disc is an anatomical area where the optic nerves and blood vessels enter the retina; this leads to a bulge — a cup — in the middle of the disc. While in a normal eye, the cup-to-disc ratio is around 0.3, a larger ratio (that is, increased size of the cup) will indicate glaucomatous condition.
Recognising other parameters, the software can categorise the condition as mild, moderate, or severely glaucomatous — all in accordance with accepted diagnostic guidelines. “The software tool can en masse analyse hundreds of images. Each image takes a few seconds to process,” says Prof. Seelamantula.
Currently, the measuring of the CDR and analyses of the images are done manually, which takes up numerous clinical hours in an already overburdened system. The software, however, promises to hasten the process, allowing for effective detection during glaucoma screening.
Plans to patent the innovation remain in the pipeline, while the researchers are in talks with companies for technology transfer.
