Microsoft shows how Artificial Intelligence can be used to prevent blindness

Eye Blindness

Currently close to 285 million people are visually impaired, of which 55 million reside in India. The causes for visual impairment can be many, but some of them are preventable. Take for example, a disease such as diabetic retinopathy. If not treated early, diabetic retinopathy can be a major cause of irreversible blindness. India which is now being called the ‘diabetic capital of the world’ due to the highest number of diabetic patients in the world (estimated to be more than 50 million) – clearly has a huge problem to solve. Experts believe that two-thirds of all Type 2 diabetics and all Type 1 diabetics are expected to develop diabetic retinopathy over a period of time. Currently, out of 50 lakh surgeries conducted in India, atleast 15 lakh surgeries are related to diabetic retinopathy.

That said, blindness can be prevented if diabetic retinopathy is detected at an early stage. Detecting eye disorders is a specialized skill and ophthalmologists often have to look at the damage caused to the blood vessels in the eye, and screen other parameters such as fluid leakage or hemorrhages. Can technology help? Technology giant, Microsoft, certainly believes it can make a huge difference, and is experimenting with using Artificial Intelligence technology to identify eye disorders and diseases such as diabetic retinopathy at an early stage.

By training computers to recognize early signs of eye-related issues, Microsoft is looking at building a large ecosystem that has like-minded organizations, which will collaborate and collectively work on diverse datasets of patients across geographies to come up with machine learning predictive models for vision impairment and eye disease.

Microsoft India has already collaborated with L V Prasad Eye Institute to launch Microsoft Intelligent Network for Eyecare (MINE). The partner organizations of this consortium include Bascom Palmer – University of Miami, Flaum Eye Institute – University of Rochester (USA), Federal University of Sao Paulo (Brazil) and Brien Holden Vision Institute (Australia). The availability of a large amount of data together with the means to analyze it and gather insights is transforming business like never before. Having the right infrastructure to gather, analyze, visualize and make sense of the data is therefore going to be critical to all institutions and Artificial Intelligence technologies can help organizations build, innovate and transform their businesses. As the datasets grow in size, the ability of the program to detect eye-related issues will improve significantly.

Explaining the vision behind forming the consortium, Anil Bhansali, Managing Director, Microsoft IndiaAnil Bhansali MD Microsoft India R&D (R&D), says, “In our shared vision to eradicate preventive blindness, Microsoft Intelligent Network for Eyecare (MINE) will help redefine eyecare by bringing together the power of technology and knowledge of global experts. Through Microsoft Intelligent Network for Eyecare (MINE), the partner eyecare institutions around the world will come together to build a pool of patient data from various geographies. This will include the rate of change of myopia in children, conditions that impact children’s eyesight, predictive outcomes of refractive surgery, optimal surgery parameters as well as ways to personalize a surgery and maximize its probability of success.” Over the next few years, the consortium hopes to establish a common set of problems to work on with all the consortium partners. Over the long-term, the consortium aims to have a deeper understanding of the geographically diverse disease patterns and to help tackle avoidable blindness at large.

According to press reports, India currently has 11,500 ophthalmologists against six million people who are affected by the disease. Using Artificial Intelligence techniques can significantly improve the speed, efficiency and accuracy of identifying eye-disorders at an early stage, before they become a major problem. By using machines to perform automated detection, the diagnosis can be made more accurate and efficient – and could prove invaluable in remote areas, where the required expertise is not available.