The disease, whose root cause is diabetes, can be detected as well as categorised by just scanning the retina and taking multiple frames of it, Chandan Chakrabarty of the IIT Kharagpur said on Wednesday.
“The software uses data analytics capabilities to automatically compare and analyse retina images of the patient. It can not only tell if the patient has diabetic retinopathy but also provides risk categorisation ranging from low to medium and high,” he said.
Called the “Computer Vision Approach to Diabetic Retinopathy Screening”, the project is being funded by IBM India.
Started in March, 2011, the project is at present being used on a pilot basis at Susrut Eye Foundation and Research Centre (SEFRC).
“The solution is being used in our clinic and the results are very encouraging. In fact, the accuracy level is as high as 92 per cent,” said Anirudh Maity of SEFRC.
Chakrabarty said that the project would take another year to be commercially viable.
“It should take a year to be ready completely. We are using the results generated by the prototype at the SEFRC to enhance its operations. Once operational, the solution will help prevent and cure DR, which has been increasingly spreading across the country,” he pointed out.
