07.08.2017

Google Shares AI System With Indian Hospitals To Combat Diabetic Retinopathy

Unfortunately, many of the eye diseases causing blindness around the world are entirely preventable. One such disease is diabetic retinopathy. As the name suggests, this disease only strikes diabetic patients and can lead to serious retinal deterioration over time.

India is just one of many developing nations struggling with high levels of diabetic retinopathy patients. There are simply too many diabetic patients and too few well-trained ophthalmologists get a handle on this disease.

Current estimates show that there are 69 million diabetics in India. It's extremely difficult for diabetics who don't have healthcare and who live far away from big hospitals in cities like New Delhi or Mumbai to get regular eye exams. Almost 50 percent of diabetics in India who have diabetic retinopathy experience severe visual symptoms before they are diagnosed with the disease. By the time these patients reach the doctor's office, there's very little that can be done to reverse the symptoms.

In order to address this public health crisis, Google has offered to help numerous Indian hospitals with a novel AI detection system. Developers at Google have designed a special image-recognition algorithm that can instantly spot lesions on thousands of retinal photographs at an amazing speed. Indeed, some tests show that Google's AI system performs with more accuracy than actual doctors.

So far, Google has been working closely with Aravind Eye Care and Sankara Eye Hospital, two of the largest hospital chains in the Asian nation. The tech giant has just finished two major clinical trials at these hospitals that show positive results for the new AI detection system.

With the positive results from these initial trials, Google hopes to initiate pilot deployments across India in the ensuing years. In particular, Google wants to bring their AI software into underserved rural areas of the country.

The good news about diabetic retinopathy is that doctors can prevent all the major symptoms of the disease if caught early. Indian doctors want to use the AI system to quickly detect and prescribe treatments to the millions of people who already carry diabetic retinopathy in their eyes. Doctors also want to encourage as many Indians as possible to get regular eye screenings to detect this and other potentially blinding diseases.

Despite the great promises this AI technology offers to India's healthcare industry, there's still a great need for more trained eye doctors in the South-Asian nation. Technology can only go so far, and more trained ophthalmologists will be necessary as India's population continues to grow.

Lily Peng, a product manager at Google Research, told reporters that while the AI software is extremely powerful, it can't work unless there's a strong base of eye care professionals all across India. Peng said "Machine learning's true potential will only be realized when deployed in partnership with healthcare providers."

India currently has a population of around 1.27 billion. The only nation more populous than India is China, which has a population of 1.37 billion.


« Back to list