Bringing together an AI network for eyecare to prevent avoidable visual impairment

Nearly 285 million people around the world suffer from some form of visual impairment. Recent medical data indicates that nearly 70% of these cases could have been prevented with early detection and screening. Today, eyecare is set to leverage Artificial Intelligence (AI) and machine learning to help millions of people to predict and reduce the loss of vision.

Refractive errors, such as near-sightedness (myopia) and farsightedness (hyperopia), are some of the most common visual impairments in the world. Myopia is relatively more prevalent and affects a younger cohort. Children and young adults tend to develop signs of myopia early on. If left untreated, myopia could become pathological. Pathological myopia can alter the shape and globe of the eye, leading to acute vision loss.

Although hyperopia isn’t as common as myopia in young children, severe cases could have an equally detrimental effect. High degrees of hyperopia are associated with sensory and visual motor problems that may result in serious visual impairments.

Is a global visual impairment epidemic on the horizon?  

Visual impairments and refractive errors have become a global health issue in recent years. The rate of impairments continues to increase. Studies suggest that 2.5 billion people could suffer from myopia by 2020 and the number is expected to increase to 5 billion by 2050.

“Uncorrected refractive error is the leading cause of visual impairment and one of the major causes of blindness in the world,” says Dr. G.N. Rao, Founder and Chair, L V Prasad Eye Institute.

The magnitude of the problem appears to be on the rise all over the world as per the currently available data. To prevent blindness from turning into a global epidemic, eyecare providers need better tools to monitor cases and detect the ones with a higher risk of turning severe.

AI-driven healthcare network is tackling the challenge

At Microsoft, we are committed to leveraging machine learning and cloud technology to address this rising global concern.  MINE, now part of Microsoft AI Network for Healthcare, is a global network of eyecare providers and academic institutions.

“The utility of artificial intelligence is that it can make these associations and integrate data within seconds or minutes, whereas it would take humans years to aggregate all that data and find patterns within it. While research papers have an N of 30 patients, AI-powered technology can help report on an N of a billion patients,” says Dr. Ranya Habash, Assistant Professor of Ophthalmology, Bascom Palmer Eye Institute and a MINE partner.

Microsoft in partnership with LVPEI, has developed a machine learning approach for prediction and progression of refractive error in children and young adults. The objective is to combine anonymous medical records and therapy data from the above healthcare providers and train cutting-edge machine learning models to predict the progression of refractive errors over a two-year period. With AI and cloud computing, monitoring and predictive analysis can be delivered on a much larger scale.

The Microsoft model is currently integrated and being validated in EMR systems at 174 centers of LVPEI in India. The model was trained on a data set of 335,799 instances from 176,037 patients aged between 0 and 25 years to predict the progression of refractive errors for a period of two years. The specialty of this model is that it predicts the refractive error for a two-year period considering a person’s medical history, eye condition and gender. Accuracy measures showed that the model proved statistically accurate and provided optimal results when used to predict cases of myopia that could become more severe and cause physical changes to the eye’s sphere and cylinder. High myopia can cause blindness for which there exists no effective restorative treatment currently. In this context, the machine learning model that predicts the refractive error will be critical in identifying patients at risk of developing high myopia, when integrated into clinical practice.

“Technological tools allowing predictive analytics and application of artificial intelligence to arrive at these insights for doctors will be of immense help in developing strategies to control the problem. L V Prasad Eye Institute’s partnership with Microsoft will help pave the path for this progress,” Dr. G.N. Rao believes.

When expanded, this program should help eyecare providers, to closely monitor cases of myopia, find the patients who are at high risk of eventual blindness and implement anti-myopia strategies and remedies to prevent loss of vision. This should help bring down the rate of blindness across the world.

The MINE work was presented by Raghu Gullapalli, Executive Director of Emerging Technologies, L V Prasad Eye Institute (LVPEI) of India and Dr. Ranya Habash, Assistant Professor of Ophthalmology, Bascom Palmer Eye Institute; (USA) – recently at HIMSS 2018 - one of the world’s largest health IT conference in Las Vegas.

The work is expected to contribute to refractive error monitoring, application of various available anti-myopia strategies, assessment of response to interventions and prognostication of risk for various refractive errors, especially high myopia. In a first-of-its-kind initiative, the Government of Telangana has agreed to screen children for visual impairments using AI as part of the Rashtriya Bal Swasthya Karyakram program under the National Health Mission.

Paving the way for predictive healthcare

Our progress with MINE has encouraged us to create the Microsoft AI Network for Healthcare. Under this new, broader program, a similar AI network is being developed for Cardiology.

AI networks for healthcare are part of our mission to empower everyone, everywhere through the power of technology. These networks are a testament to the power of human collaboration and technology to improve people’s lives. With the help of our partners in academia and medical services, we can create powerful models that can predict the progression of serious ailments. When this model is integrated into the existing medical infrastructure, we hope to improve the well-being of billions of people across the world.

 

Prashant Gupta is the Director – Microsoft Cloud Division and Lead – AI Network for Healthcare at Microsoft India.