Eyeworld

EW NEWS & OPINION 26 August 2017 point of care—will "drive strategies to tackle avoidable blindness and help increase efficiency in the deliv- ery of eyecare worldwide." "Healthcare is on the cusp of ap- plying such advanced analytics and [artificial intelligence] to improve patient care and operational effi- ciencies," Mr. Bhansali said. "With MINE, we aim to bring technology and healthcare together to help reduce avoidable blindness." Though data is being contribut- ed currently from a few major insti- tutes around the world, Dr. Habash said the findings will be applicable to every ophthalmologist for every treatment modality that's studied. "Everyone will benefit even though we're the ones doing the legwork. Whereas research papers have an N of 30 patients, we'll be able to report on an N of a billion patients," she said. "We are developing the eyecare of now," Dr. Das said. "The eyecare of now is a philosophy where we are trying to bring application and insight in real time to people to whom it matters. … The whole idea is you can make a difference on a real-time basis. That's very much possible right now with the current technology and solutions we have. That is where MINE hopes to make a difference." EW Contact information Bhansali: anilb@microsoft.com Das: vipin@lvpei.org Habash: ranya@hipaachat.com by Liz Hillman EyeWorld Staff Writer it can make these associations and integrate data within seconds or minutes, whereas it would take hu- mans years to aggregate all that data and find patterns within it." Dr. Das said he sees us as being in the "KMB era"—thousands, mil- lions, billions. Currently, published literature, for the most part, is limit- ed in terms of number of people to the thousands. "The volume of research we are able to do today is limited in terms of amount of people that present in each of these centers," he said. We are moving into the mil- lions, though, he said, as more multicenter and combined research is being done. "Right now, data on refractive error is close to 1 million," he said, for example. "We will get much stronger insights because of the sheer volume." Then, Dr. Das continued, we're on a shared mission to bring analy- sis into the billions. "Hopefully, 3, 4, 5 years from now, we will build our first billion data point. That will only happen when you use technology to pool this information in almost real time," he said, expressing that he hopes this is what MINE will help facilitate. "In that context, the KMB era will be a very powerful transition moving forward in our profession." Mr. Bhansali said three steps— aggregating data from patterns, automating insights with machine learning and artificial intelligence, and applying these insights at the Microsoft India and L V Prasad started working together on ad- vanced analytics in eyecare in 2015, using the Microsoft Azure Cloud platform and Power BI technology to map diseases geographically and in real time. Vipin Anthony Das, MD, L V Prasad Eye Institute, said they used their data from the 165 L V Prasad centers in the country and Micro- soft analytics to predict refractive error in children as well as refractive surgery outcomes. The latter has allowed the institute to identify re- fractive surgery patients predisposed to regression, who might be at risk, and what the appropriate procedure might be for the patient. "The differentiating factor that MINE wants to bring, and is already doing, is applying the knowledge gained from [real-time machine learning and data analysis] at the point of care," Dr. Das said. "The point is to take insights and help the physician and the surgeon to make better decisions." Ranya Habash, MD, Bascom Palmer Eye Institute, said the data put into the system from different sources allows machine learning to make higher-order connections and come up with different treatment algorithms. The algorithm becomes more structured with more data. "We can learn from long-term patterns in our treatment," Dr. Habash said, noting that data from both the past and present are being contributed to the system. "The utility of artificial intelligence is that The intention is to apply analysis from large datasets at the point of care to eliminate avoidable blindness A rtificial intelligence is continuing to make its way into ophthalmology with a relatively recent initiative between Mic- rosoft and several leading eyecare institutes around the world. What started as a collaboration between Microsoft and L V Prasad Eye Institute, Hyderabad, India, has expanded to include other part- ners—Bascom Palmer Eye Institute at the University of Miami, Flaum Eye Institute at the University of Rochester, New York, Federal Univer- sity of Sao Paulo, Brazil, and Brien Holden Vision Institute, Sydney, Australia—announced in December 2016 as the Microsoft Intelligent Network for Eyecare (MINE). "Technologies like machine learning (ML) and artificial intel- ligence have the power to harness data and build predictive models. In the world of ophthalmology, such technology platforms can help predict the gravity of an eye disease and its potential outcomes, enabling the doctors to pinpoint the most accurate procedures needed to prevent and treat the impairments," said Anil Bhansali, corporate vice president, Cloud & Enterprise, and managing director, Microsoft India (R&D) Pvt. Ltd., Ahmedabad, India. Microsoft and international group of eye institutes team up to develop real-time machine learning Microsoft India, L V Prasad Eye Institute, and others have partnered to create the Microsoft Intelligent Network for Eyecare, which uses Microsoft's Azure platform and Power BI technology to map diseases geographically and in real time. Source: Microsoft

• Blank Page
• Blank Page (1)