Eyeworld

20 Ophthalmology Business • December 2017 that play chess and can beat the world's best chess players. There's an expert system with a general knowl- edge base that has beaten the best Jeopardy players in the country. I think it's a natural extrapolation— can we build better machine learning systems that can help doctors make a better diagnosis?" Dr. Chiang said. AI in medicine The system that defeated the best Jeopardy players is IBM's Watson, a supercomputer that in the health field not only stores medical infor- mation, but finds meaning in it—in other words, learns. Michael Abramoff, MD, PhD, Department of Ophthalmology and Visual Sciences, University of Iowa Hospitals and Clinics, Iowa City, Iowa, has been working on AI in medicine for the past 30 years. The retinal specialist explained that there are different types of AI or machine learning systems, depend- ing on the goal. It can be used in discovering new associations or for diagnosis. "Let's say you have a bunch of images or data and you want to see whether that predicts some disease and what can be helpful; you can use machine learning or AI to discover the associations," Dr. Abramoff said. "Our partner, IBM's Watson, uses AI a lot for scientific discovery," he said. "They say, there's all this scien- tific literature out there, so let's build an algorithm that can analyze the literature and advise on the condi- tion [according to] what is the most valid evidence from the scientific literature." Google (Mountain View, Cali- fornia) is also investing in AI for the Chiang, MD, Department of Oph- thalmology, Casey Eye Institute, Oregon Health & Science University, Portland, Oregon. "Since the 1970s, there have been expert systems that have been shown to perform as well as or better than experts in certain areas of medical diagnosis." He continued, "Some of the early work was done in infectious disease; for example, there was a system called MYCIN developed by Edward Shortliffe at Stanford University." The system identified causes of meningitis and performed as well as or better than infectious disease specialists, Dr. Chiang said. Since then, technology has made tremendous advancements and along with it, computing power. "In the real world, there have been computer systems developed Exploring the advancement of artificial intelligence in ophthalmology and what it means for clinicians A rtificial intelligence (AI) is a subject ap- proached by many as being a thing of the future, in continual development, but not yet developed. Depending on which side of technol- ogy one stands, relevant discourse on the matter is accompanied either by excitement about the possibilities or trepidation about the inevitabilities. Talk to subject experts about it, however, and they will often tell you that it's been around for a long time. "The concept has been around since the 1950s," said Michael F. by Ophthalmology Business Contributing Reporter Robot doctors?

• Blank Page
• Blank Page (1)