Eyeworld

EW NEWS & OPINION 32 March 2017 Insights by J.C. Noreika, MD, MBA interface. The AI software has been taught about iris lesions and has learned from the acumen of experts like Jerry and Carol Shields. 5 This electronic counterpart directs you to refine your history (has it always been there?), your exam (does it stretch when the pupil is dilated?), the tests you order (anterior segment OCT or B-scan ultrasonography?). If necessary, it reminds you to consider transesophageal ultrasonography and imaging to rule out thoracic aor- tic aneurysm. Reaching a diagnosis deductively, a curbside consultation is initiated to determine if treatment is necessary; what works best in the hands of the experts? Injection of nanobots with an integrated scleros- ing agent into the cyst may just be the ticket. No second injection. Incredulous? What if I told you in 1997 that you'd attend Grand Rounds in Philadelphia from a patio in St. Croix? And, virtual and augmented reality was positioned on the proximate horizon. The trends—artificial intelli- gence, nanobots, augmented reality, robotics—are in place. But the "New New Thing" that changes every- thing in 20 years isn't invented yet. By the way, its 76°F with a pleasant sea-scented breeze. Rum punch anyone? EW References 1. Stuart, Annie. ROP Screening and Tele- medicine, Parts I and II: Has its time arrived? EyeNet, March and April 2014. 2. Orellana J. Teleretinal screening for diabetic retinopathy. JAMA Ophthalmol. 2015;133(10):1221–2. 3. Benaroch, R. Telemedicine: An idea with many potential pitfalls. MedPage Today. December 27, 2016. 4. Das T, Pappuru RR. Telemedicine in diabetic retinopathy: Access to rural India. Indian J Ophthalmol. 2016;64:84–6. 5. Shields JA, et al. Iris cysts in children: Classification, incidence, and management. The 1998 Torrence A Makley Jr Lecture. Br J Ophthalmol. 1999;83:334–8. Editors' note: Dr. Noreika has practiced ophthalmology since 1981. He has been a member of ASCRS for more than 35 years. Dr. Noreika would like to thank PGY-2 Lucas Bonafede for his presenta- tion on iris cysts. Contact information Noreika: JCNMD@aol.com erwise. GPS devices (and squirrels) have far greater spatial intelligence. Google's search machine exhibits impressive long-term recall. Emo- tional intelligence? Try one of the Cognitive Behavioral Therapy apps. Machines can be taught and, more importantly, learn. Kelly maintains that, although combining specific forms of intelligence, computers will never consciously drive cars; therein lay their potential. Unlike humans who are easily sidetracked by a cell phone, double cheese burrito, or blonde in a Ferrari, a self-driving automobile is never distracted. But intelligent bots won't replace humans, especially doctors, although radiologists and pathol- ogists may be at risk. Specializing in the interpretation of fluorescein angiograms and optical coherence tomography (OCT) images might suggest a career reset. Bots are remarkably efficient, doctors and other scientists noto- riously inefficient. Creativity oozes from this messy glop of inefficiency; you can see where this is going. Big Blue defeated Garry Kasparov in a chess match. But Big Blue has not defeated an opponent partnering a human with an AI machine. Intel- ligent bots do tasks and redefine work. But greatest value emerges from combining the specialized focus of a cognificating bot with the off-the-wall brilliance of an astute clinician. Performing a slit lamp exam, an ophthalmologist notices a pigment- ed mass at the pupillary margin. What is it? Does it pose a threat? What's next? Imaging. What kind of imaging? What do I need to know now? Let's consult my AI colleague. Employing sophisticated voice recognition technology even better than Amazon's, there is no keyboard nosis of a vitreous floater even in tech-friendly California? Let's try to divine the future. Nanotechnology has expe- rienced growing pains. After the investment of billions of dollars in exchange for the promise of chang- ing everything, researchers admit that building functional structures from nanoparticles—a nanometer is one-millionth of a millimeter—is a lot tougher than it looked. But some, like George Tulevski, use chemistry to do what nature has done for eons: self-assembling mac- rostructures from billions of random molecules. Tulevski's interest is com- puter science, and he reports that carbon nanotubes can speed chips exponentially with game-chang- ing effect. In medicine, miniscule nanobots may carry drugs, geneti- cally modified particles and other material through the bloodstream to fight disease, inflammation, and neoplasm. The eye's vitreous is a compelling medium. Speed amplifies computing power, and computing power facilitates cognification (i.e., artifi- cial intelligence). Kevin Kelly, the founder of WIRED magazine, sees machine cognification as the next Industrial Revolution. He maintains that humans are not very good at defining intelligence because, as it turns out, there are many different kinds of intelligence. He refers to human intelligence as "a symphony of different notes and each of these notes is played on a different instru- ment of cognition." Each person is allocated a specific, perhaps unique, combination of intelligences. Machines focus on one or a few of these. But they do so extremely well. Calculators have more mathe- matical intelligence than all but the greatest human savants, idiot or oth- Distance learning in real time is a mere piece of what portends for the future of medicine and ophthalmology A n ophthalmologist I know attends the weekly Chiefs' Rounds presented by Wills Eye Hospital, Philadelphia. A commu- nity-based generalist, he enjoys the scholarly stimulation of unusual cas- es elaborated by enthusiastic house staff and evaluated by a faculty that writes the books. He ponders clinical information new to him or remotely consigned to recesses of his cortical gyri. Pupillary margin cysts can be congenital—he knew that—and also associated with thoracic aortic aneu- rysms. Great board question! What makes his situation un- usual? Earning Continuing Medical Education credit with laptop and headset, he sips coffee while surf gently kisses the beach. He has signed in to Wills' live webinar from St. Croix, USVI. Technology at its most friendly. Webinars are yesterday's news relative to utility and ubiquity. Microsoft introduced NetMeeting in 1996. With the maturation of the internet, lightning-fast broadband, and advanced chips, participation in the Chiefs' Rounds is quotidian. What might we expect in the next 20 years? Trends are already in place. Telemedicine is no longer a trend; it's here. I believe it will have limited application, a niche. In oph- thalmology, it has been touted for screening retinopathy of prematuri- ty 1 and diabetes. 2 Maybe. It still has technical and logistical limitations 3 and, at least in the United States, its medico-legal ramifications remain murky. It may offer advantages in underserved countries such as India 4 where the supply of vision care is vastly outstripped by demand. That is not the case in, say, San Diego; is it likely that a patient will resort to telemedicine to confirm the diag- Staying current, getting smarter J.C. Noreika, MD, MBA

• Blank Page
• Blank Page (1)