Ophthalmology on the brink of digital innovation in AI, machine learning
Click Here to Manage Email Alerts
Common sight-threatening retinal diseases seen by all comprehensive ophthalmologists include wet age-related macular degeneration, geographic atrophy and diabetic macular edema.
Reviewing several sources, about 20 million Americans aged 40 years and older have AMD. Most sources suggest 10% have advanced AMD, with approximately 45% being wet AMD, another 45% GA and 10% a combination of wet AMD and GA. Another 3.8% of adults aged 40 years and older have vision-threatening DME. Seventy percent of the U.S. population is aged 40 years and older, or about 233 million people. So, as I do the math, approximately 11 million Americans have wet AMD, GA or DME.
Vision loss is the chief complaint that brings most of these patients into our offices. Many can be diagnosed with a careful fundus examination, but more and more we rely on OCT to help us precisely diagnose these patients, formulate and personalize their treatment plan, and decide whether or not to refer them to an ophthalmologist skilled in the art of intravitreal injections, usually a vitreoretinal specialist.
Ideally, interpretation of an OCT requires a review of 24 or more radial cuts across the macula. This is a demanding task, and most of us simply review a single radial OCT cut image. Digital innovation in the form of artificial intelligence or machine learning is especially well suited to help us better interpret OCT findings in the macula, analyzing 24 or more OCT cuts rather than one. Much like AI will provide a preliminary diagnosis to a cardiologist reviewing an EKG or a radiologist reviewing a chest X-ray, AI is especially useful for analyzing and interpreting images.
The eye care professional (ECP) remains responsible for the ultimate diagnosis and personalized treatment plan, but AI can save the ECP significant time and enhance the accuracy of diagnosis and effectiveness of patient management. Ophthalmology has been a laggard in the adoption of AI, but our specialty is well positioned to benefit from its capabilities, as we analyze and interpret many images. The accompanying cover story suggests to me that digital innovation, AI and machine learning are destined to benefit nearly every ECP and their patients —another example of the tremendous power of the innovation cycle.
Collapse
Click Here to Manage Email Alerts