I recently spent an afternoon working from my local coffee shop. As I pulled out my laptop, I couldn’t help but think how fast the year had gone by. I was there to prepare for a course I first taught in 2017 at the Joint Commission on Allied Health Personnel in Ophthalmology (JCAHPO). It was called “Using Artificial Intelligence to Improve Retina Care: The Future Is Now.” One thing that struck me was the staggering progress that had been made in a AI in the short 12 months since I first taught the course. I resisted the temptation to change the title to “The Future Was Yesterday.”
AI is indeed coming to your clinic, and it is doing so at lightning fast speed. For instance, RETINA-AI has just developed and released Fluid Intelligence, the world’s first mobile AI app for eye care providers, capable of detecting macular edema and subretinal fluid on OCT scans. And earlier this year, the FDA issued the first approval of a diagnostic AI device in medicine, IDx-DR, for use in primary care settings as an automated diabetic retinopathy screening tool. And these two developments are just the beginning. One potential source of worry for many people – not just in ophthalmology but across all industries – is how AI will affect the workforce. “Will I lose my job to AI or to a robot? If a robot can someday perform cataract surgery flawlessly, how will such a development affect my income?” These concerns are understandable. Some AI proponents swear that there will be no changes to the healthcare workforce in the age of AI. This is not true. Some AI antagonists, on the other hand, swear that AI will lead to massive job loss and overall apocalyptic change. Also not true. Undeniably, AI will change the way we care for our patients. It will indeed eliminate the need for humans to perform certain types of healthcare tasks; however, it will also create a need for new healthcare tasks that can only be done by humans. It is no wonder some are calling AI the fourth industrial revolution. The first was steam-powered, the second was electricity-powered, the third was information technology and internet-powered, and now the fourth is AI-powered. Just like the three prior, this revolution also represents advancement in human technological capacity, which is generally a good thing.
Of note, the development of AI systems is necessarily a ‘cottage industry,’ which requires direct input and direction from human experts — indeed, the use of AI systems within ophthalmology will always require oversight by ophthalmologists. For instance, though AI can now diagnose macula edema and a number of other conditions from OCT scans, an ophthalmologist is still needed to confirm the diagnosis and to make the final treatment decision. We live at an exciting time in history. We are entering the era where ophthalmic care will be driven by ophthalmologists but enhanced by AI. This time presents the opportunity to both build and use revolutionary AI technology to attain unprecedented benefits for our patients. It is a time of smarter diagnostics, smarter treatments and smarter AI-enhanced physicians.
