In the Winter 2024 issue of the Retina Times, Digital Diagnostics founder and executive chairman Dr. Michael Abramoff discusses his professional journey, the present and future of AI, and his candid advice to researchers and aspiring researchers and health tech entrepreneurs.
The full, in-depth interview is available to members of the American Society of Retina Specialists. Below are some takeaways from his conversation with Dr. Eliot Dow.
We chose the entrepreneurial path for LumineticsCore™ because that was the only option.
Before starting Digital Diagnostics and developing LumineticsCore, Dr. Abramoff assumed he would follow the typical path for academics working in the healthcare space: He’d publish his research then license the new technology to a healthcare company which could commercialize it.
The problem was, LumineticsCore wasn’t “typical.” When researchers develop a new drug, there’s a clear framework in place for the FDA to approve it and for pharmacies and hospitals to distribute it. That framework didn’t exist for AI when Digital Diagnostics entered the market with LumineticsCore. Dr. Abramoff and the fledgling company worked with the FDA to pioneer a framework to obtain clearance, then worked with healthcare systems to distribute the system. None of this could have happened if they had just sat back and waited for existing health tech companies to handle it.
Autonomous AI and assistive AI are both valuable to healthcare, but autonomous has the edge when it comes to expanding access.
When people talk about AI, they’re often referring to assistive AI: It helps people to do something, such as assisting a physician with reading a scan, but the physician is ultimately the one who makes the diagnosis. Autonomous AI goes one step further, by offering an actual diagnosis on its own.
Both assistive and autonomous AI can play important roles in improving access and accuracy, but since autonomous AI can operate without a physician, it has the potential to bring testing to even more people, more conveniently.
The key to winning public trust in healthcare AI is listening with empathy.
Back in 2010, an editorial jokingly called Dr. Abramoff “The Retinator” and predicted his research would lead to lost jobs and decreased accuracy. He knew that wasn’t the case, but that moment made him realize the importance of being proactive in speaking with AI’s critics.
Peer-reviewed studies and publications are an important part of countering this skepticism. These studies’ transparency, as well as the work that goes into developing these AI systems, can go a long way in building trust in these new systems.
However, the trust-building work doesn’t stop there. Researchers need to maintain an ongoing dialogue about all the benefits AI can bring to patients and physicians alike, as well as responding with empathy and honesty to the concerns critics raise.
Center patients, justice, and autonomy in your ethical framework—and quantify it.
When researchers work with AI, medical ethics encompass numerous considerations, from justice to autonomy to beneficence. Balancing all of these principles can be difficult, but by employing “metrics for ethics,” as Dr. Abramoff calls them, researchers can quantify each and apply those calculations to different problems and solutions to find the most ethical path forward.
Tips for would-be entrepreneurs: Focus on patient outcomes, consider business-side obstacles.
One piece of advice Dr. Abramoff gives is to center everything around patients as you’re developing an AI system. It’s easy to fall into the trap of what he calls “glamour AI”—tools that seem cool but ultimately don’t improve patient outcomes. If you can’t explain how your system is going to lead to a better patient outcome, that’s a problem.
Additionally, Dr. Abramoff has found the business side is even more complex than researchers tend to assume, and that can have big implications for whether and how their work makes it to the front lines of care. From reimbursement policies to regulations to basic medical coding, a wide range of factors affect the cost and uptake of medical innovations. Understanding these issues, and approaching them with resilience, is key to overcoming them.
What’s next?
Over the last 15 years, Digital Diagnostics has played a crucial role in advancing diagnostic AI, from helping to pioneer the FDA clearance process to establishing a blueprint for bringing the technology to patients. The future will bring new opportunities, but Dr. Abramoff’s guiding light will remain the same: Delivering patient benefit and accurate, accessible testing to as many patients as possible.
As Dr. Abramoff notes in the interview, “Every step of my journey, from pursuing my career as a retina specialist, to designing machine learning algorithms, to founding Digital Diagnostics, has been driven by a desire to use my unique skill set to help improve the patient and provider experience in health care.”