Medical AI: How to raise money and eventually make money

Dr. Michael Abramoff, the founder of an AI medical diagnostic device company, recently appeared on the NEJM AI Grand Rounds podcast. The interview covered a range of topics, but what resonated with me were his experiences raising capital and navigating the reimbursement medical AI landscape.


At the 16 minute mark, Dr. Abramoff discussed fund-raising, first describing the approaches he tried that didn't work:

  • Traditional funding routes such as the National Institutes of Health (NIH) were not viable because they do not fund the interactions with the FDA or the cost of expensive consultants.
  • The pharmaceutical model of patenting the AI algorithm and waiting for a large tech company like IBM or Google to invest didn't work. Despite patenting his AI algorithms, there was no movement from these large companies.
  • Raising money through philanthropy was also not successful. While philanthropic donations are often made to causes such as helping children with inherited eye diseases, the cause of generating paperwork for ISO certification and FDA approval was not as compelling to potential donors.

Dr. Abramoff pivoted towards angel investors and venture capital for funding when traditional methods failed. This decision led to the founding of his company, Digital Diagnostics. Success followed, with the company eventually securing its first growth equity investment, marking a significant shift in the perception of AI in healthcare. This desire to meet high safety standards was a driving factor in this commercial journey.


At the 32-minute mark, they discussed the complex topic of reimbursement, first describing some of the challenges:

  • Charging based on the marginal cost of a diagnosis made by AI was not sustainable. The cost of R&D is too high to be covered by such small returns, making this model a money drain and unattractive to investors.
  • Cost-effective analysis, another suggested method, would mean charging just below the threshold at which a service remains cost-effective. However, this would make the service more expensive than it currently is, which was against Dr. Abramoff's goal of improving access and lowering costs.
  • Legal complications associated with reimbursement are considerable, with serious penalties for missteps. The Social Security Act only allows the Center for Medicare and Medicaid Services (CMS) to reimburse what a physician charges, adding complexity to the process.

Dr. Abramoff instead developed the 'equity enhancing payment' concept, charging based on society's current willingness to pay for 20-30% of patients receiving these exams. The goal was to use the same amount of money to cover 100% of patients, which translated to a rate of $55 per service. After many meetings with payers and discussions about what constitutes an appropriate payment, this model was established as a viable path forward. Despite the complexity of the reimbursement process, this approach aims to decrease healthcare costs and improve accessibility, particularly for those currently underserved by the healthcare system.

Feedback welcome: Twitter and LinkedIn

Full Podcast

No Doctor Needed? Dr. Michael Abramoff on the Potential of Autonomous AI | NEJM AI Grand Rounds
Dr. Michael Abramoff is a renowned ophthalmologist and medical AI pioneer. In this episode, we explore his groundbreaking work that led to the first FDA-authorized device that does not require a physician, IDx-DR, which detects more than mild diabetic retinopathy from digital images of the eye. Dr.…

Additional References

A reimbursement framework for artificial intelligence in healthcare - PubMed
Responsible adoption of healthcare artificial intelligence (AI) requires that AI systems which benefit patients and populations, including autonomous AI systems, are incentivized financially at a consistent and sustainable level. We present a framework for analytically determining value and cost of…
A framework to evaluate ethical considerations with ML-HCA applications— valuable, even necessary, but never comprehensive

Subscribe to Ian Maurer's Notes

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.