Sumit Verma is a board-certified specialist in cardiac electrophysiology and cardiovascular disease with Baptist Health Care in Pensacola. In a recent interview with FLORIDA TREND, he discussed the impact that artificial intelligence (AI) is having on the diagnosis and treatment of heart diseases.

Assistive Tool: “The way I see AI’s role in cardiology is as an assistive tool for busy cardiologists to help screen data and prioritize tasks. We are often overloaded with too much information that can lead to delays in acting upon important information, or that information gets completely buried in volumes of data… Obviously, in heart care many things are urgent and require prompt action. AI can help sort through large amounts of data and prioritize more urgent tasks and improve the accuracy of diagnostic tests.”

Cost Equation: “The growing use of AI in cardiology will increase costs at least initially, and it will have to demonstrate both substantial patient benefits and eventual savings in cost. It can do this by allowing physicians more time for clinical care and less for administrative tasks and charting, etc. Also, predictive AI models can allow for savings by reducing unnecessary testing and possibly invasive procedures that are likely to be futile or unhelpful in a particular patient’s situation.”

Spotting Subtleties: “The computer (with AI software) keeps learning what is right, what is wrong. And then it gets to a point where it can be much more accurate than a human, and secondly, it can develop predictive ability. Then it can start seeing what we can’t even see. For example, there may be several little changes on the electrocardiogram that the human eye can’t even pick up. But the computer, having seen billions of those subtle changes, can say this patient is developing (heart problems).”

Getting Acquainted: “It is important for cardiologists to have a certain degree of familiarity with AI, especially to be knowledgeable about which tasks are being performed and how data is being screened and processed before being presented to the clinician. Cardiologists currently are not required to learn about AI, but I suspect this will change in the future … If you’re going to be smart about what you do, you better know where your information is coming from, how it’s developed and presented to you. So, I’m actually signing up for a course this year to learn more about AI.”