AI in Healthcare, Trusting Your Microwave, and Colonizing the Moon

In this episode, our hosts interview Dr. Eric Eskioglu, neurosurgeon and Chief Medical Officer of Novant Health about how his organization stays forward-thinking in artificial intelligence (AI) developments and promising AI healthcare applications of the near future.

Dr. Eskioglu describes his first career in aerospace engineering, which sparked his original interest in AI. He recognized the value of data analytics and how it could be used to plan out future scenarios. After making the shift towards medicine, Dr. Eskioglu found himself being drawn to translational research and neurosurgery.

Today, he describes his role within Novant Health as an interpreter between physicians and administrators, and also a champion of behavioral economics and AI innovation within the health network. He emphasizes that focusing on improving quality and safety measures is the key to gaining organizational and funding support for AI projects. A large focus of Novant Health has been the use of AI to recognize and reduce non-beneficial variations in clinical practice between different providers.

Next, we shift to discussing how AI can be used to equalize healthcare access across different communities. Dr. Eskioglu gives the example of Viz.ai, a stroke detection system that improves diagnosis and response team efficiency, no matter where the patient and medical team are located. (If you’re interested in hearing more about Viz.ai, check out Episode 7, in which we interview its founder, Dr. Chris Mansi!)

Dr. Eskioglu also discusses how AI can help prevent physician burnout. For example, Aidoc is currently using image scrubbing to stratify ER images based on clinical severity, which effectively reduces anxiety and burnout for radiologists. He also mentions how natural language processing that automatically generates progress notes as physicians converse with patients. Furthermore, he speculates that AI will have the largest impact on family medicine, since this field has frequent patient contact and an abundance of patient data collected over time that can be used to glean patterns to guide clinical practice.