Takeaways from the One-on-One with Jack Krouse, featuring Anand Devaiah, MD

by Christopher Vickery, MD, FAAOA

In a thought-provoking discussion on artificial intelligence (AI) in medicine, Dr. Krouse and Dr. Anand Devaiah explore the growing influence of AI in healthcare. Dr. Devaiah is an otolaryngologist specializing in skull base surgery at Boston University with a background in computer science ranging from writing code to his involvement in start-ups in the technology space.  He defines artificial intelligence as a machine or system designed to mimic human intelligence, with the goal of becoming more sentient. He emphasizes that AI is not here to replace clinicians but to augment their abilities, allowing for more accurate diagnostics and streamlined workflows. 

Dr. Devaiah highlights how AI applications are already a part of the landscape in patient care. He discusses the integration of AI into electronic medical records (EMR), revenue cycle management, and patient follow-up. He reports that radiology and histology are using AI forimage processing, chatbots are able to interact with patients and allow for asynchronous communication, and AI systems analyze documentation to help craft better notes or more efficient coding.  AI can identify inefficiencies in medical practices, automate data analysis, and reduce manual errors, ultimately improving patient care. 

The discussion continues with the challenges created by the widespread adoption of AI in healthcare. Maintenance of HIPAA compliance and ensuring patient data privacy are at the heart of some of these obstacles. Dr. Devaiah warns that it is essential to understand the data security measures taken in the programming and teaching the models when integrating AI tools.  If a vendor cannot explain the data protection, then he advocates against using their software.  Dr. Krouse recognizes the possibility of bias introduced into the analysis and the challenge of the digital divide leading to the exacerbation of disparity in healthcare.   To better navigate these challenges, Dr. Devaiah emphasizes the importance of physician education and clinical collaboration as AI’s role in healthcare continues to evolve and expand.

The conversation moves to the role of AI in medical research and publication. The lack of technical expertise in this field presents difficulties with evaluating AI-generated content in medical research. The possibility of ”hallucination“ by the AI system is highlighted as a challenge in implementation. Dr. Devaiah shares an example in which the intelligent agent he queried presented an article that he had authored but misrepresented the paper in an erroneous way.  Dr. Krouse advocates for understanding how the AI model is trained and ensuring it reflects diverse, real-world data. The speakers advocate for transparency and patient engagement by encouraging patients to discuss the information they find as they use this technology.  

In conclusion, the conversation underscores AI’s potential in healthcare while acknowledging the ethical, practical, and technical challenges ahead. By fostering trust, transparency, and collaboration, healthcare professionals can harness AI to improve patient outcomes while ensuring equitable and responsible integration into medical practice.