ChatGPT Shows ‘Remarkable Ability’ to Process Thoracic Surgery Data, Could Help with Training and Patient Care

A lab focused on how artificial intelligence technology can improve the treatment of cardiovascular disease has tested the ability of ChatGPT to answer assessment questions from the American Board of Thoracic Surgery. ChatGPT-4 achieved an accuracy of 81.3 percent. The results prove that the AI model shows the ability to understand complex thoracic surgical clinical information, which the authors suggest could be developed into personalized learning platforms for trainees.