Medicine

This project at the Texas College of Osteopathic Medicine (Fort Worth) evaluated the use of an artificial-intelligence-derived measure, "Knowledge-Based Inference Tool " (KBIT), as the basis for assessing medical students ' diagnostic capabilities and designing instruction to improve diagnostic skills. The instrument was designed to address the problem that, in medicine, diagnostic expertise is problem-specific and appears to be more a factor of the student's knowledge base than cognitive skills. This study determined that the KBIT produced reliable and valid (based on comparisons of diagnostic accuracy of experts with those of novices) for four different problem areas: Weakness, Red Eye, Papulosquamous Disorders, and Elevated Creatinine. Additionally the study showed that two expert/KBIT-derived instructional approaches significantly improved the diagnostic accuracy of treatment student groups when compared to a control group and to students conventionally trained. After the executive summary and a project overview, this report describes the project's background and origins, its components and activities, and results. Attached is a related article titled "An Exper