About >
History
History
Our roots as a pioneering center for medical artificial intelligence (AI) reach back to the early 1970s, when Jack D. Myers, MD, chairman of the Department of Medicine, began collaborating with AI researcher Harry E. Pople, Jr., Ph.D. Together, they developed one of the first clinical decision support systems, initially called DIALOG (for diagnostic logic) and later renamed INTERNIST-1. The system used a knowledge-based, heuristic approach to match patient symptoms with possible diagnoses across the broad scope of general internal medicine, and was more ambitious than contemporaries such as MYCIN, which focused specifically on diagnosing bacterial infections and recommending antibiotics. Randolph A. Miller, M.D., also played a key role in developing the program and its knowledge base.
INTERNIST-1 evolved through several iterations. CADUCEUS, developed in the mid-1980s, could diagnose up to 1,000 diseases and was described at the time as the most knowledge-intensive expert system ever built. It extended INTERNIST-1's reasoning capabilities to better handle the complexity of simultaneous diseases and imperfect data and was eventually commercialized as Quick Medical Reference (QMR), a microcomputer-based diagnostic tool adopted at hospitals and medical schools across the country.
The legacy of that early work continues to shape Pittsburgh's health AI landscape. Major health systems, including UPMC and Allegheny Health Network, now deploy modern AI tools for clinical documentation and administrative tasks, partnering with companies such as Microsoft and Abridge. This region's early investment in medical AI, rooted in the Myers-Pople collaboration, established the foundation for Pittsburgh's ongoing role as a hub for healthcare technology innovation.
Building on this early work, we formalized our commitment to biomedical informatics through a series of institutional milestones. Our National Library of Medicine (NLM)-funded training program, established in 1987 as one of the first of its kind, began with a small cohort of students focused on AI applications in medicine. The Intelligent Systems Program, founded in 1988, served as the home for graduate study and added a dedicated biomedical informatics track in 1991. In 1996, the University established the Center for Biomedical Informatics under founding Director Charles P. Friedman, Ph.D., who expanded the program to students from across the health sciences, information sciences, and public health. By 2001, the center had launched a formal degree-granting program in biomedical informatics, consolidating several earlier degree tracks. The program's growth culminated in 2006 with the creation of our department, formed by merging the Center for Biomedical Informatics with the Centers for Oncology and Pathology Informatics under the leadership of Michael J. Becich, M.D., Ph.D., as founding chair, and Gregory F. Cooper, M.D., Ph.D., as vice chair. Continuously funded by the NLM since its inception, our training program has grown from a small AI-focused initiative into a broad, interdisciplinary department spanning schools and disciplines, reflecting our enduring role as a national leader in biomedical informatics.