I. Present Concepts of the Meaning and Limitations of Medical Diagnosis

Abstract

For the past six years a small group of physicians, electronic engineers, mathematicians, logicians, and computer programmers has been meeting regularly to explore ways in which modern electronic computers might aid physicians in diagnosis. Early in the discussions several interesting problems became obvious. First, before any real progress could be made, it would be necessary to define terms more precisely than previously required. Second, because the definitions rest on certain philosophical and historical issues, there was a need to review these fundamental disciplines and their relationship to diagnosis. Finally, rapid changes in our concepts of diagnosis require that we try to visualize diagnosis of the future. Only after these considerations can we attempt a realistic evaluation of the role computers might some day play as aids to diagnosis. We are attempting here to explore these questions so that both physicians and electronic engineers may have a common background for their