The role of perception in imaging: past and future

Abstract

The accurate and efficient interpretation of medical images relies on a host of factors. Clearly the technologies and methods used to acquire, process, transmit, store, and display the image and associated data are critical, but they are only one-half of the equation. In the end, the final diagnostic interpretation and recommendations for further action lie with the clinician. Ideally we would like to believe that all decisions rendered by competent clinicians are correct, but the interpretation task is not always easy or black and white. Thus, decisions are not always absolutely conclusive, are often formulated with plausible alternatives, and errors in interpretation can and do occur regularly. The discipline of medical image perception seeks an improved understanding of the perceptual factors that underlie the creation and interpretation of medical images, with the belief that improved diagnostic performance with the use of imaging devices can be achieved by the development of systems that are optimized for the interpretation of visual diagnostic information. Perception research can identify specific reasons for missed diagnoses, thereby helping to train physicians and eliminate diagnostic errors, and clarifying situations in which errors are a consequence of fundamentally ambiguous information rather than poor reader performance. The goal of this article is to provide a short review of the history of the discipline of medical image perception, highlight key research areas, and provide a look toward the future regarding the role that medical image perception research will continue to fill as imaging technology in medicine advances and develops.