Computing in Medicine

Reviewer: Bruce Ivan Blum

This is one of 23 overview articles in the special “State of Computing” issue of IEEE Computer. This article is written by some well-known workers in the field who begin with the caveat, “A short survey article is, of course, bound to be superficial to some extent and to leave gaps in certain aspects of the subject covered. As you might expect, we cover areas in our fields of expertise, but we also briefly survey activities as a whole.” Unfortunately, the article lacks a framework, and it is difficult for the reader to extract or understand what is obsolescent, commonplace, or research. The weaknesses of the paper are most evident in its opening sections. The discussion begins with the problem of automating medical history taking, and the historic works of Collen et al. [1] and Slack et al. [2] are referenced. This is followed by a discussion of automating follow-up data entry, in which an anatomic pathology reporting system is referenced as an illustration (with no citation). Following this, there are sections on automating instrumentation derived data analysis (e.g., ECG), automatic medical imaging (e.g., digital radiography), automatic image storage and retrieval (e.g., image archiving and retrieval), automating visual observations (e.g., digital microscopy), and automating decisionmaking (i.e., AI). Each section seems to have a single author and there are slight variations in style, the level of detail, and the use of illustrations. There is also a three-page sidebar by Douglas A. Fenderson on computers to aid the handicapped. Because this is a short, multi-author paper, it lacks a clear perspective. It provides a useful survey and catalog of work in selected areas; its citations and suggestions for additional reading will be helpful to persons new to this area. It would be unfair to criticize the paper for what it does not include. However, some additional sources which add a balance to the paper are the issue of IEEE Computer [3] dedicated to this topic, the proceedings of the annual Symposia on Computer Applications in Medical Care [4], and the work of Bronzino [5] and Schwartz [6]. As can be seen from the citations in the paper and this review, there are few good surveys or textbooks available for this very important field. Since, as the authors point out, the health service industry expends more per year than the Department of Defense, that is unfortunate indeed.