Abstract
Medical consultation systems were among the first to prove the feasibility of knowledge-based artificial intelligence approaches for the interpretation of clinical data. Yet these consultation systems, despite impressive performance results, have not been followed by widespread commercially available systems. One of the reasons may lie in their complexity, which makes evaluation and validation of their knowledge bases more difficult, time consuming, and expensive. Another is the stand alone nature of most of the systems. Most practical systems need to be integrated into the practice of medicine to succeed, and the best way of doing this is to incorporate them in instruments. At least one expert system for biomedical signal interpretation has been widely disseminated, that for serum protein interpretation. This paper will outline the experience gained in building the system from knowledge acquisition through prototype testing to final implementation. It will also discuss issues involved in the possible regulation of medical software. Concern with regulation may well be a factor in the scarcity of other decision support and AI systems produce commercially to date.
| Original language | American English |
|---|---|
| Pages | 20 |
| Number of pages | 1 |
| State | Published - 1986 |
ASJC Scopus subject areas
- General Engineering