Reflections on building medical decision support systems and corresponding implementation in diagnostics shell D3
We take a closer look at the medical environment in which decision support systems will have to operate and which ultimately determine their success of failure. Based on the experience accumulated in ten years of active involvement into research revolving around the construction of expert systems, we put forward for discussion a couple of judgements including representation of symptomatic detail, temporal reasoning, case data validation, rule syntax, intermediate conclusions, modularity, test indication/sequence and domain choice. For each of the items touched we describe how our preferences and conclusions have been implemented in the diagnostics shell D3 and thereupon-based diagnostic systems.
KeywordsDecision Support System Bone Marrow Failure Temporal Reasoning Diagnostic Task Puncture Maximum
Unable to display preview. Download preview PDF.
- (1).Why Doctors? The Economist, December 10, p.93–94, 1994Google Scholar
- (3).Puppe F, Poeck K, Gappa & others: Reusable Components for a configurable diagnostics shell (in German); Künstliche Intelligenz 2:13–18, 1994Google Scholar
- (4).Puppe B: Building a medical knowledge base: Tricks facilitating the simulation of the expert's reasoning. AIME '93: 5th Conference on Artifical Intelligence in Medicine Europe, ProceedingsGoogle Scholar
- (7).Kassirer U, Gorry A: Clinical problem solving: a behavioral analysis; Annals of Int. Me. 89:245–255, 1978Google Scholar
- (8).Puppe B, Puppe F: Standardized forward and hypothetico-deductive reasoning in medical diagnosis; Proc. of MEDINFO-86, p 199–201, 1986Google Scholar