KNOWLEDGE-BASED APPROACH TO QUALITY-CONTROL IN DIAGNOSTIC-RADIOLOGY

The design and prototype implementation of a Knowledge Based System (K BS) devoted to the Quality Control (QC) of X ray imaging (mammographic ) equipment is briefly reported. Following the steps in KBS developmen t reported in the Artificial Intelligence literature a conceptual anal ysis of the problem domain is first performed, identifying the relevan t objects, concepts and their relations, as well as the tasks/goals in volved in current practice. In particular, plans are individuated and explicitly represented that are sequences of relevant actions followed by the experts in QC practices. The computable form of plans constitu tes the core part of the KBS. One of the plans developed aims at diagn osing malfunctions of the mammographic equipment on the basis of the d ata which are daily collected during a routine QC programme, i.e. a pe rformance rest with a phantom and the sensitometric strip.