Designing experiments for medical diagnosis evaluation using membership functions

Training is an important issue in all professions and it is specifically es sential in medicine. The ability to make a correct diagnosis has to be asse ssed as a part of training. The evaluation of the diagnosis ability is usua lly performed based on senior physician diagnosis. Medical diagnosis evaluation in a clinical context is the focus of this cha pter. To solve this problem, one proposes a fuzzy set-based experiment desi gn. The membership functions are built based on the Theory of Psychological Measurement, using scaling methods. The main tasks related to the scaling methods are analyzed, and conditions for extracting the membership function s are derived. The proposed method and the related experiments that can be built upon are aimed to allow the acquisition of medical experience in diag nosis. Although initially the hypotheses are informally stated, by using th e framework of the proposed methodology here, they can be precisely determi ned and then validated.