[The assisted-diagnosis by identification program, an expert system for acquired ocular hypertension in adults].

Computer Aided Diagnosis by Identification packages are expert systems which proceed by iterative comparisons sign by sign between patient's cases and data encased in their clinical memory. They draw straight forward inference from thought processes analyses that lead to diagnosis: comparison between patient's own case report and acquired identified situation, error factors: ignorance or non-performance and, lastly rules to subscribe to: perfect knowledge of every kind of glaucoma and for every kind, perfect knowledge of every sign. Formalization of the problem specifies the algorithm, the miscellaneous solutions: single hypothesis and multiple hypothesis, and their respective values. In order to build up the clinical library, literature search is required to be comprehensive: treaty, reviews, papers, communications and personal case-reports. Outputs are retrieved in an average time-lag of 15 seconds, and are expressed as a list from which only the physician can make out the meaning.