GAIT-ER-AID: An expert system for diagnosis of human gait

A general knowledge-based system for diagnosis of human gait pathologies is presented. Gait analysis is a non-trivial medical diagnosis problem which involves a vast amount of data and complex reasoning. By capturing an expert's diagnostic knowledge compactly in diagnostic reference frames, the system expedites decision making from large and heterogeneous data sets. The variability of the data as well as its complexity dictated a ‘divide and conquer’, modular inferencing strategy, using only partial data and pruning the tree of potential diagnoses at each step. The concluding section of the paper presents our plans for augmentation of the system by making use of model-based qualitative reasoning to handle cases not currently in the knowledge base.