Hot Links
EMG - Knowledge-Based Assistant for ElectromyographyESPRIT I, Project 599, First Framework Programme
Achievements: The aim of the project was to develop a knowledge based assistant to support physicians in all stages of an electromyographical (EMG) examination of patients with neurological diseases. The objective was to produce a system sufficiently robust to withstand clinical trials in a neurophysiological laboratory. Particular attention was given to involving users in the definition of requirements and in system acceptance testing, and to bringing medical knowledge based systems to a fully functional state. The prototype EMG expert system supports the diagnostician in the analysis of EMG signals and advises on the test procedures to be performed. It includes a report generator, and contains a database of case studies. It incorporates a casual probabilistic network model to allow a unified approach to planning, diagnosis, explanation and reporting. The following major features were developed subsequently: robust inference systems; new ways of handling uncertainty by probabilistic methods; and methodologies of general applicability in knowledge representation, blackboard architecture, and user interface specification. General information: The prototype EMG expert system constructed in Phase I supports the diagnostician in the analysis of EMG signals and advises on the test procedures to be performed. It includes a report generator, and contains a database of case studies. It incorporates a causal-probabilistic network model to allow a unified approach to planning, diagnosis, explanation and reporting. Phase II saw a substantial improvement in real-time performance and the development of the following major features: Project Details:Duration: 63 months Coordinator: Partners: |