Knowledge based support for EEG recording

The authors have developed a real-time signal analyzer, which, by monitoring EEG signals, can provide the optimal control of the parameters of the EEG recorder. The signal processing scheme and the implementation issues of the analyzer are described, with emphasis on the knowledge-based subsystem. The first phase of the signal processing is a feature extraction process, which determines selected parameters of the sampled input signals. The second phase is a decision-making process, based on the actual parameter set, which produces the necessary control activities. The decision making is carried out by a forward-chaining inference engine using knowledge in rule-based form.<>