Spatial analysis of multi-channel EEG recordings through a fuzzy-rule based system in the detection of epiletiform events.
Type of content
UC permalink
Publisher's DOI/URI
Thesis discipline
Degree name
Publisher
Journal Title
Journal ISSN
Volume Title
Language
Date
Authors
Abstract
A system has been developed which utilises fuzzy logic to perform spatial analysis of the multichannel EEG recording and forms the final stage of a multi-stage system to detect the presence of epileptiform events (EVs) in the EEG. This spatial-combiner consists of a set of 127 fuzzy-rules which define our expectation of the spatial distribution of an EV as measured across a 4- channel bipolar chain of scalp electrodes. A set of probabilities assigned to each channel by the previous ANN-based stage of the EV-detection system are fuzzified into 5 fuzzy variables and the best matching fuzzy rule gives an output of either definite, probable or possible to indicate a detection of an EV on spatial grounds. The system was tested on 8 clinical EEG recordings (7 epileptiform and 1 normal which indicated a sensitivity of 55.3%, a selectivity of 82.0% and a false detection rate of 7.2/hour. These results show a 50-fold decrease in the false detection rate when compared to the performance of the system without spatial analysis, whilst maintaining a comparable level of sensitivity.