Automatic eeg spike detection by use of adaptive decision criteria to individual eeg records

摘要

This paper proposed a new method for automatic detection of spikes and sharp waves (SSWs) in electroencephalogram (EEG) records. All the items which were necessary to express the features of SSWs in visual inspection of EEG records were first determined, and the mathematical models were constructed for detecting the SSWs in EEG. The proposed method was evaluated by applying it into 71 EEG records of 9 subjects, and was in a good agreement with the electroen-cephalographer's visual inspection. The proposed method is based on adaptive decision criteria, therefore it can adapt to the various waveforms with SSW of respective subjects automaticaly. It will be expected to be adaoted in the medical field as an assistant tool for electroencephalographers.