[Detection of EMG artifacts for use in automatic routine analysis of clinical EEG's]

Automatic computerized EEG-evaluation in several stages requires information about whether or not EMG-artifacts are present. Furthermore, algorithms for recognition of EMG-artifacts may be helpful while reproducing low pass filtered EEG-signals. Depending on the kind of application three algorithms for the recognition of EMG-artifacts are proposed, taking into account the different kind of results required as well as the different nature of input data. The algorithms have been constructed under the guideline of numerical simplicity to match the requirements of routine EEG-evaluation. They have shown to be a sufficient tool for any EMG-recognition task arising from computerized analysis of clinical EEG-signals.