DETECTION OF MUSCLE ARTIFACT IN THE NORMAL HUMAN AWAKE EEG

Objectives: A study was performed to investigate automatic detection o f muscle artefact, using time domain and frequency domain methods. The evaluation focussed on epoch length and performance of detection. Met hods: EEG data were recorded in 21 normal adult subjects for 50 min du ring awake state. Investigated positions included central, temporal an d parietal scalp electrodes. Expert annotation of muscle artefact was performed by accurate visual marking in a randomised test-set of the d ata, which allowed for intra-expert comparison. For time domain detect ion, the parameter set consisted of slope and maximum/ minimum amplitu de. Parameters in the frequency domain were absolute and relative 'hig h beta' power (>25 Hz) and spectral edge frequency. Distributions as c alculated from a reference period in each subject were used to investi gate the statistics of the parameter ranges. Detection thresholds were calculated from these distributions per subject, and performance was compared to constant (empirical) thresholds for the entire data set. R esults: Results indicate a 1 s epoch length as optimal for detection o f muscle artefact. The analysis using a slope threshold or absolute 'h igh beta' power showed the best results in sensitivity (80%) and speci ficity (90%), matching the expert's performance. Conclusions: Constant threshold settings performed better than statistical thresholds per s ubject. (C) 1998 Elsevier Science Ireland Ltd. All rights reserved.