An automatic method for scoring leg movements in polygraphic sleep recordings and its validity in comparison to visual scoring.

STUDY OBJECTIVES The study presents an automatic method for scoring leg movements in polysomnographic recordings and describes an empirical investigation of its validity. DESIGN Leg movements measured by means of the surface electromyogram activation of the right and left tibialis anterior muscle contained in 24 digitally recorded all-night polysomnograms were analyzed visually according to the American Sleep Disorders Association guidelines by 2 experienced raters and automatically scored using a newly developed electromyogram-based analytical method. Two visual scorings and the automatic scoring were compared in pairs using descriptive and confirmative statistical methods. SETTING N/A. PARTICIPANTS All-night polysomnograms of adaptation nights at the sleep laboratory of 10 patients with idiopathic restless legs syndrome (RLS) according to the International RLS Study Group. INTERVENTIONS N/A. MEASUREMENTS AND RESULTS Agreement rates between the 2 well-trained scorers and the automatic method were comparable. Based on the first scorer's results (100%) 92.5% of the movements were detected by the second scorer and 94.3% by the automatic method. When the visual scorings were compared, the rate of false-positive and false-negative errors were 7% and 3%, respectively. Comparing both visual scorings with the results of the automatic scoring yielded false-positive and false-negative rates in the range from 3% to 8%. CONCLUSIONS The degree of accordance between the 2 visual scorings and between the visual and the automatic scorings were comparable. Therefore, this method is valid and may be used for the automatic detection of leg movements in future studies.