Defining morphology of periodic leg movements in sleep: an evidence-based definition of a minimum window of sustained activity

Objective: Current standard guidelines for scoring periodic leg movements (PLM) define the start and end of a movement but fail to explicitly specify the movement morphology necessary to classify an EMG event as a PLM, rather than some other muscle event. This is currently left to the expert visual scorer to determine. This study aimed to define this morphology to provide a consistent standard for visual scoring and to improve automatic periodic leg movements in sleep scoring. Methods: A review of expert PLM scoring produced a hypothesized morphology criterion: a window of high EMG activity within the movement lasting at least 0.5 s. Two diverse expert visual scorers were independently presented with images of EMG tracings from candidate leg movements (CLM) that either passed or failed this requirement (aka “full” or “empty” movements, respectively), and indicated whether each should be scored as CLM. The 0.5-s window was compared with alternatives of 0.25 and 0.75 windows. Results: Expert scorers on average identified 94 % of “full” movements as CLM in contrast to only 8.5 % of “empty” movements. The proposed minimum window of 0.5 s also resulted in the highest agreement between visual scorers and between scorers and an automatic program. Conclusion: An added criterion requiring 0.5 s of high EMG activity within a valid CLM improves the accuracy of automatic scoring algorithms in relation to the gold standard of expert visual scorers. Our results suggest that this rule is an accurate representation of the morphology feature used by experts. This new rule has the potential to improve consistency and accuracy of visual and automatic scoring of PLM.