Objective: The aim of this study is to identify a cut-off value for the burst duration of cortical myoclonus in a large cohort of myoclonus patients. This abstract presents preliminary data on a small subset of patients.

Background: The burst duration of cortical myoclonus is used in diagnostic criteria of anatomical myoclonic subtypes; especially, to differentiate between cortical and non-cortical myoclonus. The recommended cut-off value of 100ms for cortical myoclonus has however, not yet been validated.

Method: We included two patients with definite cortical myoclonus, caused by progressive myoclonus epilepsy, and one patient with non-cortical myoclonus, caused by ataxia-telangiectasia. We defined the diagnosis of definite cortical myoclonus as the presence of a cortical spike on the EEG preceding a myoclonic jerk on the EMG, measured with either jerk-locked back-averaging, or cortico-muscular coherence. To calculate the burst duration of myoclonic jerks, we registered EMG of the wrist flexor or extensor muscle and manually selected the onset and offset of the myoclonic bursts. In addition to duration, we quantified the pattern of each burst by extracting Statistical Signal Characterization features of the power spectrum, the mean amplitude and period of the signal.[1,2] To identify a cut-off value, we selected 50 bursts per patient and used open-label K-means cluster analysis to divide all bursts in two clusters, a presumed cortical and non-cortical cluster. The mean burst duration and proportion of cortical and non-cortical jerks were calculated for each cluster.

Results: We identified 150 individual bursts of which ten were excluded due to poor signal quality. Two patients with cortical myoclonus (n=48 bursts of each) showed a mean ±SD burst duration of 35.3 ±11.5ms and 57.1 ±19.3ms, the patient with non-cortical myoclonus (n=44 bursts) 65.9 ±33ms. The K-means cluster analysis performed most accurately when based on the two features of the burst pattern. Cluster 1 consisted of 99 bursts with a mean burst duration of 43.5 ±18.1ms; cluster 2 consisted of 41 bursts with a duration of 81.4 ±21.1ms. A total of 89.6% of the cortical myoclonus bursts were part of cluster 1.

Conclusion: The preliminary data show that our approach by means of open-label cluster analysis of independent bursts is promising for the objective validation of a cut-off value for the burst duration of cortical myoclonus.

References: 1. Hossen A, Muthuraman M, Al-Hakim Z, Raethjen J, Deuschl G and Heute U. Discrimination of Parkinsonian tremor from essential tremor using statistical signal characterization of the spectrum of accelerometer signal. Biomed. Mater. Eng. 2013; 23: 513–531. 2. Hirsch HL. Statistical Signal Characterization. Artech House; 1992.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/the-burst-duration-in-cortical-myoclonus-patients-what-is-the-cut-off/