EMG coherence and spectral analysis in cervical dystonia: discriminative tools to identify dystonic muscles?

Botulinum toxin injections in the dystonic muscles are the preferred treatment for cervical dystonia (CD), but proper selection of the dystonic muscles remains a challenge. We investigated the use of EMG coherence and autospectral analysis as discriminative tools to identify dystonic muscles in CD patients. We compared the occurrence of 8-14 Hz autospectral peaks and 4-7 Hz intermuscular coherences between 10 CD patients and 10 healthy controls. Secondly, we compared the muscles with significant 4-7 Hz coherences with the muscles that were selected clinically for botulinum toxin treatment. Autospectral peaks between 8 and 14 Hz were significantly more often absent in the splenius capitis (SPL) muscles of CD patients compared to controls (p <0.01). Contrary to previous findings, there was no significant difference in the occurrence of 4-7 Hz intermuscular coherences between patients and controls and the diagnostic accuracy of coherence analysis to identify the clinically dystonic muscles was low. Intermuscular EMG coherence analysis cannot reliably discriminate patients from controls. Autospectral changes in the SPL muscles are a more discriminative feature of CD. In patients, coherence analysis does not seem to be a reliable method to identify dystonic muscles. The clinical relevance and the origin of the autospectral changes need further study