Teager-Kaiser energy operator improves the detection and quantification of nociceptive withdrawal reflexes from surface electromyography

This paper presents a simple and robust method to improve detection and quantification of nociceptive withdrawal reflexes in surface electromyography, based on pre-processing the signals with the Teager-Kaiser energy operator. Withdrawal responses were recorded from tibialis anterior and soleus muscles in three hundred subjects, and a subset of fifty recordings of each muscle were classified by an expert as either exhibiting a reflex or not. Performance of five detection algorithms was compared against the expert's classification using a receiver operating characteristic analysis. Results showed improvement in the performance of all algorithms when the pre-processing algorithm was applied.