Signal processing methods for automatic F-wave latency assignment

F-waves are motor responses resulting from antidromic activation of motor neurones. Any successful F-wave analysis algorithm must satisfactorily address the following challenges: the uncertainty of F-wave occurrences, the variability of F-wave morphology and poor signal quality. F-wave latency is the most frequently used parameter for clinical diagnosis, and it is defined as the time difference between the stimulus and the initial deflection of the F-wave from its baseline. Our goal is to develop an automatic latency assignment algorithm that will yield F-wave latencies that are highly correlated with those assigned by human operators. In this paper, we present signal processing techniques for F-wave noise suppression and baseline removal, and report the results of latency assignment improvement.