SVD and higher-order statistical processing of human nerve signals

Human afferent whole nerve signals recorded using an implanted nerve-cuff electrode were analyzed using two algorithms based on the statistical properties of the signals. The processing method typically described in the literature (Rectification and Bin-Integration-RBI) has serious shortcomings in processing these signals, which have very poor signal-to-noise ratios. Algorithms based on a Singular Value Decomposition (SVD) of the signal´s 2nd and Higher-Order Statistics (HOS) have resulted in more robust signal detection. Reliable detection of afferent nerve signals is essential if such signals are to be of use in artificial sensory-based functional electrical stimulation neural prosthetics