Estimation of the nerve conduction velocity distribution by peeling sampled compound action potentials

Abstract

The nerve conduction velocity distribution is estimated by a "peeling" concept method. The compound action potential is a linear summation of the single fiber action potentials propagating along the nerve fibers and can be expressed as the convolution of a delay sequence and the single fiber action potential wavelet. An algorithm based on the comparison of the front part of a continuously deconstructed compound action potential signal to a single fiber action potential wavelet is developed (i) to separate the delay sequence from the sampled compound action potential signal, and (ii) to estimate the distribution of the conduction velocities.