Analysis of Continuous Motor Nerve Conduction Velocity Distribution from Compound Muscle Action Potential

Abstract

A new approach to non-invasively uncover the pattern of myelinated Aα motor nerve fiber conduction velocity distribution (CVD) from the corresponding compound muscle action potential (CMAP), also known as the inverse problem of motor nerve conduction study, has been explored in this work. The previous works to solve this type of problem were mostly in the discrete manner. We leveraged a continuous approach to exploit the gradient optimization technique to solve this problem. A diphasic sinusoidal function was taken to model the motor unit action potential (MUAP) signal and a 5th order polynomial function was taken to model the assumed continuous CVD curve. The continuous CVD instead of a discrete CVD helps us to perform more computation. The inverse results derived using the proposed methodology closely matched (almost 100%) the predicted CVD with the original CVD of different shapes from the corresponding simulated CMAP data using the forward solution of nerve conduction.