Decomposition of surface EMG signals into single fiber action potentials by means of neural networks

A neural network approach to decomposing surface EMG signals into a multitude of single muscle-fiber action potentials (SFAP) is described. This decomposition yields the signal forms of the SFAPs and allows the particular SFAPs to be localized relative to the recording (surface) electrodes. The goal is to allow the physician and the medical researcher to observe the waveforms and localize SFAPs below the surface of the skin in a noninvasive manner and without discomforting the patient. The approach utilizes a Hopfield neural network.<
>