Experimental validation of the nerve conduction velocity selective recording technique using a multi-contact cuff electrode.

The earthworm (Lumbricus terrestris) is presented as an in vitro model of a peripheral nerve containing only two fibers each with distinctly different conduction velocities, the median and lateral giant fibers (MGF and LGF). The worm model is used with a multi-contact cuff electrode to validate the spatial-temporal filtering effect of different electrode contact configurations and the effect of applying a delay adder and matched filter tuned to either the MGF or LGF action potential (AP) to extract conduction direction and velocity from the recording. The results confirmed the known effect of inter-electrode spacing and bipolar and tripolar recording configuration on the AP amplitude. It also demonstrates a crossover point where the amplitude of the tripolar recording is larger than the monopolar recording, an effect of the slower action potential conduction velocities in the worm. The delay adder was found to be an effective velocity sensitive filter, able to discriminate units based on conduction velocity. The matched filter to be an effective means to eliminate artifact and increase signal to noise ratios, however was not found to improve selectivity.