Deconvolution estimation of nerve conduction velocity distribution.

A conduction velocity distribution (CVD) estimator that incorporates volume conductor modeling of the nerve-evoked response is introduced in this paper. The CVD estimates are obtained from two compound nerve action potentials (CNAP) recorded at the skin surface. A third channel is introduced in order to assess the estimator performance in the experimental case. The relevance of using an accurate signal model is shown by comparing the performance of the proposed estimator with a previous approach based on a different CNAP model. The performance of the proposed estimator is evaluated for simulated and experimental data. The study assesses signal-to-noise ratio immunity and sensitivity to errors in the model parameters.