Models and analysis of compound nerve action potentials

Upon electrical stimulation of a peripheral nerve, a compound action potential (CAP) can be recorded, a procedure that is widely used to study the functional condition of a nerve. The CAP provides relevant information about such parameters as the number of active myelinated fibers and their propagation velocities. This paper reviews methods of model-based CAP analysis that have emerged during the last decade. First, the basic model formulation for the CAP as resulting from a linear summation of the underlying single fiber action potentials (SFAPs) is discussed. Subsequently, a survey is given of volume conductor model approaches for the prediction of the SFAPs. Attention is then concentrated on the "inverse" problem of extracting relevant nerve parameters from experimentally recorded CAPs. Finally, the applicability of such methods is discussed and a direction for future developments is outlined.