A numerical model of the human nervous fiber and surrounding tissues is pre
sented, which is able to Pi edict the outcome of evoked-potential (EP) test
s. Ii is designed to be helpful in far-field based diagnosis, as it has bee
n validated through experimental results from evoked potential tests. Regul
ar and pathological situations have been simulated, and models for disconti
nuities in fiber and tissue have been developed, the goal being to locate a
nd classify anomalies - either pathological or physiological - based on the
Potential trace arising in EP tests. Some already known qualitative models
have been improved in order to become quantitatively reliable. and realist
ic parameters have been taken into account.