Evaluating neural tissue regeneration by ANN based classification of MGN signals

One of the major challenges of microsurgery is nerve regeneration across lesions. In this respect it is very important to assess whether a peripheral nerve is reconstructed and how much of its initial function has been restored. Several methods are available for this purpose: function tests, histological techniques, neurophysiologic tests. The present paper addresses magnetoneurography (MNG), a neurophysiology test which is very suitable because of its high sensitivity features. This technique has been applied to animal models with a strictly controlled procedure to obtain a set of highly comparable and replicable results. The aim of this study is to assess whether the time interval between surgery and MNG measurement is a parameter according to which the MNG signals can be classified obtaining insights into the state of the regeneration process. The classification has been performed using two types of supervised artificial neural networks (ANN)