Determining hearing threshold from Brain Stem Evoked Potentials - Optimising a neural network to improve classification performance

Brain Stem Evoked Potentials (BAEPs) are considered the most objective measure currently available with which to determine the functional integrity of the peripheral auditory nervous system. Estimating hearing threshold from BAEP signals is a time consuming and laborious task, and therefore one which recommends itself to automation. In this paper we demonstrate that neural networks trained by back-propagation are an effective method to classify BAEPs. In order to achieve maxumum generalsiation, it was necesary to fine tune the learning parameters. Although this step can be time consuming, it has important clinical consequences