INTRODUCTION: This article presents the investigation of the clinical utility of artificial neural networks as an alternative in interpreting vestibular evoked myogenic potential (VEMP). METHODS: We developed an application that incorporates a neural network. This software reproduces and analyses the vestibular potential under study. The software also gives an estimated error for this process. RESULTS: The correlation obtained in this study between the neural network and an experienced human observer was always above 95%. CONCLUSION: The neural network created offers useful, reliable interpretations of vestibular evoked myogenic potential, in agreement with current clinical parameters.
INTRODUCTION: This article presents the investigation of the clinical utility of artificial neural networks as an alternative in interpreting vestibular evoked myogenic potential (VEMP). METHODS: We developed an application that incorporates a neural network. This software reproduces and analyses the vestibular potential under study. The software also gives an estimated error for this process. RESULTS: The correlation obtained in this study between the neural network and an experienced human observer was always above 95%. CONCLUSION: The neural network created offers useful, reliable interpretations of vestibular evoked myogenic potential, in agreement with current clinical parameters.