Avaliação da atividade muscular de indivíduos com disfunção temporomandibular usando redes neurais artificiais auto-organizadas

Abstract

Background: Although many clinicians still use the electromyographic (EMG) signal to assess patients with temporomandibular disorders (TMD), so far, a method for analyzing the EMG signal that demonstrates clear differences between these patients and healthy individuals has not yet been found. Objective: The aim of this study was to evaluate the muscle activity of individuals with TMD using self-organized artificial neural networks. Methods: This was a cross-sectional study, consisting of consecutive samples, consisting of 36 women with TMD and 24 healthy women aged between 18 and 45 years. The EMG signal from the masseter and temporal muscles, both sides, was collected under conditions of rest, chewing (CHW) and maximum habitual intercuspation (MHI). The EMG signal was processed using self-organized artificial neural networks and the movement deviation profile curve (MDP) was calculated in relation to the control group (healthy). Results: The TMD group had a significantly higher MDP value (p < 0.05) with effect size ranging from moderate to high (0.26 to 0.62) for all analyzed muscles (masseter and right and left temporalis) under REST, CHW and MHI condictions. Conclusions: In this study, it was possible to observe significant differences between healthy individuals and individuals with temporomandibular disorders in the electromyographic signal analyzed from self-organized neural networks of the masseter and anterior temporal muscles on both sides, recorded in REP, ISTO and MIH.