Correlative Spectral Mapping of Surface Electromyography for the Detection of Temporomandibular Joint Disorder

In this study, we introduce a novel approach to develop the detection of temporomandibular joint disorder using surface electromyography of Masticatory muscle with wavelet analysis which includes Denoising and feature extraction using spectral coding. The normalized electromyographic characteristics of masticatory muscles in patients with temporomandibular joint disorders (TMD) and healthy controls were studied. The existing approach of entropy-based coding has a spatial domain limitation, where magnitudes are used as feature vectors. A new approach of spectral domain coding for spectral mapping and feature selection process is developed. The results show that, the analysis of the developed approach which is combined model with Discrete wavelet transform (DWT) and Multiclass SVM demonstrated an improvement of parametric valuation of accuracy, precision and signal to noise ratio to the processing system. The classification performance of the proposed approach has 94.7% of accuracy. Statistical analysis shows that there is a significant difference in mean rank of control and experimental group among the subjects with p < 0.05.