International Research Journal of Education and Technology

Amyotrophic Lateral Sclerosis (ALS) is a progressive neurodegenerative disease that severely impacts motor functions, necessitating early and accurate diagnosis for better patient management. This study proposes a hybrid approach combining the Improved Wigner-Ville Distribution (WVD) for feature extraction with Convolutional Neural Networks (CNNs) for spatial feature learning and Long Short- Term Memory (LSTM) networks for capturing temporal dependencies in Electromyography (EMG) data. The WVD enhances the time-frequency representation of the signals, while CNNs extract complex spatial patterns, and the LSTM network ensures robust classification by modeling sequential relationships in the data. This hybrid methodology demonstrates significant improvements in diagnostic accuracy over traditional methods, offering a non-invasive, efficient solution for ALS prediction with the potential for clinical implementation.

Keywords: ALS, EMG signal analysis, Wigner-Ville Distribution, Convolutional Neural Networks, Long Short- Term Memory, neurodegenerative disease prediction, time- frequency analysis.