Unsupervised clustering and classification of upper limb EMG signals during functional movements: a data-driven
The study developed a method to automatically group and classify upper-limb surface electromyography (sEMG) signals during functional movements like reaching and grasping. By preprocessing signals, extracting temporal and frequency features, and using clustering techniques, the researchers identified six representative gestures. These were then classified using Artificial Neural Networks or Extra Trees with high accuracy near 95%, providing a robust tool for understanding muscle activation patterns.