Optimization of an AI Prosthetic Leg Using Machine Learning Algorithms Trained with EMG Signals

Human amputations have been happening throughout the entire history of humankind, mostly due to medical reasons or accidents. Prosthetic limbs have usually been the most proper solution to amputations, and hence, have been of significant interest in medical science, biomedical engineering, and mechatronic engineering. In the recent decades, with the appearance of artificial intelligence in parallel with the development of technology, intelligent prosthetic limbs have become a vibrant research interest. Nowadays, different researches are being conducted aiming to the recognition of performance patterns of different natural limbs, or simulation of the performance patterns. This research is dedicated to the optimization of an intelligent prosthetic knee joint by the utilization of proper existing artificial intelligence algorithms. Utilized well-known machine learning algorithms are described and the implementation outcomes of each algorithm are presented and discussed in details, in this article. Even though there is still room for improvement, some of the results obtained in this research have shown more than ۹۹ percent accuracies, which could be considered as a quite desirable and reliable outcome for such research

کلیدواژه ها:

Smart Prosthetic Limbs ، Electromiography ، Biopotentials ، Machine Learning