A comprehensive study of EEG-based control of artificial arms

Abstract

Introduction/purpose: The electroencephalography (EEG) signal has a great impact on the development of prosthetic arm control technology. EEG signals are used as the main tool in functional investigations of human motion. The study of controlling prosthetic arms using brain signals is still in its early stages. Brain wave-controlled prosthetic arms have attracted researchers' attention in the last few years. Methods: Several studies have been carried out to systematically review published articles as a means of offering researchers and experts a comprehensive summary of the present, state-of-the-art EEG-based control techniques used in the prosthetic arm and other technologies. Results: 175 articles were studied, compared, and filtered to only include the articles that have strong connections to the study. Conclusion: This study has three goals. The first one is to gather, summarize, and evaluate information from the studies published between 2011 and 2022. The second goal is to extensively report on the holistic, experimental outcomes of this domain in relation to current research. It is systematically performed to provide a wealthy image and grounded evidence of the current state of research covering EEG-based control of prosthetic arms to all experts and scientists. The third goal is to recognize the gap in knowledge that demands further investigation and to recommend directions for future research in this area.

The LIBRA NeuroLimb: Hybrid Real-Time Control and Mechatronic Design for Affordable Prosthetics in Developing Regions

“…In this sense, exclusive EEG control improves the accuracy of the discernment of the gesture [ 14 ]. Due to the mapping of brain commands that it applies, it is usually necessary to implement deep learning and machine learning models to decode the EEG data of the headband with an arrangement of sensors operating as a part of the brain–computer interface (BCI) [ 15 ]. However, although this has advantages compared to EMG reading alone, the training process for control models usually varies from user to user, in time and comfort [ 16 ].…”