Brigham R. Boice  Samantha E. Shomo  Jonathan Banfill

Vol. 1,  No. 1, pp. 1-16, Aug.  2024
10.23246/AAIRJ.2024.01.01.05

motor imagery BCI immersive digital environments motor classification

PDF

This paper will systematically review techniques for classifying motor imagery in brain- computer interfaces (BCIs) and their application in immersive digital environments. The primary focus will be to examine the technical methodologies currently used to interpret motor imagery, exploring how these advancements might improve user interactions within flat screen, virtual (VR), augmented (AR), or mixed-reality (MR) experiences. While BCIs have shown great promise, their methods for synthesizing user commands are not without their limitations, especially for a commercial entertainment product. The accuracy of motor imagery classification can be affected by various factors, such as user fatigue and electrode placement. The classification of motor imagery from BCIs is essential for predicting user reactions and facilitating these new interaction methods in immersive environments. These advancements will offer alternative control mechanisms, significantly benefiting users with impairments, enhancing system responsiveness to user states, and addressing the constraints of existing physical control instruments. The main objective of this review is to consolidate current research on BCI technology and motor imagery classification, investigate existing applications, identify current limitations, and suggest future directions for this technology. This review addresses critical questions about the methodologies employed in BCIs for motor imagery interpretation and their practical applications in immersive digital settings. Key findings highlight the predominance of certain machine learning techniques, such as CNNs, LDAs, and SVMs, which were favored for their efficiency in processing EEG data. Additionally, the review identifies significant challenges, including the underrepresentation of AR and MR applications in MI-BCI research and a need for more diversity among study participants. Addressing these gaps is crucial for improving the generalizability and inclusivity of BCI systems. By providing a comprehensive knowledge base, this paper aims to support further research and development in BCI technologies and their integration into immersive digital environments, thereby paving the way for more efficient and inclusive human-machine interactions