Next-generation BCI spellers: a comparative study of ERP responses and fatigue in mixed, virtual, and desktop realities

Arif S.M.U. Shomanov A. Tyler B. Yazici A.
March 2026 Springer Science and Business Media Deutschland GmbH

Virtual Reality
2026 #30 Issue 1

This study provides critical insights into next-generation BCI systems, specifically addressing the challenges of visual and physical fatigue, as well as performance limitations, in VR and MR environments. Conventional Event-Related Potential (ERP)-based spellers typically rely on two-dimensional visual stimuli presented on desktop LED screens. With the emergence of immersive head-mounted displays (HMDs), there is growing interest in integrating (VR) and (MR) into brain-computer interface (BCI) systems. In this study, we implemented ERP speller systems across Desktop, VR, and MR environments, evaluating (1) P300/N200 mean amplitudes, (2) spelling performance, and (3) physiological and physical fatigue. Using Linear Discriminant Analysis (LDA) alongside Random Forest (RF), k-Nearest Neighbors (kNN), and XGBoost classifiers, we found that the Desktop environment shows higher P300 and N200 amplitudes, achieving a spelling accuracy of 75% and an Information Transfer Rate (ITR) of 26.1 bits/min, compared to MR (63.4% accuracy, 20.5 bits/min) and VR (60.8% accuracy, 19.8 bits/min). Cross-session experiments, analyzed via two-sided paired t-tests, revealed stable P300 signals in Desktop settings but significant performance declines in VR and MR during prolonged use. Symptom questionnaires indicated that MR induced the highest visual fatigue, while VR caused the most physical fatigue. These findings highlight critical challenges in VR/MR BCI systems, including visual and physical stress and performance degradation, while demonstrating the potential of MR-P300 spellers for applications in neurorehabilitation and assistive communication for motor-impaired individuals. Our results underscore the need for optimized 3D rendering, adaptive user-system mechanisms, and advanced machine learning to enhance next-generation BCI performance in immersive environments.

Brain computer interface (BCI) , Event-related potential (ERP) , MR (MR) , VR (VR)

Text of the article Перейти на текст статьи

Department of Computer Science, Nazarbayev University, Qabanbay Batyr Ave 53, Astana, 010000, Kazakhstan

Department of Computer Science

10 лет помогаем публиковать статьи Международный издатель

Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026