Immersive Technologies for Cognitive Rehabilitation in Dementia and Mild Cognitive Impairment: Systematic Review

Background: Cognitive decline across the mild cognitive impairment (MCI)-dementia continuum is a major driver of loss of independence and growing health- and social-care burden. Immersive technologies, such as virtual reality (VR), augmented reality (AR), and Cave Automatic Virtual Environment (CAVE) systems, are increasingly explored as tools to enhance engagement, personalization, and ecological validity in cognitive rehabilitation. Objective: This systematic review synthesizes current evidence on the usability, therapeutic effects, and implementation challenges of immersive technologies for cognitive rehabilitation in MCI and dementia. Methods: A systematic search of Scopus and Web of Science was conducted for peer-reviewed journal articles published between 2021 and 2026. Eligible studies investigated VR, AR, or CAVE interventions targeting cognitive rehabilitation outcomes in MCI and/or dementia and reported measures related to usability or acceptability, or cognitive, functional, or behavioral outcomes. Due to heterogeneity in technologies, intervention content, and outcome measures, findings were synthesized narratively with comparisons across modalities and study designs. Results: In total, 119 studies met the inclusion criteria. Across immersive VR interventions, signals of benefit were most consistently reported for memory, attention, and executive functioning, with several studies also targeting outcomes with higher ecological relevance (eg, everyday task performance and functional skills). AR approaches primarily support context-aware cueing and task guidance in real-world settings, aiming to strengthen daily functioning and independence. CAVE-based systems were frequently used for spatial navigation and embodied interaction, offering advantages for supervised clinical deployment. Key barriers included cybersickness and comfort issues, interface complexity, and onboarding demands in cognitively impaired users, limited accessibility and standardization of outcome measures, small samples and short follow-up periods, and practical constraints related to cost, space, staffing, and caregiver involvement. Conclusions: Immersive VR, AR, and CAVE systems are feasible and often engaging for cognitive rehabilitation in MCI and dementia, with promising therapeutic signals but substantial uncertainty driven by methodological and implementation heterogeneity. Future work should prioritize standardized reporting (intervention components, dose, and adverse events), clinically meaningful outcomes (including functional end points), adequately powered comparative trials, and explicit evaluation of scalability and real-world deployment pathways.