Design and Validation of a Multimodal Immersive Virtual Environment for Cognitive Intervention in Patients with Alzheimer's Disease

Virtual reality (VR) environments have been widely applied in cognitive assessment and intervention research; however, most existing studies on cognitive load rely on fixed-difficulty paradigms that are insufficient for addressing Alzheimer’s disease (AD)-specific deficits, limit ecological validity, and risk cognitive overload and interpretive bias. To overcome these limitations, we proposed and validated a multimodal VR method with adaptive adjustment capabilities, grounded in the affective–cognitive impairments of AD patients. In this study, we further introduce a closed-loop adaptive VR method that dynamically modulates task parameters across trials in accordance with users’ real-time cognitive states. Eye-tracking and electroencephalography (EEG) modules are incorporated to enable synchronous acquisition and integration of multimodal physiological signals, supporting real-time cognitive load modeling and adaptive task modulation. The task battery combines emotion recognition and executive function evaluation with multi-level visual and auditory interference to effectively elicit AD-related cognitive deficits. Experimental results demonstrate that the proposed method effectively maintains participants within an optimal cognitive load range, mitigating risks of overload and underload while significantly improving EEG signal reliability and behavioral validity. Our findings provide strong empirical evidence for the effectiveness of adaptive VR approaches in individualized cognitive assessment and intervention for AD, and lay a solid foundation for the broader application of immersive VR technology in neuropsychological research and practice.