# The use of fully immersive virtual reality for screening neurodegenerative diseases: A systematic review of behavioral and diagnostic outcomes

## TL;DR

This paper reviews how immersive virtual reality can help detect early signs of Alzheimer's, Parkinson's, and cognitive decline by simulating real-life tasks and using machine learning to improve accuracy.

## Contribution

The study systematically evaluates the diagnostic potential of immersive VR for neurodegenerative diseases, highlighting its feasibility and the role of machine learning in enhancing accuracy.

## Key findings

- Immersive VR assessments showed strong diagnostic accuracy (up to 89% area under the curve) for detecting mild cognitive impairment and early-stage Alzheimer's.
- Combining VR data with machine learning improved diagnostic accuracy to 87-100% in some studies.
- VR-based screening is ecologically valid and engaging but requires standardization and longitudinal validation for clinical adoption.

## Abstract

Early detection of Alzheimer's disease (AD), Parkinson's disease (PD), and mild cognitive impairment (MCI) is crucial for timely intervention. Traditional cognitive screening tools lack ecological validity and sensitivity. Virtual reality (VR) provides realistic, controlled environments for assessing multidimensional cognition. This systematic review evaluated the diagnostic accuracy, feasibility, and applicability of immersive VR assessments for neurodegenerative screening. We searched PubMed, PsycINFO, and Embase for studies published June 2005 to April 2024. Eligible studies used head‐mounted displays in adults with MCI, early AD/PD, or dementia. Ten studies (n = 472) met criteria. Tasks targeted spatial memory, executive function, attention, and navigation. Several reported strong discriminations (area under the curve up to 0.89) and, when combined with machine learning, accuracies of 87% to 100%. Immersive VR shows promise as an ecologically valid, engaging, and scalable screening approach; however, standardization of tasks and outcomes, real‐world validation, and robust longitudinal evidence are needed to support clinical adoption.

This review systematically describes the application of fully immersive virtual reality (VR) in the early screening of neurodegenerative diseases, with a focus on studies using head‐mounted devices to simulate real‐life tasks.Task types such as spatial memory, daily living simulations, and executive function assessments have demonstrated high sensitivity and specificity in diagnosing mild cognitive impairment (MCI) and early‐stage Alzheimer's disease (AD).Approximately one third of studies combined machine learning techniques to analyze multimodal behavioral data (e.g., path deviations, task duration, and language responses), significantly improving diagnostic accuracy.This study highlights methodological heterogeneity, small sample sizes, and the lack of longitudinal studies as current research limitations, and calls for future standardized, multicenter, and long‐term follow‐up studies to validate the predictive validity and real‐world applicability of VR tools.

This review systematically describes the application of fully immersive virtual reality (VR) in the early screening of neurodegenerative diseases, with a focus on studies using head‐mounted devices to simulate real‐life tasks.

Task types such as spatial memory, daily living simulations, and executive function assessments have demonstrated high sensitivity and specificity in diagnosing mild cognitive impairment (MCI) and early‐stage Alzheimer's disease (AD).

Approximately one third of studies combined machine learning techniques to analyze multimodal behavioral data (e.g., path deviations, task duration, and language responses), significantly improving diagnostic accuracy.

This study highlights methodological heterogeneity, small sample sizes, and the lack of longitudinal studies as current research limitations, and calls for future standardized, multicenter, and long‐term follow‐up studies to validate the predictive validity and real‐world applicability of VR tools.