JMIR Publications

ABSTRACT

Background:

Virtual reality (VR) technologies have shown promising potential in the early diagnosis of dementia by enabling regular assessment thanks to the portability and accessibility of the devices. However, previous VR studies were restricted to the analysis of behavioral responses, so information about degenerated brain dynamics could not be acquired in these studies. One of the possible approaches to enabling inspection of brain dynamics using a portable platform is the application of wearable EEG devices, which provide simple and convenient neuroimaging tools for identifying characteristic neural activities of dementia.

Objective:

The aim of this study was to develop a cognitive impairment (CI) screening tool based on a wearable EEG device integrated into a VR platform, and to evaluate whether it is feasible and effective to screen for CI using this device.

Methods:

Subjects were asked to use a hardware setup consisting of a frontal six-channel EEG device mounted on a VR device and to perform tasks designed for assessing cognitive abilities in four domains (focused attention, selective attention, depth perception and working memory). Behavioral response profiles and EEG features were extracted during the tasks, and investigated to determine whether patients with CI could be differentiated from healthy controls (HCs). Furthermore, random forest classifiers were trained on the extracted features to evaluate whether EEG characteristics during the cognitive tasks were helpful for screening subjects with CI.

Results:

A total of 59 participants (30 HCs and 29 CI) participated in this study. The CI group achieved notably worse cognitive task scores. More importantly, power spectral analysis of the EEG signals revealed notable differences between the two groups. Characteristic observations from participants with CI included reduced alpha activity during the resting-state (P = .03) and other cognitive tasks (P = .04 and .006 for selective attention and depth perception tasks, respectively), which are well-known pathological EEG patterns for dementia. Therefore, the delta-to-alpha ratio (DAR) and theta-to-alpha ratio (TAR), which are well-known biomarkers for dementia screening, were significantly greater in the CI group. The performance of the patient classification consistently improved when EEG characteristics measured during cognitive tasks were additionally included in feature attributes, suggesting that our protocol provides discriminative information that could improve the performance of dementia screening.

Conclusions:

This study demonstrated that the integration of EEG devices into a VR framework could emerge as a powerful and synergistic strategy for constructing an easily accessible EEG-based dementia screening tool.