Investigating the neural activity of various 3D visual fatigue degrees using depth-related visual evoked potentials

摘要

Severe visual fatigue symptoms occur while users view three-dimensional (3D) content. These symptoms adversely affect the user experience and hinder the widespread application of 3D media. Existing studies mainly used electroencephalography (EEG) power (based on frequency domain analysis) to evaluate 3D visual fatigue (3DVF), but their results were often contradictory and confusing. To address this problem, we investigated the relationship between EEG features and visual fatigue from the perspective of both time and time-frequency domain. Random-dot stereograms (RDS) with various disparities were employed to elicit depth-related visual evoked potentials (VEPs). The experiment lasted about 27 min for each subject to induce fatigue, and the EEG was recorded synchronously. VEPs of different fatigue levels were processed to extract both time (such as P-120 and N-170 components in VEPs) and time-frequency (like event-related spectral perturbation [ERSP]) domain characteristics. A significant increase in P-120 amplitude was observed in the parietal region, and time-frequency analysis also revealed a significant ERSP decrease of theta bands in the central-frontal region with a latency of 200 to 300 ms when subjects suffered fatigue. Our results confirm the hypothesis that activity of theta decreased with 3DVF level, or rather, we found ERSPs of theta bands on FC4 and Cz in the 200- to 300-ms window decreased significantly across four fatigue levels. Furthermore, we first reported the relationship between P-120 amplitude and 3DVF, which implied this feature could be considered as a feature of 3DVF evaluation. In summary, our studies could provide a new insight for 3DVF evaluation using EEG.