Estimation of mental workload induced by different presentation rates in rapid serial visual presentation tasks

Area of Science:

• Neuroscience
• Human-Computer Interaction
• Biomedical Engineering

Background:

• Brain-computer interfaces (BCIs) leverage event-related brain responses for information detection.
• Rapid serial visual presentation (RSVP) paradigms are common in BCI, with parameters like presentation rate influencing task difficulty and mental workload.
• Understanding how presentation rate impacts mental workload is crucial for optimizing BCI performance and user experience.

Purpose of the Study:

• To investigate the influence of presentation rate on mental workload within an RSVP-BCI paradigm.
• To assess the separability of brain states associated with different presentation rates.
• To determine if mental workload can be accurately recognized based on electroencephalographic (EEG) data.

Main Methods:

• Recorded 64-channel EEG data from ten healthy subjects performing RSVP tasks.
• Varied the presentation rate across three distinct levels.
• Utilized one-way repeated measures ANOVA on z-scored RMSE to analyze differences in mental workload.
• Applied classification algorithms to distinguish between mental workload levels.

Main Results:

• Significant differences in mental workload were observed across different presentation rates (ANOVA, p < 0.05).
• Higher presentation rates led to a significant decrease in both behavioral and single-trial recognition performance.
• Classification accuracy for mental workload levels reached a mean of 65.5% and a maximum of 88.3%.

Conclusions:

• Presentation rate is a significant factor influencing mental workload in RSVP-BCI tasks.
• Mental workload induced by varying presentation rates can be accurately recognized using EEG data.
• This research offers a potential method for monitoring mental workload in RSVP-BCI applications.