Generalizability of EEG-based Mental Attention Modeling with Multiple Cognitive Tasks

Area of Science:

• Cognitive Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Attention is a fundamental cognitive function crucial for various mental processes.
• Electroencephalogram (EEG) offers a quantifiable measure of attention levels.
• Existing research lacks empirical comparisons of different cognitive tasks for attention detection.

Purpose of the Study:

• To empirically evaluate and compare the performance of three distinct cognitive tasks in detecting and quantifying attention levels using EEG.
• To assess the generalizability of EEG-based attention recognition across subjects and tasks.

Main Methods:

• An experiment was conducted with ten subjects performing Stroop, Eriksen Flanker, and Psychomotor Vigilance Task (PVT) in a randomized order.
• Six standard band power features were extracted from EEG data.
• Classification of attention levels was performed using subject-specific and subject-independent cross-validation methods.

Main Results:

• Subject-independent classification accuracies were 61.6% (PVT), 63.7% (Stroop), and 65.9% (Flanker).
• The highest accuracy achieved was 74.1% for the Flanker test in the subject-dependent case.
• No statistically significant differences in classification accuracy were observed among the three cognitive tasks.

Conclusions:

• EEG-based attention recognition demonstrates generalizability across different subjects.
• The chosen cognitive tasks showed comparable performance for EEG-based attention detection.
• This research validates EEG as a robust tool for attention monitoring in diverse cognitive contexts.