Detection of multitask mental workload using gamma band power features

Korkmaz, Onur Erdem; Korkmaz, Sevde Gül; Aydemir, Onder · 2024 · Crossref

DOI: 10.1007/s00521-024-09627-9

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the detection of cognitive fatigue and mental workload by distinguishing between single-task and multitasking scenarios using electroencephalography (EEG) signals. Motivated by the need for reliable methods to monitor cognitive states in high-stress environments, the research investigates how different task types affect neural activity, specifically focusing on the gamma frequency band. The authors aim to provide a robust classification method that can identify multitasking-induced fatigue, which is critical for applications requiring sustained attention, such as brain-computer interfaces and decision-making tasks. The experimental design involved eight participants (five males, three females) who performed a cognitive workload paradigm consisting of three blocks. Blocks 1 and 2 involved single-task mental arithmetic, while Block 3 required multitasking: solving arithmetic problems while simultaneously listening to news recordings. EEG data were recorded using a 32-channel actiCHamp system at a sampling rate of 250 Hz. The data underwent preprocessing with fourth-order Butterworth band-pass filters to isolate delta, theta, alpha, beta, and gamma frequency bands. The signals were segmented into 1-second, 3-second, and 5-second epochs. Feature extraction utilized the spectrogram method to calculate band power. Classification was performed using Artificial Neural Networks (ANN), with Support Vector Machines (SVM) and Linear Discriminant Analysis (LDA) used for comparison. The dataset was split into 50% training, 25% validation, and 25% testing, with the process repeated 50 times to ensure statistical reliability. The results demonstrated that the gamma frequency band was the most effective feature for distinguishing between single-task and multitasking conditions. For the comparison between Block 1 (single-task) and Block 3 (multitask), the highest classification accuracy was 97.11% using 5-second EEG segments and the gamma band. For the comparison between Block 2 (single-task) and Block 3 (multitask), the highest accuracy was 90.88% under the same conditions. The study found that classification performance decreased as the experiment progressed, likely due to increasing cognitive fatigue in Block 2 compared to Block 1. The gamma band consistently outperformed other frequency bands, while the delta band yielded the lowest accuracy. The ANN model proved superior to SVM and LDA in this specific configuration. The significance of this work lies in its high-accuracy method for detecting multitask mental workload, offering a new benchmark in cognitive fatigue research. By identifying specific neural markers associated with multitasking, the proposed model can be applied to detect attention deficits and focus impairments. This approach provides a non-invasive tool for monitoring cognitive states in real-time, potentially enhancing safety and performance in fields such as aviation, healthcare, and human-computer interaction. The study highlights the utility of short-duration gamma band features for efficient and accurate mental state classification.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• neuro workload indices
• workload measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data