Neural Networks Meet Neural Activity: Utilizing EEG for Mental Workload Estimation
DOI: 10.1007/978-3-031-78195-7_22
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 challenge of accurately estimating mental workload using electroencephalography (EEG), a critical capability for high-stakes domains like aviation, medicine, and brain-computer interfaces. While EEG offers real-time, non-invasive monitoring, its utility is often limited by high dimensionality, intrinsic noise, and non-stationarity. To overcome these limitations, the authors introduce a customized ConvNeXt architecture, a convolutional neural network originally designed for computer vision, adapted specifically for EEG signal analysis. The research aims to demonstrate that this advanced architecture can extract subtle neural patterns more effectively than traditional models, thereby improving classification precision and establishing new standards for cognitive state measurement. The experimental design utilized the Simultaneous Task EEG Workload (STEW) dataset, which contains raw EEG data from 48 subjects performing multitasking exercises. Data was recorded using a 14-channel Emotiv EPOC headset at 128 Hz. Preprocessing involved bandpass filtering, independent component analysis to remove artifacts, and epoching into one-second intervals with half-second overlaps, resulting in 26,910 samples. The dataset was split into training, validation, and test sets in a 70:15:15 ratio. The proposed ConvNeXt model was modified for EEG by reducing channel dimensions and adjusting block structures to handle the specific signal characteristics. It was evaluated against four baseline classifiers: Support Vector Machine (SVM), EEGNet, TSception, and a Transformer model. Performance was assessed using stratified five-fold cross-validation on two tasks: a binary classification (No Task vs. SIMKAP) and a ternary classification (low, medium, and high workload levels). The results indicate that the customized ConvNeXt model significantly outperformed all baseline methods. It achieved an accuracy of 95.76% for the binary classification task and 95.11% for the multi-class ternary task. These findings demonstrate the model’s resilience to noise and its efficiency in handling the high dimensionality of EEG data. The study highlights that ConvNeXt’s architectural features, such as layer normalization and inverted residual structures, allow for more stable training and better feature extraction compared to conventional CNNs or SVMs. The superior performance was consistent across varying complexities of mental workload representation, confirming the robustness of the approach. The significance of this work lies in its demonstration that computer vision architectures can be effectively repurposed for neurophysiological signal analysis, bridging a gap between deep learning domains. By achieving higher accuracy and reliability, the proposed method advances the feasibility of real-time mental workload monitoring in practical applications. This contributes to the broader fields of cognitive neuroscience and human-computer interaction by providing a more precise tool for assessing cognitive load, which can enhance safety, efficiency, and user experience in critical operational environments. The study establishes ConvNeXt as a powerful baseline for future EEG-based classification research.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-08-09 |
| archive | success | unpaywall | — | — | 2 | 2026-08-09 |
| extract | success | cached | — | — | 3 | 2026-08-10 |
| clean | success | clean | — | — | 1 | 2026-08-09 |
| chunk | success | chunk | — | — | 1 | 2026-08-09 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-08-09 |
| enrich | success | semantic_scholar | — | — | 1 | 2026-08-09 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 10 | 2026-08-11 |
| verify | success | — | — | — | 2 | 2026-08-10 |
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.
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