一种基于泄漏感知评估协议的混合神经深度信念网络-门控循环单元算法用于心电图-心率变异性认知负荷分类
A hybrid neural deep belief networks-gated recurrent unit algorithm with a leakage-aware evaluation protocol for ECG-HRV-based cognitive load classification
该研究提出并评估了一种混合深度信念网络-门控循环单元(DBN-GRU)算法,用于基于心电图-心率变异性的心认知负荷分类。在WESAD和SWELL数据集上的实验结果显示,该模型在留一参与者验证下的准确率为68.4%,平衡准确率为62.8%,宏F1值为0.614,优于多数类预测器。参与者相关的特征改善了13名参与者中的13名的表现,但掩蔽参与者身份并未提升分类效果。
- Article
- Open access
- Published: 10 November 2026
Scientific Reports (2026) Cite this article
We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.
Abstract
Cognitive underload and overload can impair knowledge workers’ performance. Electrocardiogram-derived heart rate variability (HRV) offers a practical signal for assessing cognitive demand. However, sample-wise evaluation may reward participant recognition rather than cognitive-state classification. This study proposes and evaluates a hybrid deep belief network–gated recurrent unit (DBN–GRU) algorithm on WESAD and SWELL. The model combines restricted Boltzmann machine pre-training with a gated classifier head. The primary WESAD evaluations use participant-disjoint protocols and validation-based model selection. Under leave-one-subject-out validation, the model achieves 68.4% accuracy, 62.8% balanced accuracy and 0.614 macro-F1. A majority-class predictor achieves 52.8% accuracy and 0.230 macro-F1. Participant-relative features improve performance for 13 of 15 participants (p = 0.0015), whereas adversarial suppression of participant identity does not improve classification. Participant-relative features require unlabelled calibration data from each new participant. None of the eight alternative architectures differs significantly from the proposed model. Zero-shot cross-dataset transfer falls below the target majority-class baseline in both directions. Fine-tuning underperforms direct training on the target dataset. These results indicate a larger effect of participant-relative features than of architectural choice under the evaluated conditions as a quality assurance metrics.
Acknowledgements
The authors express gratitude to the teams who collected the WESAD and SWELL datasets, and to those who publicly released the derived heartrate-variability feature tables analyzed here; the corresponding links are given in the Data Availability declaration below. This article was part collaborative research based on the funding titled” Formulation and classification of mental workload in industrial workplaces using a hybrid big data analytical model for better health and social well-being” (1001/PKOMP/8580093), granted to Pantea Keikhosrokiani as the Principal Investigator (PI). In addition, the authors acknowledge and appreciate innovation center and 6G-anabaled Sustainable Society (6GESS) Pantea keikhosrokinai is part of it, University of Oulu, for the two granted innovation funds of Proof of Concept (24304435124) “AI-Mental: An AI-augmented Cognitive Balance digital health solution for employees’ mental workload management” and 6GESS New Innovative Initiative fund (2464010616) entitled “AI-Cognitive Data Driven Software Solution For Mental Well-Being and Automated Decision-making Process” to Pantea Keikhosrokiani, for proceeding with the study as part of a collaborative project. The studies and funds are derived based on her copyrighted projects.
Funding
This work was supported by grant 1001/PKOMP/8580093, awarded to Pantea Keikhosrokiani as PI. The funded project is entitled “Formulation and classification of mental workload in industrial workplaces using a hybrid big data analytical model for better health and social well-being”. Additional support came from two innovation awards from the Innovation Centre and 6GESS at the University of Oulu. Both were awarded to Pantea Keikhosrokiani for projects based on her copyrighted work. The Proof of Concept award supported “AI-Mental: An AI-augmented Cognitive Balance digital health solution for employees’ mental workload management”. The 6GESS New Innovative Initiative award supported “AI-Cognitive Data-Driven Software Solution For Mental Well-Being and Automated Decision-making Process”.
Author information
Author notes
Pantea Keikhosrokiani and Ning Wang have contributed equally to this work.
Authors and Affiliations
Faculty of Information Technology and Electrical Engineering, Faculty of Medicine, University of Oulu, Pentti Kaiteran Katu 1, 90570, Oulu, Finland
Pantea Keikhosrokiani
School of Computer Science, Universiti Sains Malaysia, 11800, Penang, Malaysia
Ning Wang & Suzi Iryanti Fadilah
Bijie Medical College, Jinhaihu New District, Bijie, 551700, Guizhou, China
Jianshu Shi
Sichuan Ertan Construction Consulting Co., Ltd., Chengdu, Sichuan, China
Wanhui Liu
Authors
- Pantea Keikhosrokiani
- Ning Wang
- Jianshu Shi
- Wanhui Liu
- Suzi Iryanti Fadilah
Corresponding author
Correspondence to Pantea Keikhosrokiani.
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Keikhosrokiani, P., Wang, N., Shi, J. et al. A hybrid neural deep belief networks-gated recurrent unit algorithm with a leakage-aware evaluation protocol for ECG-HRV-based cognitive load classification. Sci Rep (2026). https://doi.org/10.1038/s41598-026-80282-x
Received: 16 February 2026
Accepted: 06 November 2026
Published: 10 November 2026
DOI: https://doi.org/10.1038/s41598-026-80282-x
Keywords
来源:Nature 健康护理主题 · nature.com