Skip to content
DraconiticResearch
Search Sources API
182,549 studies31 sourcesUpdated 24 min ago
Peer-reviewedCC BYPLOS One

An interpretable multimodal deep learning model for physiological stress detection

  • Ahmed Tibermacine1
  • Imad Eddine Tibermacine2
  • M'hamed Mancer1
  • Abdelaziz Rabehi3
  • Walid Guettala4
  • Mustapha Habib5
5 affiliations
  1. 1Department of Computer Science, University of Mohamed Khider Biskra, Algeria.
  2. 2Department of Computer, Control and Management Engineering, Sapienza University of Rome, Italy.
  3. 3Laboratory of Telecommunications and Smart Systems, Faculty of Sciences and Technologies, University of Djelfa, Djelfa, Algeria.
  4. 4Department of Artificial Intelligence, ELTE Eotvos Lorand University, Budapest, Hungary.
  5. 5Division of Building Technology and Design, Department of Civil and Architectural Engineering, KTH Royal Institute of Technology, Stockholm, Sweden.
Read full textPDFDOICite

Abstract

Stress is a major contributor to physical and mental health disorders, underscoring the need for reliable and interpretable detection systems. We present an interpretable multimodal deep learning framework that integrates electrocardiogram (ECG), electrodermal activity (EDA), and respiration signals using modality-specific neural architectures complemented by feature-based classical models. To ensure robustness and generalizability, the framework was evaluated under both subject-independent and subject-dependent protocols. Results show that multimodal fusion consistently outperformed single modalities, achieving 94.8% accuracy, 93.9% F1-score, and 0.97 AUC under subject-independent validation, compared with 86–91% accuracy for individual signals. In subject-dependent experiments, performance reached 97.2% accuracy and 96.8% F1-score. Comparative analysis with recent state-of-the-art methods demonstrated the competitiveness of our approach, while ablation studies confirmed the contribution of each modality and fusion strategy. Statistical significance testing with (p < 0.05) further validated the reliability of these gains. Importantly, interpretability analyses using Grad-CAM and attention visualization provided physiological insights, revealing that the model emphasizes clinically meaningful regions such as ECG R-peaks and phasic EDA fluctuations. By uniting high performance, robustness, and interpretability, this work advances stress detection research and offers a physiologically grounded framework for continuous mental health monitoring.

Related

  1. YieldLearn: An interpretable machine-learning pipeline for estimating yield-related traits based on physiological indicators
    X. Zhang, H. Shaterian, P. Ashe
    bioRxivPreprintCC BY-NCFull textBioinformatics & computational biologybioRxiv
  2. Learning from Adaptive Oscillatory Modes: An Interpretable Synchrosqueezing Transformer for Patient-Independent Seizure Detection
    Afrah Alanazi, Kais Khaldi, Sarah Almaghrabi +1
    MathematicsPeer-reviewedCC BY 4.0NeuroscienceCrossref
  3. A Multimodal Machine Learning Framework Integrating Genomic, Epidemiological, Environmental, and Clinical Imaging Data for Accelerated Measles Detection and Transmission Inference
    Jiada Li
    arXivPreprintCC BY 4.0Full textBioinformatics & computational biologyarXiv
  4. Replacing EEG with peripheral physiological signals for driving behavior recognition: a physiology-informed multimodal approach
    Huaqiang Xu, Yuwei Zhang, Xiaoqing Sun +2
    Cognitive NeurodynamicsPeer-reviewedLicence not statedFull textNeurosciencePMC
  5. Circadian multimodal physiological, ocular and behavioural data from a 40-hour forced-desynchrony protocol
    H. S. Heinrichs, C. Zanoli, S. M. Thalji +3
    bioRxivPreprintCC BYFull textNeurosciencebioRxiv

Cite

BibTeXRIS
BibTeX
@article{tibermacine2026interpretable,
  title = {{An interpretable multimodal deep learning model for physiological stress detection}},
  author = {Ahmed Tibermacine and Imad Eddine Tibermacine and M'hamed Mancer and Abdelaziz Rabehi and Walid Guettala and Mustapha Habib},
  journal = {PLOS One},
  year = {2026},
  month = oct,
  doi = {10.1371/journal.pone.0360314},
  url = {https://doi.org/10.1371/journal.pone.0360314},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - An interpretable multimodal deep learning model for physiological stress detection
AU  - Ahmed Tibermacine
AU  - Imad Eddine Tibermacine
AU  - M'hamed Mancer
AU  - Abdelaziz Rabehi
AU  - Walid Guettala
AU  - Mustapha Habib
PY  - 2026
DA  - 2026/10/29
JO  - PLOS One
DO  - 10.1371/journal.pone.0360314
UR  - https://doi.org/10.1371/journal.pone.0360314
AB  - Stress is a major contributor to physical and mental health disorders, underscoring the need for reliable and interpretable detection systems. We present an interpretable multimodal deep learning framework that integrates electrocardiogram (ECG), electrodermal activity (EDA), and respiration signals using modality-specific neural architectures complemented by feature-based classical models. To ensure robustness and generalizability, the framework was evaluated under both subject-independent and subject-dependent protocols. Results show that multimodal fusion consistently outperformed single modalities, achieving 94.8% accuracy, 93.9% F1-score, and 0.97 AUC under subject-independent validation, compared with 86–91% accuracy for individual signals. In subject-dependent experiments, performance reached 97.2% accuracy and 96.8% F1-score. Comparative analysis with recent state-of-the-art methods demonstrated the competitiveness of our approach, while ablation studies confirmed the contribution of each modality and fusion strategy. Statistical significance testing with (p < 0.05) further validated the reliability of these gains. Importantly, interpretability analyses using Grad-CAM and attention visualization provided physiological insights, revealing that the model emphasizes clinically meaningful regions such as ECG R-peaks and phasic EDA fluctuations. By uniting high performance, robustness, and interpretability, this work advances stress detection research and offers a physiologically grounded framework for continuous mental health monitoring.
SN  - 1932-6203
ER  -