Key result
Contrastive SSL detects mental stress from ECGs with ~0.72 AUROC using 10% labeled data.
Why the study?
Supervised deep learning for ECG-based mental stress detection requires large amounts of costly labeled data and struggles to generalize across datasets. Contrastive self-supervised learning can leverage accessible unlabeled physiological data, but remains underexplored for ECG-based mental stress detection.
Do time-series-specific contrastive self-supervised learning methods improve label-efficient ECG-based mental stress detection and cross-dataset generalization compared to supervised deep learning and feature-engineered baselines?
Do time-series-specific contrastive self-supervised learning methods improve label-efficient ECG-based mental stress detection and cross-dataset generalization compared to supervised deep learning and feature-engineered baselines?
Time-series-specific contrastive self-supervised learning enables label-efficient ECG-based mental stress detection by recovering most of the performance of fully supervised models with only 10% of labeled data, though feature-engineered baselines remain highly competitive across datasets.
Answers are grounded in this paper and cite their sources.
Cautious exploration of time-series methods for stress detection warranted; leaves open clinical translation pending larger validation studies.
Mental stress contributes to heart disease, depression, and lowered quality of life, yet traditional assessment methods such as self-reports are unsuitable for continuous, real-world monitoring. Electrocardiogram (ECG)-based mental stress detection using machine learning offers a non-invasive alternative, but progress faces key challenges: supervised deep learning requires large amounts of costly labeled data, while unlabeled physiological data are readily accessible. Models also struggle to generalize across datasets due to domain shifts, especially when trained on small datasets. Contrastive self-supervised learning (SSL) shows promise by learning from unlabeled data, but remains underexplored for ECG-based mental stress detection. We present, to our knowledge, one of the first comparative evaluations of multiple time-series-specific contrastive SSL methods for ECG-only mental stress detection, comparing four approaches (SimCLR, TS-TCC, TS2Vec, InfoTS) against supervised deep learning and feature-engineered baselines. Using a dataset of 127 participants across 26 experimental conditions, we evaluated label efficiency over different annotation budgets (1%–100%) and cross-dataset generalization to two external datasets. Time-series-specific methods substantially outperformed the general-purpose SimCLR framework, with TS-TCC achieving 0.719 [0.703, 0.734] AUROC by leveraging unlabeled ECG data and only 10% of labels, recovering 90% of fully supervised CNN performance (74% after baseline adjustment; CNN trained on 100% of labels: 0.796 [0.779, 0.814] AUROC). Cross-dataset evaluation showed that fine-tuning did not meaningfully outperform training from scratch for either SSL or supervised encoders, and that feature-engineered baselines remained highly competitive. Our findings show the potential of contrastive SSL for label-efficient ECG-based mental stress detection, while underlining the continued value of domain knowledge.
No takes yet. Share an insight, caveat, or question.
Uendes et al. (2026) studied Mental stress (n=127). Contrastive self-supervised learning (SimCLR, TS-TCC, TS2Vec, InfoTS) vs. Supervised deep learning (CNN) and feature-engineered baselines was evaluated on ECG-based mental stress detection performance (AUROC) (AUROC 0.719, 95% CI 0.703-0.734). Contrastive SSL (TS-TCC) with 10% labeled ECG data achieved 0.719 AUROC (95% CI 0.703-0.734) for mental stress detection vs 0.796 AUROC for a fully supervised CNN.
Synapse has enriched 10 related papers. Consider them for comparative context: