Key result
Dual-input CNN-BiLSTM improves ECG-based sleepiness detection to ~65% external accuracy vs 1D CNN.
Why the study?
Fatigue and sleepiness arise under varying conditions, yet ECG-based detection models are typically evaluated only within their development dataset rather than across independent datasets and paradigms.
Does a dual-input CNN-BiLSTM model combining RR interval sequences with HRV descriptors improve participant-independent and cross-dataset classification of subjective sleepiness compared to an HRV-based 1D CNN in healthy adults undergoing fatigue induction?
Observational (n=34)
No
Does a dual-input CNN-BiLSTM model combining RR interval sequences with HRV descriptors improve participant-independent and cross-dataset classification of subjective sleepiness compared to an HRV-based 1D CNN in healthy adults undergoing fatigue induction?
ECG-derived HRV and RR interval deep learning models retain moderate discriminative ability for detecting subjective sleepiness when evaluated across unseen participants and heterogeneous external experimental datasets without retraining.
Answers are grounded in this paper and cite their sources.
Moderate ECG-based sleepiness detection accuracy may support further model development; leaves open clinical adoption pending larger validation studies.
Introduction Fatigue and sleepiness can arise under different experimental and behavioral conditions, yet ECG-based detection models are often evaluated only within the dataset used for development. This study investigated the participant-independent and cross-dataset performance of two ECG-derived approaches: a 1D CNN based on six HRV features and a dual-input CNN-BiLSTM combining RR interval sequences with statistical HRV descriptors. Methods Models were developed using the DROZY database, in which sleepiness was induced by prolonged wakefulness. Internal performance was assessed using five-fold subject-wise evaluation on 25 recordings from 14 participants. The models were then applied without retraining or parameter adaptation to the independent FatiQData dataset comprising 20 participants performing a prolonged monotonous driving task. Results On DROZY, the HRV-based 1D CNN achieved 60.0% accuracy, 61.5% F1-score, and a ROC-AUC of 0.593, whereas the dual-input CNN-BiLSTM achieved 68.0% accuracy, 73.3% F1-score, and a ROC-AUC of 0.753. In FatiQData, Stanford Sleepiness Scale scores increased significantly from 2.40 ± 0.82 to 4.15 ± 0.81. For binary external discrimination, the HRV-based model achieved 60.0% accuracy, 69.2% F1-score, and a ROC-AUC of 0.766, while the dual-input model achieved 65.0% accuracy, 77.4% F1-score, and a ROC-AUC of 0.750. Discussion The results indicate that ECG-derived HRV and RR dynamics retain information related to subjective sleepiness, while participant-independent and cross-dataset performance remains moderate. These findings emphasize the importance of subject-wise validation, independent external testing, and cautious interpretation across heterogeneous experimental conditions.
No takes yet. Share an insight, caveat, or question.
Králiková et al. (2026) conducted an observational in Fatigue and sleepiness (n=34). Dual-input CNN-BiLSTM model (combining RR interval sequences and statistical HRV descriptors) vs. HRV-based 1D CNN model (6 HRV features) was evaluated on Binary external classification accuracy and ROC-AUC for high versus lower/intermediate subjective sleepiness on the independent FatiQData dataset. Dual-input CNN-BiLSTM achieved 68.0% accuracy (ROC-AUC 0.753) internally and 65.0% accuracy (ROC-AUC 0.750) externally for ECG-based sleepiness detection compared to 60.0% for the 1D CNN.
Synapse has enriched 10 related papers. Consider them for comparative context: