Evaluating Long-Range Temporal Structure in Foundation Model-Based Forecasts of Heartbeat Dynamics
Abstract
We examine the long-range temporal structure of forecasts produced by Time-Series Foundation Models (TSFMs) on heartbeat dynamics using the MIT-BIH Normal Sinus Rhythm Database (NSRDB). Our findings indicate that these models do not adequately capture long-range dependencies, as reflected in growing errors in RR-interval predictions over longer forecast horizons. Code is available at https://github.com/SubramanianLab/ecg-tsfm-benchmark.
02 /Specification & Classification
Click any tag to browse matching papers- Citations
- 0
- Organism
- Assay
- Research Focus
- Approach
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- Lognormal Neural Point Process Models for Interpretable Heartbeat Dynamics
Bhuvana Ravikumar, Bharath Ramsundar, Sandya Subramanian·bioRxiv·Aug 20, 2026
- The Feasibility of Predicting Impending Malignant Ventricular Arrhythmias on the basis of Signal Complexity of Heartbeat Intervals
Zheng Chen, Koshiro Kido, … Ming Huang·bioRxiv·Dec 25, 2019
- Model-based concept to extract heart beat-to-beat variations beyond respiratory arrhythmia and baroreflex
M. Baranauskas, R. Stanikunas, … A. Lukosevicius·bioRxiv·Mar 23, 2023
- A Convex Point Process Model of Heartbeat Dynamics for Inference, Prediction, and Information Quantification
Andrew S. Perley, Matthew Martinez, … Todd P. Coleman·bioRxiv·Aug 28, 2025
- Enhanced EEG Forecasting: A Probabilistic Deep Learning Approach
Hanna Pankka, Jaakko Lehtinen, … Timo Roine·Neural ComputationPeer-reviewed·Mar 3, 2025
- Resting-state heartbeat-evoked potentials are associated with Kalman-derived cardiac prediction errors
Takahiro Hirao, Kokono Terada, … Makiko Yamada·bioRxiv·May 18, 2026