PaperarXiv:2609.33501
PulseFlow: PPG Counterfactual Generation Via Latent Transport
arXiv preprint (Singapore Management University / University of Cambridge)
(opens in a new tab)Pham, Hung Manh et al., "Pulseflow: PPG Counterfactual Generation Via Latent Transport", arXiv:2609.33501, CC BY 4.0
TL;DR
Source-conditioned counterfactual PPG generation: conditional VAE representation + invertible latent transport (conditional RealNVP) to edit cardiac rhythm (SR↔AF) while retaining source demographics / residual. Trained on MIMIC-III-Ext-PPG (~5,668 patients; 30 s @ 125 Hz); same-patient SR–AF pairs guide transport. Reports strong AF conversion (~93.1%) and distributional realism vs generative baselines; on unseen VitalDB scarce-label AF classification, PulseFlow augmentation lifts AUROC (e.g.
- Form factor: lab
Abstract
Photoplethysmography (PPG) has become an important modality for continuous cardiovascular monitoring, including atrial fibrillation (AF) detection. However, labeled AF recordings remain limited in many clinical settings, making model adaptation difficult when only limited target data are available. Generative modeling offers a natural way to alleviate this scarcity by synthesizing additional AF signals. Existing approaches, however, mainly generate samples that match the target condition without explicitly modeling how an observed source recording should be transformed, making it difficult to leverage abundant source recordings from a specific population or cohort for targeted augmentation. We introduce PulseFlow, a source-conditioned counterfactual generation framework that combines conditional representation learning with invertible latent transport to edit cardiac rhythm while retaining information from the source. Experiments across two clinical cohorts demonstrate effective rhythm transformation, measurable source correspondence, and improved AF classification under limited labels.
Verbatim from arXiv (opens in a new tab) · arXiv metadata (abstracts are released under CC0)
Results on ChatPPG leaderboards
All leaderboardsNo results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.
Tasks & signals
Tasks
Signals
Keywords
How it works
Cite this paper
APA
Pham, H. M., Ma, D., Zhu, B., & Zhou, P. (2026). PulseFlow: PPG Counterfactual Generation Via Latent Transport. arXiv preprint. https://doi.org/10.48550/arXiv.2609.33501
BibTeX
@misc{pham2026pulseflow,
title = {{PulseFlow: PPG Counterfactual Generation Via Latent Transport}},
author = {Pham, Hung Manh and Ma, Dong and Zhu, Bin and Zhou, Pan},
year = {2026},
howpublished = {arXiv preprint},
doi = {10.48550/arXiv.2609.33501},
eprint = {2609.33501},
archivePrefix = {arXiv}
}Built from the hub record. Check it against the publisher before submitting.
Added to the ChatPPG research hub
