Pulseflow creates realistic heart rhythm changes for better AF detection
Pulseflow: PPG Counterfactual Generation Via Latent Transport
Machine Learning
Summary
Detecting atrial fibrillation (AF), a heart rhythm problem, needs lots of data, but such labeled data is often scarce. The authors developed PulseFlow, a method that can take an existing heart signal and cleverly modify it to simulate AF or normal rhythms while keeping the original person's heartbeat details. This makes it possible to create more training samples from limited data. Their tests show that PulseFlow can transform heart rhythms effectively and helps improve AF detection when there is a shortage of labeled recordings.
What this means in practice
- •For hospital data teams: Enhance AF detection models by augmenting limited labeled heart rhythm data with source-conditioned synthetic signals.
- •For wearable device developers: Improve arrhythmia detection algorithms by training on realistic counterfactual PPG signals generated from existing user data.
Authors
Hung Manh Pham, Dong Ma, Bin Zhu, Pan Zhou
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.