Paperdoi:10.1161/circ.154.suppl_1.16003

Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003)

Venue listed2026Added Nov 15, 2026

  • CM
  • ET
  • YL
  • MK
  • AD
  • CK
  • MA
  • YM
  • +4

Authors: Christian M. Massad, Eli Tsakiris, Yingshuo Liu, Michel Abou Khalil, Alex El Darzi, Carlo El Khoury, Mohammad Montaser Atasi, Yara Menassa, Maximilian Moersdorf, W. Herbst,

+2 moreGhassan Bidaoui, Hadi Younes

Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003 (2026-11-02)

TL;DR

First-month passive smartwatch PPG from HEARTBEAT Study participants encoded with a previously trained deep-learning encoder; patient-level embedding mean/SD fed a shallow random forest to separate CHA2DS2-VASc ≥ 4 vs ≤ 1. Held-out cohort (n=72): AUPRC 0.819, AUROC 0.813; internal 5-fold CV (n=401): mean AUPRC up to 0.825, AUROC ~0.77. Encoder identity and intermediate-risk handling not described in the abstract.

  • Form factor: wrist

Abstract

Background: Wearable photoplethysmography (PPG) from wearable devices offers continuous, noninvasive cardiovascular data. While PPG-derived heart rate and rhythm detection are well-validated, whether passive PPG signal morphology encodes broader thromboembolic and cardiovascular risk burden remains largely unexplored.

Methods: Participants enrolled in the HEARTBEAT Study with available first-month passive smartwatch PPG recordings were included. PPG signals were processed using previously trained deep-learning encoder, and patient-level embedding means and standard deviations were used as model inputs. A shallow random-forest classifier was trained to distinguish high-thromboembolic risk (CHA 2 DS 2 -VASc ≥4) from low-risk (CHA 2 DS 2 -VASc ≤1) patients. Performance was evaluated on a held-out validation cohort (n=72) and further assessed via internal 5-fold cross-validation (n=401). Area under the precision-recall curve (AUPRC) was emphasized as the primary performance metric, with area under the receiver operating characteristic curve (AUROC) reported as a complementary measure.

Results: On the held-out validation cohort, the random-forest classifier achieved AUPRC 0.819 and AUROC 0.813 for distinguishing participants with high vs low CHA 2 DS 2 -VASc scores. Internal 5-fold CV confirmed stability, with tuned random-forest models achieving mean-fold AUPRC up to 0.825 and mean-fold AUROC of approximately 0.77. Overall, AUPRC remained consistently strong across the held-out and cross-validation analysis, and AUROC further supported model discrimination and risk stratification.

Conclusion: These findings suggest that wearable PPG morphology encodes cardiovascular risk information beyond just heart rate or arrhythmia detection.

Verbatim from the publisher's DOI record (opens in a new tab) · Abstract as deposited by the publisher with Crossref; © the authors/publisher

Results on ChatPPG leaderboards

All leaderboards

No results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.

Tasks & signals

Tasks

AF / rhythm

Signals

Keywords

EmbeddingsCVD RiskCHA2DS2 VAScAFibHEARTBEATConference Abstract

How it works

How Method works: Setup → Method → TargetSetup: Signals (PPG); Form factor (wrist). Method: Method. Target: AF / rhythm.SetupSignals: PPGSignalsPPGForm factor: wristForm factorwristMethodMethodTargetAF / rhythmAF / rhythm

Cite this paper

Cite this paper with the fields recorded here. Check it against the publisher before submitting.

Massad, C. M., Tsakiris, E., Liu, Y., Khalil, M. A., Darzi, A. E., Khoury, C. E., Atasi, M. M., Menassa, Y., Moersdorf, M., Herbst, W., Bidaoui, G., & Younes, H. (2026). Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003). Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003. https://doi.org/10.1161/circ.154.suppl_1.16003
Massad, Christian M., Eli Tsakiris, Yingshuo Liu, Michel Abou Khalil, Alex El Darzi, Carlo El Khoury, and Mohammad Montaser Atasi, et al. 2026. "Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003)." Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003. https://doi.org/10.1161/circ.154.suppl_1.16003.
Massad, C. M., Tsakiris, E., Liu, Y., Khalil, M. A., Darzi, A. E., Khoury, C. E. and Atasi, M. M. et al. (2026) 'Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003)', Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003. doi: 10.1161/circ.154.suppl_1.16003.
Massad, Christian M., et al. "Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003)." Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003, 2026. https://doi.org/10.1161/circ.154.suppl_1.16003.
C. M. Massad et al., "Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003)," Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003, 2026, doi: 10.1161/circ.154.suppl_1.16003.
@article{massad2026using,
  title         = {{Using Smartwatch Photoplethysmography to Identify CHA2DS2-VASc Risk: Subanalysis from the HEARTBEAT Study (AHA 2026 Abstract 16003)}},
  author        = {Massad, Christian M. and Tsakiris, Eli and Liu, Yingshuo and Khalil, Michel Abou and Darzi, Alex El and Khoury, Carlo El and Atasi, Mohammad Montaser and Menassa, Yara and Moersdorf, Maximilian and Herbst, W. and Bidaoui, Ghassan and Younes, Hadi},
  year          = {2026},
  journal       = {Circulation 154 suppl 1, AHA Scientific Sessions 2026, Abstract 16003},
  doi           = {10.1161/circ.154.suppl_1.16003}
}

Added to the ChatPPG research hub