ChatPPG.
Join free
Foundation model

ECGTFormer (NYU Abu Dhabi; physiology-aware ECGTokenizer + causal V-JEPA ECG FM)

ECGNeurIPS 2026 Workshop GenAI4Health
Paper Cite
Venue listedCatalog paper date: 2026Added to atlas: Oct 26, 2026Source

Activity counts measure attention, not model quality. Missing counts are unknown. Publication status follows the linked source; a listed venue alone does not confirm peer review.

ECGTFormer is a compact causal ECG foundation model that operates over physiology-aware tokens for P, QRS, T, PR, ST, and TP regions per lead, with continuous-time rotary position encoding and V-JEPA-style pretraining. It supports single- or multi-lead streaming inference. Presented at the NeurIPS 2026 GenAI4Health workshop.

Signals
ECG
Form factor
Other
Year
2026
Architecture

Compact causal Transformer over ECGTokenizer tokens (per-lead P, QRS, T, PR, ST, TP regions), continuous-time rotary position encoding, V-JEPA self-supervised pretraining; single- and multi-lead input; streaming inference with prefix caching

At a glance
Type
Foundation model
Modalities
ECG
Form factor
Other
Parameters
Not reported
Architecture
Compact causal Transformer over ECGTokenizer tokens (per-lead P, QRS, T, PR, ST, TP regions), continuous-time rotary position encoding, V-JEPA self-supervised pretraining; single- and multi-lead input; streaming inference with prefix caching
Venue
NeurIPS 2026 Workshop GenAI4Health (poster)
Year
2026
arXiv
Not reported
DOI
Not reported
Code
Not reported
Weights
Not reported
Licence
Not reported
Caveats
Not reported
Datasets
Papers
Added
Oct 26, 2026
Updated
Oct 26, 2026
ECGTFormer (NYU Abu Dhabi; physiology-aware ECGTokenizer + causal V-JEPA ECG FM): ChatPPG research summary card

How it works

How ECGTFormer (NYU Abu Dhabi; physiology-aware ECGTokenizer + causal V-JEPA ECG FM) works: Input signals → Foundation model → Downstream tasksInput signals: ECG (Electrocardiogram). Dataset: PTB-XL. Foundation model: ECGTFormer (NYU Abu Dhabi; physiology-aware ECGTokenizer + causal V-JEPA ECG FM) (Transformer).Input signalsECG: ElectrocardiogramECGElectrocardiogramDataset: PTB-XLDatasetPTB-XLFoundation model×NECGTFormer (NYU Abu Dhabi…Transformer

Datasets linked in the hub; pretraining vs evaluation roles follow the paper.

Architecture
Compact causal Transformer over ECGTokenizer tokens (per-lead P, QRS, T, PR, ST, TP regions), continuous-time rotary position encoding, V-JEPA self-supervised pretraining; single- and multi-lead input; streaming inference with prefix caching
Inputs
ECG
Approach
ECG · foundation model · tokenization · V JEPA · causal · streaming

Datasets· 1

Papers· 1

Cite

APA

Zhylko, D., Shamout, F. E., & list, P. T. O. A. (2026). Rethinking ECG Tokenization: A Unified Taxonomy and Clinically-Grounded ECG Foundation Model. NeurIPS 2026 Workshop GenAI4Health. https://openreview.net/forum?id=6iqTIcGBqt

BibTeX

@inproceedings{zhylko2026rethinking,
  title         = {{Rethinking ECG Tokenization: A Unified Taxonomy and Clinically-Grounded ECG Foundation Model}},
  author        = {Zhylko, Dmytro and Shamout, Farah E. and list, per the OpenReview author},
  year          = {2026},
  booktitle     = {NeurIPS 2026 Workshop GenAI4Health},
  url           = {https://openreview.net/forum?id=6iqTIcGBqt}
}

Built from the hub record. Check it against the publisher before submitting.

History

  1. Added to the ChatPPG research hub

Connections

Open in the graph explorer
Dataset: PTB-XL (uses dataset)PTB-XLPTB-XLSignal: ECG (uses signal)ECGECGPaper: Rethinking ECG Tokenization: A Unified Taxonomy and Clinically-Grounded ECG Foundation Model (paper and model)Rethinking ECG Toke…Rethinking…ECGTFormer (NYU Abu Dhabi; ph…ECGTFormer (NYU A…
Datasets 1
PTB-XL
Signals 1
ECG

Same tasks, data and authors

Papers

Models

Datasets