AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Benchmarking ECG Foundational Models: A Reality Check Across Clinical Tasks

M A Al-Masud, Juan Lopez Alcaraz, Nils Strodthoff

ICLR 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use S4 backbone instead of Net1D CNN in supervised trainingsupervised baseline · PTB-XL (all)macro AUROC: 0.929 -> 0.941Improved
use ECG-CPC SSM foundation model instead of supervised S4finetuning · PTB-XL (all)macro AUROC: 0.941 -> 0.949Improved
use ECG-CPC instead of supervised S4finetuning · EchoNext (Echo)macro AUROC: 0.819 -> 0.831Improved
use ECG-FM instead of supervised S4finetuning · MIMIC (Age)standardized MAE: 0.455 -> 0.412Improved
use ST-MEM foundation model instead of supervised S4finetuning · MIMIC (Age)standardized MAE: 0.455 -> 0.504Hurt
pretrain HuBERT-ECG with masked language modeling at 9.1M samples instead of S4 from scratchfinetuning · MIMIC (Age)standardized MAE: 0.455 -> 0.579Hurt
pretrain with ECG-JEPA instead of training S4 from scratch (label efficiency)EchoNext scaling · EchoNextlabel efficiency ratio: 0.11-0.42Improved
pretrain with ECG-CPC instead of S4 from scratch (label efficiency)EchoNext scaling · EchoNextlabel efficiency ratio: 0.21-0.40Improved
pretrain with ECGFounder instead of S4 from scratch (label efficiency)EchoNext scaling · EchoNextlabel efficiency ratio: 0.30-0.62Improved
add layer-dependent learning rates in finetuning (ablation)finetuned FMs · PTB-XL, EchoNext, CPSC-ExtraperformanceImproved
use 2.5-second crops with test-time averaging instead of full 10-second inputs (ablation)finetuning · ECG benchmarkperformanceImproved
use linear head instead of learnable query-attention head in frozen evaluation for ECG-CPCECG-CPC · ECG benchmarkrankingHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy