mlsanityData to 14 Oct 2026

BenchECG and xECG: a benchmark and baseline for ECG foundation models

arXiv 2509.10151 ↗First posted 12 Sep 2025cs.LG cs.AIcited by 9

Not on the map. Why?

Abstract (arXiv)

Electrocardiograms (ECGs) are inexpensive, widely used, and well-suited to deep learning. Recently, interest has grown in developing foundation models for ECGs - models that generalise across diverse downstream tasks. However, consistent evaluation has been lacking: prior work often uses narrow task selections and inconsistent datasets, hindering fair comparison. Here, we introduce BenchECG, a standardised benchmark comprising a comprehensive suite of publicly available ECG datasets and versatile tasks. We also propose xECG, an xLSTM-based recurrent model trained with SimDINOv2 self-supervised learning, which achieves the best BenchECG score compared to publicly available state-of-the-art models. In particular, xECG is the only publicly available model to perform strongly on all datasets and tasks. By standardising evaluation, BenchECG enables rigorous comparison and aims to accelerate progress in ECG representation learning. xECG achieves superior performance over earlier approaches, defining a new baseline for future ECG foundation models.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 34 of its 70 references that are in the map. First posted on arXiv 12 Sep 2025.

Builds on

34 of 70 references.

References · 34

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  4. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  5. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  6. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  7. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  8. 2020Scaling Laws for Neural Language Modelscited by 6,586
  9. 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
  10. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,848
  11. 2023Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecturecited by 1,039
  12. 2021Scaling Vision Transformerscited by 999
22 more
  1. 2017Deep Reinforcement Learning that Matterscited by 754
  2. 2024xLSTM: Extended Long Short-Term Memorycited by 327
  3. 2025Perception Encoder: The best visual embeddings are not at the output of the networkcited by 326
  4. 2020Improving Reproducibility in Machine Learning Research (A Report from the NeurIPS 2019 Reproducibility Program)cited by 322
  5. 2020How Well Do Self-Supervised Models Transfer?cited by 213
  6. 2019Automatic diagnosis of the 12-lead ECG using a deep neural networkcited by 164
  7. 2020Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scanscited by 158
  8. 2025TiRex: Zero-Shot Forecasting Across Long and Short Horizons with Enhanced In-Context Learningcited by 120
  9. 2021MedPerf: Open Benchmarking Platform for Medical Artificial Intelligence using Federated Evaluationcited by 88
  10. 2024Guiding Masked Representation Learning to Capture Spatio-Temporal Relationship of Electrocardiogramcited by 84
  11. 2024Vision-LSTM: xLSTM as Generic Vision Backbonecited by 51
  12. 2022Lead-agnostic Self-supervised Learning for Local and Global Representations of Electrocardiogramcited by 38
  13. 2018Deep learning cardiac motion analysis for human survival predictioncited by 31
  14. 2024Unlocking the Power of LSTM for Long Term Time Series Forecastingcited by 29
  15. 2025Simplifying DINO via Coding Rate Regularizationcited by 21
  16. 2024Learning General Representation of 12-Lead Electrocardiogram with a Joint-Embedding Predictive Architecturecited by 19
  17. 2024Bio-xLSTM: Generative modeling, representation and in-context learning of biological and chemical sequencescited by 13
  18. 2022Advancing the State-of-the-Art for ECG Analysis through Structured State Space Modelscited by 9
  19. 2024A Systematic Review on Foundation Models for Electrocardiogram Analysis: Initial Strides and Expansive Horizonscited by 8
  20. 2024CardioLab: Laboratory Values Estimation from Electrocardiogram Features - An Exploratory Studycited by 5
  21. 2023AIOSA: An approach to the automatic identification of obstructive sleep apnea events based on deep learningcited by 2
  22. 2025xLSTM-ECG: Multi-label ECG Classification via Feature Fusion with xLSTMcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 9.

9 papers.

  1. 2026Echo2ECG: Enhancing ECG Representations with Cardiac Morphology from Multi-View Echoscited by 2
  2. 2026SignalMC-MED: A Multimodal Benchmark for Evaluating Biosignal Foundation Models on Single-Lead ECG and PPGcited by 1
  3. 2026Pretraining Strategies and Scaling for ECG Foundation Models: A Systematic Studycited by 1
  4. 2026DeepArrhythmia: Segment-Contextualized ECG Arrhythmia Classification via Selective Evidence Acquisitioncited by 1
  5. 2026How Do Electrocardiogram Models Scale?cited by 1
  6. 2026Cross-Modal Contrastive Learning of ECG and Angiography Representations for Severe Stenosis Classificationcited by 1
  7. 2026Next-Token Prediction Learns Generalisable Representations of Sleep Physiologycited by 1
  8. 2026Looking Beyond Accuracy: A Holistic Benchmark of ECG Foundation Modelscited by 0
  9. 2026Do ECG Foundation Models Transfer to Rare Cardiac Diseases? Evidence from Brugada Syndrome Detectioncited by 0