mlsanityData to 14 Oct 2026

GeoECG: Data Augmentation via Wasserstein Geodesic Perturbation for Robust Electrocardiogram Prediction

arXiv 2208.01220 ↗First posted 2 Aug 2022stat.ML cs.LG eess.SPcited by 8

Not on the map. Why?

Abstract (arXiv)

There has been an increased interest in applying deep neural networks to automatically interpret and analyze the 12-lead electrocardiogram (ECG). The current paradigms with machine learning methods are often limited by the amount of labeled data. This phenomenon is particularly problematic for clinically-relevant data, where labeling at scale can be time-consuming and costly in terms of the specialized expertise and human effort required. Moreover, deep learning classifiers may be vulnerable to adversarial examples and perturbations, which could have catastrophic consequences, for example, when applied in the context of medical treatment, clinical trials, or insurance claims. In this paper, we propose a physiologically-inspired data augmentation method to improve performance and increase the robustness of heart disease detection based on ECG signals. We obtain augmented samples by perturbing the data distribution towards other classes along the geodesic in Wasserstein space. To better utilize domain-specific knowledge, we design a ground metric that recognizes the difference between ECG signals based on physiologically determined features. Learning from 12-lead ECG signals, our model is able to distinguish five categories of cardiac conditions. Our results demonstrate improvements in accuracy and robustness, reflecting the effectiveness of our data augmentation method.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 49 of its 90 references that are in the map. First posted on arXiv 2 Aug 2022.

Builds on

49 of 90 references.

References · 49

  1. 2017Attention Is All You Needcited by 60,361
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  3. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  4. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
  5. 2013Intriguing properties of neural networkscited by 5,155
  6. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,548
  7. 2014Sequence to Sequence Learning with Neural Networkscited by 2,986
  8. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,316
  9. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,340
  10. 2016Adversarial Machine Learning at Scalecited by 936
  11. 2018Generalizing to Unseen Domains via Adversarial Data Augmentationcited by 451
  12. 2019Unlabeled Data Improves Adversarial Robustnesscited by 416
37 more
  1. 2017Certifying Some Distributional Robustness with Principled Adversarial Trainingcited by 337
  2. 2019Provably Robust Deep Learning via Adversarially Trained Smoothed Classifierscited by 330
  3. 2021Fixing Data Augmentation to Improve Adversarial Robustnesscited by 204
  4. 2021Data Augmentation Can Improve Robustnesscited by 194
  5. 2020Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Datacited by 180
  6. 2018Speaker Recognition from Raw Waveform with SincNetcited by 166
  7. 2020Deep Learning for ECG Analysis: Benchmarks and Insights from PTB-XLcited by 158
  8. 2022Prompting Decision Transformer for Few-Shot Policy Generalizationcited by 146
  9. 2020Maximum-Entropy Adversarial Data Augmentation for Improved Generalization and Robustnesscited by 121
  10. 2020Feature Purification: How Adversarial Training Performs Robust Deep Learningcited by 107
  11. 2020SoK: Certified Robustness for Deep Neural Networkscited by 105
  12. 2021Transformer-based Spatial-Temporal Feature Learning for EEG Decodingcited by 97
  13. 2020SSMBA: Self-Supervised Manifold Based Data Augmentation for Improving Out-of-Domain Robustnesscited by 96
  14. 2019Adversarially Robust Generalization Just Requires More Unlabeled Datacited by 76
  15. 2020Sequence-Level Mixed Sample Data Augmentationcited by 71
  16. 2021Minibatch optimal transport distances; analysis and applicationscited by 65
  17. 2020SeqMix: Augmenting Active Sequence Labeling via Sequence Mixupcited by 52
  18. 2020TSS: Transformation-Specific Smoothing for Robustness Certificationcited by 49
  19. 2021SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustnesscited by 49
  20. 2018Deep Learning for Electromyographic Hand Gesture Signal Classification Using Transfer Learningcited by 47
  21. 2020Debiased Sinkhorn barycenterscited by 34
  22. 2020Sharp Statistical Guarantees for Adversarially Robust Gaussian Classificationcited by 28
  23. 2021A Transformer Architecture for Stress Detection from ECGcited by 27
  24. 2020Improving Adversarial Robustness via Unlabeled Out-of-Domain Datacited by 23
  25. 2021k-Mixup Regularization for Deep Learning via Optimal Transportcited by 14
  26. 2021Fundamental Tradeoffs in Distributionally Adversarial Trainingcited by 12
  27. 2022Data Augmentation for Electrocardiogramscited by 12
  28. 2019ECGadv: Generating Adversarial Electrocardiogram to Misguide Arrhythmia Classification Systemcited by 10
  29. 2019Multimodal Emotion Recognition Using Deep Canonical Correlation Analysiscited by 10
  30. 2021ECG-Adv-GAN: Detecting ECG Adversarial Examples with Conditional Generative Adversarial Networkscited by 8
  31. 2021Functional optimal transport: map estimation and domain adaptation for functional datacited by 7
  32. 2021Robustness of convolutional neural networks to physiological ECG noisecited by 7
  33. 2022Certifying Out-of-Domain Generalization for Blackbox Functionscited by 7
  34. 2022MHMS: Multimodal Hierarchical Multimedia Summarizationcited by 6
  35. 2018Multiple Instance Learning for ECG Risk Stratificationcited by 5
  36. 2022PhysioMTL: Personalizing Physiological Patterns using Optimal Transport Multi-Task Regressioncited by 3
  37. 2021ECG-ATK-GAN: Robustness against Adversarial Attacks on ECGs using Conditional Generative Adversarial Networkscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 2; 2023: 5; 2024: 0; 2025: 0; 2026: 1.

8 papers.

  1. 2022Curriculum Reinforcement Learning using Optimal Transport via Gradual Domain Adaptationcited by 28
  2. 2023Transfer Knowledge from Natural Language to Electrocardiography: Can We Detect Cardiovascular Disease Through Language Models?cited by 21
  3. 2022Semantics-Consistent Cross-domain Summarization via Optimal Transport Alignmentcited by 6
  4. 2023Multimodal Representation Learning of Cardiovascular Magnetic Resonance Imagingcited by 4
  5. 2023Interpolation for Robust Learning: Data Augmentation on Wasserstein Geodesicscited by 3
  6. 2023Automated Cardiovascular Record Retrieval by Multimodal Learning between Electrocardiogram and Clinical Reportcited by 1
  7. 2023Learning Beyond Similarities: Incorporating Dissimilarities between Positive Pairs in Self-Supervised Time Series Learningcited by 0
  8. 2026Boosting ECG Classification Performance by Pre-training with Synthesized Datacited by 0