Pathology-Aware Multi-View Contrastive Learning for Patient-Independent ECG Reconstruction

Youssef Youssef; Jitin Singla

cs.LGMar 18, 2026

Pathology-Aware Multi-View Contrastive Learning for Patient-Independent ECG Reconstruction

Authors: Youssef Youssef, Jitin Singla

Organizations: Department of Computer Science and Engineering, Indian Institute of Technology Roorkee, Roorkee 247667, India · Department of Biosciences and Bioengineering, Indian Institute of Technology Roorkee, Roorkee 247667, India

Abstract

Reconstructing a 12-lead electrocardiogram (ECG) from a reduced lead set is an ill-posed inverse problem due to anatomical variability. Standard deep learning methods often ignore underlying cardiac pathology losing vital morphology in precordial leads. We propose Pathology-Aware Multi-View Contrastive Learning, a framework that regularizes the latent space through a pathological manifold. Our architecture integrates high-fidelity time-domain waveforms with pathology-aware embeddings learned via supervised contrastive alignment. By maximizing mutual information between latent representations and clinical labels, the framework learns to filter anatomical "nuisance" variables. On the PTB-XL dataset, our method achieves approx. 76% reduction in RMSE compared to state-of-the-art model in patient-independent setting. Cross-dataset evaluation on the PTB Diagnostic Database confirms superior generalization, bridging the gap between hardware portability and diagnostic-grade reconstruction.

Figures & tables

Fig. 1 : Proposed ECG reconstruction framework. (a) Preprocessing and segmentation, (b) Multi-View Supervised contrastive pretraining to learn pathology-aware latent representations, and (c) Reconstruction network that integrates contrastive representation with clean ECG signals via a temporal decoder.

Fig. 2 : k-NN affinity matrices ( $k=10$ ) comparing (a) the clean baseline $\mathbf{x}$ and (b) the learned latent representation $\mathbf{h}$ .

Lead	RMSE (mV) $\downarrow$			$R^{2}$ $\uparrow$			Pearson $\uparrow$
	C	C-h	$\Delta\%$	C	C-h	$\Delta\%$	C	C-h	$\Delta\%$
$V_{1}$	0.063	0.052	+18.2%	0.754	0.839	+11.2%	0.940	0.951	+1.2%
$V_{3}$	0.115	0.101	+12.7%	0.670	0.744	+11.0%	0.913	0.926	+1.5%
$V_{4}$	0.118	0.108	+8.7%	0.605	0.687	+13.5%	0.893	0.906	+1.4%
$V_{5}$	0.094	0.084	+10.3%	0.684	0.717	+4.8%	0.922	0.931	+1.0%
$V_{6}$	0.071	0.063	+11.4%	0.699	0.760	+8.8%	0.935	0.944	+1.0%

TABLE I : Reconstruction Performance: Baseline (C) vs. Proposed (C-h)

Fig. 3 : Radar plot showing lower RMSE for C-h model for few most and least abundant diagnostic class. $n$ represents the number of segments in test fold. The closer the vertex to center, the lower is the RMSE. The y-values for each ring correspond to [0.05, 0.075, 0.10, 0.125, 0.15] starting from center.

Panel of Figure 5 — Fig. 4 : Qualitative heartbeat-level reconstruction examples across different diagnostic classes (NORM, INJAS, RVH) for leads V1,V3–V6. The C-h configuration more closely follows the ground truth compared to the C baseline, particularly in capturing QRS complex amplitude and T-wave morphology.

Lead	Proposed (C-h)		Rajotte et al. [ 26 ]
	RMSE	$R^{2}$	RMSE	$\Delta RMSE\%$	$R^{2}$	$\Delta R^{2}\%$
$V_{1}$	$0.054\pm 0.002$	$0.839\pm 0.010$	$0.276\pm 0.009$	80.4%	$0.796\pm 0.011$	5.4%
$V_{3}$	$0.103\pm 0.003$	$0.747\pm 0.023$	$0.364\pm 0.011$	71.7%	$0.702\pm 0.023$	6.4%
$V_{4}$	$0.110\pm 0.004$	$0.692\pm 0.026$	$0.420\pm 0.014$	73.8%	$0.613\pm 0.040$	12.9%
$V_{5}$	$0.086\pm 0.003$	$0.719\pm 0.037$	$0.350\pm 0.011$	75.4%	$0.694\pm 0.023$	3.6%
$V_{6}$	$0.065\pm 0.002$	$0.765\pm 0.028$	$0.334\pm 0.012$	80.5%	$0.729\pm 0.018$	4.9%

TABLE II : Record-level Comparison with Rajotte et al. [ 26 ]

Method	Patient Indep.	Input Leads	RMSE	Pearson
Garg et al. (2023) [ 19 ]	No	II	–	0.805
Lu et al. (2025) [ 25 ]	No	$V_{2}$	0.160	0.860
Hebiguchi et al. (2025) [ 18 ]	No	I, II, $V_{2}$	-	0.93
Ours	Yes	I, II, $V_{2}$	0.082	0.932

TABLE III : Global Performance Benchmarks on PTB-XL

Method	Patient Indep.	Input Leads	RMSE	Pearson
Smith et al. (2021) [ 23 ]	No	6 limb leads + $V_{2}$	0.173	0.921
Moghaddam et al. (2025) [ 4 ]	Yes	3-leads	–	0.76
Ours	Yes	I, II, $V_{2}$	0.095	0.887

TABLE IV : Cross-Dataset Performance Comparison on the PTB Database

Explore similar work

Learning ECG Image Representations via Dual Physiological-Aware Alignments
Apr 2, 2026Hung Manh Pham, Jialu Tang, Aaqib Saeed +3Medical Image Representation Learning Electrocardiography
University of Cambridge · Singapore Management University · Eindhoven University of Technology, Netherlands
Graph-Based Pseudo-multimodal Contrastive Learning for 12-Lead ECG Representations
Aug 27, 2026Mengyu Wang, Kozo Okada, Takafumi Goto +4Electrocardiogram Classification Time Series Representation Learning
Graduate School of Engineering Science Yokohama National University Yokohama, Japan · Division of Cardiology Yokohama City University Medical Center Yokohama, Japan · Technology and Innovation Department Fukuda Denshi Co.,Ltd Tokyo, Japan
CoRe-ECG: Advancing Self-Supervised Representation Learning for 12-Lead ECG via Contrastive and Reconstructive Synergy
Apr 13, 2026Zehao Qin, Xiaojian Lin, Ping Zhang +6Representation Learning Electrocardiography
Department of Electronic Engineering, Shenzhen International Graduate School, Tsinghua University, Shenzhen 518071, PR China · Department of Cardiology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, Beijing, 102218, PR China · Department of Anesthesiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100021, PR China +3

Apr 2, 2026J/K move · Enter open · S save

Learning ECG Image Representations via Dual Physiological-Aware Alignments

Hung Manh Pham, Jialu Tang, Aaqib Saeed, Dong Ma+2

University of Cambridge · Singapore Management University · Eindhoven University of Technology, Netherlands

Medical Image Representation Learning Electrocardiography Self-Supervised Visual Representation Learning Multimodal Contrastive Learning

Open paper AskPDF arXiv

Electrocardiograms (ECGs) are among the most widely used diagnostic tools for cardiovascular diseases, and a large amount of ECG data worldwide appears only in image form. However, most existing automated ECG analysis methods rely on access to raw signal recordings, limiting their applicability in real-world and resource-constrained settings. In this paper, we present ECG-Scan, a self-supervised framework for learning clinically generalized representations from ECG images through dual physiological-aware alignments: 1) Our approach optimizes image representation learning using multimodal contrastive alignment between image and gold-standard signal-text modalities. 2) We further integrate domain knowledge via soft-lead constraints, regularizing the reconstruction process and improving signal lead inter-consistency. Extensive benchmarking across multiple datasets and downstream tasks demonstrates that our image-based model achieves superior performance compared to existing image baselines and notably narrows the gap between ECG image and signal analysis. These results highlight the potential of self-supervised image modeling to unlock large-scale legacy ECG data and broaden access to automated cardiovascular diagnostics.

Abstract

Explore similar work

Learning ECG Image Representations via Dual Physiological-Aware Alignments

Graph-Based Pseudo-multimodal Contrastive Learning for 12-Lead ECG Representations

CoRe-ECG: Advancing Self-Supervised Representation Learning for 12-Lead ECG via Contrastive and Reconstructive Synergy