# Bidirectional Fusion Guided by Cardiac Patterns for Semi-Supervised ECG Segmentation Source: [Bidirectional Fusion Guided by Cardiac Patterns for Semi-Supervised ECG Segmentation]() Collection index: [llms.txt]() Data updated: 2026-05-19T05:00:05.916Z Authors: Jeonghwa Lim, Minje Park, Sunghoon Joo Published: 2026-05-15T08:18:42.000Z ## Abstract Accurate delineation of electrocardiogram (ECG), the segmentation of meaningful waveform features, is crucial for cardiovascular diagnostics. However, the scarcity of annotated data poses a significant challenge for training deep learning models. Conventional semi-supervised semantic segmentation (SemiSeg) methods primarily focus on consistency from unlabeled data, underutilizing the information exchange possible between labeled and unlabeled sets. To address this, we introduce CardioMix, a framework built on a bidirectional CutMix strategy guided by cardiac patterns for ECG segmentation. This approach enriches the labeled set with realistic variations from unlabeled data while simultaneously applying stronger supervisory signals to the unlabeled set, as the cardiac pattern-guided mixing ensures all augmented samples remain physiologically meaningful. Our framework is designed as a plug-and-play module, demonstrating high compatibility with various SemiSeg algorithms. Extensive experiments on SemiSegECG, a public multi-dataset benchmark for ECG delineation, demonstrate that CardioMix consistently outperforms existing CutMix-based fusion strategies across diverse datasets and labeled ratios as a plug-and-play module compatible with various SemiSeg algorithms. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [ECG Segmentation on ISP (test)](): ECG Segmentation; ISP (test); mIoU: 82.5. - [ECG segmentation on LUDB 1/16 labeled ratio N=84](): ECG Segmentation; LUDB 1/16 labeled ratio N=84; mIoU: 78.4. - [ECG Segmentation on QTDB S1 (1/16 (N=22))](): ECG Segmentation; QTDB S1 (1/16 (N=22)); mIoU: 60.2. - [ECG Segmentation on QTDB S1 (1/8 (N=56))](): ECG Segmentation; QTDB S1 (1/8 (N=56)); mIoU: 62.8. - [ECG Segmentation on QTDB S1 (1/4 (N=114))](): ECG Segmentation; QTDB S1 (1/4 (N=114)); mIoU: 68.4. - [ECG Segmentation on QTDB S1 (1/2 (N=222))](): ECG Segmentation; QTDB S1 (1/2 (N=222)); mIoU: 72. - [ECG Segmentation on Zhejiang 1/16 labeled ratio](): ECG Segmentation; Zhejiang 1/16 labeled ratio; mIoU: 81.1. - [ECG segmentation on LUDB 1/8 labeled ratio N=180](): ECG Segmentation; LUDB 1/8 labeled ratio N=180; mIoU: 80.2. - [ECG segmentation on LUDB 1/4 labeled ratio N=360](): ECG Segmentation; LUDB 1/4 labeled ratio N=360; mIoU: 82.2. - [ECG segmentation on LUDB 1/2 labeled ratio N=711](): ECG Segmentation; LUDB 1/2 labeled ratio N=711; mIoU: 83.5. - [ECG Segmentation on Zhejiang 1/8 labeled ratio](): ECG Segmentation; Zhejiang 1/8 labeled ratio; mIoU: 82.1. - [ECG Segmentation on Zhejiang 1/4 labeled ratio](): ECG Segmentation; Zhejiang 1/4 labeled ratio; mIoU: 82.9. - [ECG Segmentation on Zhejiang 1/2 labeled ratio](): ECG Segmentation; Zhejiang 1/2 labeled ratio; mIoU: 83.7. ## Original sources and implementation [Original paper]() [PDF]()