---
title: "Semantics-Guided Representation Learning for Long-Tailed ECG Classification"
authors:
  - "Fengyi Guo"
  - "Ying An"
  - "Jianxin Wang"
doi: "10.24963/ijcai.2026/613"
url: "https://ijcai.org/proceedings/2026/613"
pdf: "https://ijcai.org/proceedings/2026/0613.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "5504-5512"
track: "Main Track"
keywords:
  - "Multidisciplinary Topics and Applications: Health and medicine"
  - "Machine Learning: Multi-modal learning"
  - "Machine Learning: Representation learning"
---

# Semantics-Guided Representation Learning for Long-Tailed ECG Classification

Fengyi Guo, Ying An, Jianxin Wang

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 5504-5512. DOI: [10.24963/ijcai.2026/613](https://doi.org/10.24963/ijcai.2026/613)

## Abstract

Deep learning has achieved remarkable success in automated electrocardiogram (ECG) diagnosis.   However, the long-tailed distribution of  real-world ECG data remains a critical challenge. Existing methods mainly operate at the signal feature level and fail to exploit  semantic relationships among diagnostic labels, which limits their ability to effectively recognize rare categories. To address this issue, we propose a Semantics-Guided Representation Learning (SGRL) framework, designed to leverage semantic priors to guide feature learning and thereby correct feature bias under long-tailed distributions.  Specifically, we propose a Label Semantic Prompt Adaptation (LSPA) strategy, which introduces ECG-calibrated learnable parameters into label prompts to generate adaptive label semantic embeddings under a semantic-consistency constraint, thereby overcoming the limitations of coarse-grained label prompts in characterizing fine-grained intra-class morphological variations.  Subsequently, we propose Semantic-Guided Representation Decorrelation (SGRD) that fuses ECG embeddings with adaptive label semantics and suppresses inter-class semantic correlations to obtain discriminative class semantic anchors, thereby improving tail-class separability. Extensive experiments on multiple public datasets demonstrate the significant superiority of SGRL in long-tailed ECG classification tasks. Our code is available on https://github.com/gfywudi/SGRL.