Reasoning Before Diagnosis: Physician-Inspired Structured Thinking for ECG Classification
A new framework called CardioThink has been proposed to enhance ECG classification by incorporating structured reasoning inspired by physicians. This approach aims to improve diagnostic accuracy and clinical alignment by modeling the reasoning process through interpretable stages. The findings suggest that moving beyond direct label prediction towards structured reasoning can significantly enhance the quality of ECG diagnostics.
- ▪CardioThink is a physician-inspired multimodal large language model framework for ECG classification.
- ▪The framework emphasizes structured reasoning over direct label prediction, improving interpretability and clinical alignment.
- ▪Extensive experiments show that CardioThink significantly outperforms existing methods in diagnostic accuracy.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.17308 (cs) [Submitted on 17 May 2026] Title:Reasoning Before Diagnosis: Physician-Inspired Structured Thinking for ECG Classification Authors:Yang Wu, Xiaoyan Yuan, Hau-San Wong, Xiping Hu View a PDF of the paper titled Reasoning Before Diagnosis: Physician-Inspired Structured Thinking for ECG Classification, by Yang Wu and 3 other authors View PDF HTML (experimental) Abstract:Electrocardiogram (ECG) diagnosis in clinical practice relies on structured reasoning over multiple hierarchical aspects, including cardiac rhythm, conduction properties, waveform morphology, and overall diagnostic impression.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
