GEM

🚀 GEM: Empowering MLLM for Grounded ECG Understanding with Time Series and Images

GEM is a multimodal LLM that integrates ECG time series, 12 - lead ECG images, and text for grounded and clinician - aligned ECG interpretation, enabling feature - grounded analysis, evidence - driven reasoning, and a clinician - like diagnostic process.

🚀 Quick Start

This section provides essential links for the GEM project, including the project page, paper, code, model, and data.

✨ Features

GEM is a multimodal large language model (MLLM) that combines ECG time series, 12 - lead ECG images, and text. It offers the following features:

• Grounded Analysis: Enables feature - grounded analysis of ECG data.
• Evidence - Driven Reasoning: Supports evidence - driven reasoning in the diagnostic process.
• Clinician - Aligned Interpretation: Provides a diagnostic process similar to that of clinicians.

📚 Documentation

📋 Updates

• Project Page: 📄 Page
• Paper: 📄 Arxiv
• Code: 💻 GitHub
• Model: 🤖 GEM
• Data: 📊 ECG - Grounding

📖 Citation

If you find GEM helpful for your research and applications, please cite our paper:

@misc{lan2025gemempoweringmllmgrounded,
      title={GEM: Empowering MLLM for Grounded ECG Understanding with Time Series and Images}, 
      author={Xiang Lan and Feng Wu and Kai He and Qinghao Zhao and Shenda Hong and Mengling Feng},
      year={2025},
      eprint={2503.06073},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2503.06073}, 
}

📄 License

This project is licensed under the Apache - 2.0 license.