MedGemma | ATRIA EchoTrace: Echocardiographic Cardiac Chamber Contour Tracing

[habanwer]

ATRIA | EchoTrace
Artifact Transformation & Resource Interoperability in Artificial Intelligence
| LLM-tuned Data-Driven Echocardiographic Contour Tracing

---

Notebook on Google Colab:
https://colab.research.google.com/drive/1qofahQ8LztTrB_Us9j1Iyz2aYeS2_2rH

---

Core idea:
Fine-tune MedGemma-1.5-4B-it (via QLoRA) to output polygon coordinates for tracing the Left Ventricle endocardial border on echocardiographic frames (apical 2CH/4CH views, ED/ES phases). It includes:

• Data preparation pipeline (from what looks like a CAMUS-style dataset with polygon annotations).
• Conversational SFT format tailored for polygon JSON output.
• Custom data collator + TRL SFTTrainer with QLoRA.
• Evaluation with Dice and IoU metrics.
• An interactive HTML/JS human-in-the-loop editor (model predicts → user corrects points on canvas → saves JSON + visualizations).

---

REFERENCES

• Google Health - MedGemma GitHub Repository: Code and notebooks on https://github.com/Google-Health/medgemma
• Google MedGemma 1.5 Google. (2026). MedGemma 1.5: Technical reports and model card (google/medgemma-1.54b-it). Hugging Face. https://huggingface.co/google/medgemma-1.5-4b-it
• CAMUS Dataset Leclerc, S., Smistad, E., Pedrosa, J., Østvik, A., Cervenansky, F., Espinosa, F., Espeland, T., Berg, E. A. R., Jodoin, P.-M., Grenier, T., Lartizien, C., Dhooge, J., Løvstakken, L., & Bernard, O. (2019). Deep learning for segmentation using an open large-scale dataset in 2D echocardiography. IEEE Transactions on Medical Imaging , 38 (9), 2198–2210. https://doi.org/10.1109/tmi.2019.2900516
• EchoNet-Dynamic Ouyang, D., He, B., Ghorbani, A., Yuan, N., Ebinger, J., Langlotz, C. P., Heidenreich, P. A., Harrington, R. A., Liang, D. H., Ashley, E. A., & Zou, J. Y. (2020). Video-based AI for beat-to-beat assessment of cardiac function. Nature , 580 (7802), 252–256. https://doi.org/10.1038/s41586-020-2145-8
• LoRA: Low-Rank Adaptation . Hu, E. J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., & Chen, W. (2021). LoRA: Low-rank adaptation of large language models. arXiv. https://doi.org/10.48550/arXiv.2106.09685
• The Adimension & DEITY Principles Anwer, S. (2026). The Adimension: Bridging human ingenuity and machine intelligence through the DEITY principles framework. European Heart Journal - Imaging Methods and Practice , 4 (1), qyaf038. https://doi.org/10.1093/ehjimp/qyaf038

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[habanwer]

A Contributor License Agreement (CLA) has been already signed - email: shehab.anwer@gmail.com