{"id":29015316,"url":"https://github.com/abdur75648/echo-dnd","last_synced_at":"2026-02-23T08:03:12.912Z","repository":{"id":299765289,"uuid":"1004080747","full_name":"abdur75648/Echo-DND","owner":"abdur75648","description":"Official implementation of Echo-DND: A Dual Noise Diffusion Model for Robust and Precise Left Ventricle Segmentation in Echocardiography","archived":false,"fork":false,"pushed_at":"2025-06-19T07:23:36.000Z","size":1482,"stargazers_count":6,"open_issues_count":0,"forks_count":3,"subscribers_count":4,"default_branch":"main","last_synced_at":"2025-10-25T11:45:59.457Z","etag":null,"topics":["computer-vision","deep-learning","diffusion-models","echo-dnd","echocardiogram","echocardiogram-segmentation","echocardiographic-videos","echocardiography","image-segmentation","left-ventricle-segmentation","medical-image-segmentation","medical-imaging"],"latest_commit_sha":null,"homepage":"https://abdur75648.github.io/Echo-DND","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/abdur75648.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-06-18T05:05:23.000Z","updated_at":"2025-07-27T12:39:27.000Z","dependencies_parsed_at":"2025-10-10T07:34:43.480Z","dependency_job_id":"1e7f3029-b391-4bf1-882a-a8941b09f002","html_url":"https://github.com/abdur75648/Echo-DND","commit_stats":null,"previous_names":["abdur75648/echo-dnd"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/abdur75648/Echo-DND","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/abdur75648%2FEcho-DND","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/abdur75648%2FEcho-DND/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/abdur75648%2FEcho-DND/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/abdur75648%2FEcho-DND/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/abdur75648","download_url":"https://codeload.github.com/abdur75648/Echo-DND/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/abdur75648%2FEcho-DND/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29739760,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-23T07:44:07.782Z","status":"ssl_error","status_checked_at":"2026-02-23T07:44:07.432Z","response_time":90,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["computer-vision","deep-learning","diffusion-models","echo-dnd","echocardiogram","echocardiogram-segmentation","echocardiographic-videos","echocardiography","image-segmentation","left-ventricle-segmentation","medical-image-segmentation","medical-imaging"],"created_at":"2025-06-25T21:05:47.554Z","updated_at":"2026-02-23T08:03:12.895Z","avatar_url":"https://github.com/abdur75648.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Echo-DND: A Dual Noise Diffusion Model for Robust and Precise Left Ventricle Segmentation in Echocardiography\n\n[![](https://img.shields.io/badge/Paper-Discover%20Applied%20Sciences%20(Springer)-green?logo=springer\u0026style=flat-square)](https://doi.org/10.1007/s42452-025-07055-5)\n[![](https://img.shields.io/badge/arXiv-Preprint-red?logo=arxiv\u0026style=flat-square)](https://arxiv.org/abs/2506.15166)\n[![](https://img.shields.io/badge/Code-GitHub-black?logo=github\u0026style=flat-square)](https://github.com/abdur75648/Echo-DND)\n[![](https://img.shields.io/badge/Project-Page-blue?style=flat\u0026logo=githubpages\u0026logoColor=white)](https://abdur75648.github.io/Echo-DND)\n\n---\n\n## 🧠 Overview\nAccurate segmentation of the left ventricle (LV) in echocardiograms is crucial for cardiac diagnostics but is challenging due to inherent noise, low contrast, and ambiguous boundaries in ultrasound images. This repository accompanies our paper introducing **Echo-DND**, a novel diffusion probabilistic model (DPM) specifically designed to address these challenges.\n\nEcho-DND introduces several key innovations:\n*   A **Synergistic Dual-Noise Strategy:** Uniquely combines Gaussian and Bernoulli noises within the diffusion framework, effectively modeling both continuous sensor-like variations and the discrete binary nature of segmentation masks.\n*   A **Multi-scale Fusion Conditioning Module (MFCM):** Employs multi-resolution feature extraction and cross-resolution fusion to preserve high-resolution spatial details, crucial for precise boundary delineation.\n*   **Spatial Coherence Calibration (SCC):** Incorporates a pixel-wise calibration technique that complements the diffusion process to maintain spatial integrity and consistency in the output segmentation masks.\n\nOur model was rigorously validated on the public CAMUS and EchoNet-Dynamic datasets, demonstrating state-of-the-art performance and establishing a new benchmark in echocardiogram LV segmentation. Echo-DND's architecture holds promise for broader applicability in other medical imaging tasks.\n\n---\n\n## 🖼️ Architecture\n\n\u003cp align=\"center\"\u003e\n  \u003cimg src=\"docs/static/images/echo_dnd_architecture.png\" width=\"700px\" alt=\"Echo-DND Architecture\"/\u003e\n\u003c/p\u003e\n\n\u003e *Figure: Overall architecture of the Echo-DND model, illustrating the dual noise (Gaussian and Bernoulli) diffusion pathways, the Multi-scale Fusion Conditioning Module (MFCM), and the integration of various loss components including Spatial Coherence Calibration (SCC).*\n\n---\n\n## ⚙️ Setup \u0026 Installation\n\n1.  **Clone the repository:**\n    ```bash\n    git clone https://github.com/abdur75648/Echo-DND.git\n    cd Echo-DND\n    ```\n\n2.  **Install Dependencies:**\n    ```bash\n    pip install -r requirements.txt\n    ```\n\n3. **Prepare the Dataset:**\n   - Download the CAMUS and EchoNet-Dynamic datasets.\n   - Organize them into a root data directory with the following structure:\n     ```\n     \u003cyour_data_root_dir\u003e/\n     ├── CAMUS/\n     │   ├── patient0001/\n     │   │   ├── patient0001_4CH_ED.mhd\n     │   │   ├── patient0001_4CH_ED_gt.mhd\n     │   │   └── ... (other patient files)\n     │   └── ... (other patient folders)\n     └── EchoNet-Dynamic/\n         ├── Train/\n         │   ├── Frames/\n         │   │   └── 0X100037609D9A4939_image0001.png\n         │   └── Masks/\n         │       └── 0X100037609D9A4939_image0001.png (corresponding mask)\n         ├── Val/\n             └── ...\n     ```\n   - The `echo_dnd_dataset.py` script is configured to load data assuming this structure.\n\n---\n\n## 🏃‍♂️ Training \u0026 Inference\n### Training\nTo train the Echo-DND model, run the following command:\n```bash\npython training_echo_dnd.py --data_dir /path/to/your_data_root_dir --batch_size 4 --lr 1e-4 --out_dir ./results/training_run1\n```\n\n### Inference\nTo perform inference on a single image, use the following command:\n```bash\npython inference_echo_dnd.py --image_path /path/to/your/test_image.png --model_path /path/to/your/pretrained_echodnd_model.pt --out_dir ./results/inference_output\n```\n\n## 📄 Citation\n\nIf you find this work useful, please consider citing:\n\n```bibtex\n@article{Rahman2025EchoDND,\n  author    = {Rahman, Abdur and Balraj, Keerthiveena and Ramteke, Manojkumar and Rathore, Anurag Singh},\n  title     = {Echo-DND: a dual noise diffusion model for robust and precise left ventricle segmentation in echocardiography},\n  journal   = {Discover Applied Sciences},\n  volume    = {7},\n  number    = {514},\n  year      = {2025},\n  month     = {May},\n  doi       = {10.1007/s42452-025-07055-5},\n  url       = {https://doi.org/10.1007/s42452-025-07055-5},\n  publisher = {Springer Nature}\n}\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fabdur75648%2Fecho-dnd","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fabdur75648%2Fecho-dnd","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fabdur75648%2Fecho-dnd/lists"}