This project is developed to segment various biomedical imaging applications such as segmentation of brain tumors from MR scans, skin lesions from dermoscopy images, and retinal blood vessels from fundus images.
An end-to-end deep learning network called CMM-Net is designed for medical image segmentation. The proposed method is validated with three different medical imaging tasks and performed state-of-the-art performance.
We make the source code publicly available for researchers for validation and further improvement. The code includes the following files: main, train, predict, and model's network.
This work has been published in Scientific Reports
Cite this article: Al-masni, M.A., Kim, DH. CMM-Net: Contextual multi-scale multi-level network for efficient biomedical image segmentation. Sci Rep 11, 10191 (2021). https://doi.org/10.1038/s41598-021-89686-3