
by OpenGVLab
SAM-Med2D is an open-source adaptation of the Segment Anything Model for 2D medical image segmentation. It was fine-tuned on a dataset of 4.6M images and 19.7M masks.
Web appFree
Step 1: Clone the repository from GitHub and download the pre-trained weights.
Step 2: Try the online demo on OpenXLab or run the notebook demo locally.
Step 3: Prepare your dataset like the samples in data_demo and generate the required JSON file.
Step 4: Run train.py to fine-tune, or test.py to evaluate with box or point prompts.
Step 5: Export the encoder and decoder to ONNX for deployment with onnxruntime if needed.
The README states the project is released under the Apache 2.0 license.
Yes. The README links an online Gradio demo on OpenXLab. You can also run a notebook or Gradio app locally.
SAM-Med2D is for 2D images. The same group released SAM-Med3D for 3D medical imaging.
TotalSegmentatorTotalSegmentator segments most major anatomical structures in any CT or MR image. It runs from the command line and was created at University Hospital Basel.
FloraceneAI app builder for creating and deploying HIPAA-compliant internal healthcare tools without a dedicated engineering team.