As one of the most fundamental tasks in computer-assisted clinical decision support, medical image segmentation forms the basis for numerous downstream applications ranging from diagnosis, therapeutic planning, to post-treatment monitoring and outcome evaluation. To accurately and efficiently segment anatomical structures in medical images, an AI-driven promptable segmentation plugin is introduced in this project. It integrates a state-of-the-art framework for 3D promptable segmentation: nnInteractive [1] and a web-based user interface, supporting positive and negative point prompts across multiple imaging modalities (CT, MR, PET, etc.). This model has been trained on 120+ diverse volumetric 3D datasets (CT, MRI, PET, 3D Microscopy, etc.), enabling full 3D segmentation from intuitive 2D interactions with accuracy, adaptability, and usability.
[1] Isensee, F., Rokuss, M., Krämer, L.*, Dinkelacker, S., Ravindran, A., Stritzke, F., Hamm, B., Wald, T., Langenberg, M., Ulrich, C., Deissler, J., Floca, R., & Maier-Hein, K. (2025). nnInteractive: Redefining 3D Promptable Segmentation. https://arxiv.org/abs/2503.08373 *: equal contribution
- Start by launching the Medical Image Annotation plugin.
2. Click a point as a prompt within the region of intersted.
3. The mask will show in the medical image after around 30 seconds without GPU or 5 seconds with GPU.
