nnunet-segmentation

What problem does it solve?

This Skill automates the complex and time-consuming process of segmenting anatomical structures or lesions within medical images, eliminating the need for manual annotation or extensive hyperparameter tuning.

Core Features & Use Cases

• Self-Configuring Framework: nnU-Net automatically selects the optimal network architecture, preprocessing, and training strategy for any imaging modality (CT, MRI, microscopy, etc.).
• State-of-the-Art Performance: Achieves top results on diverse segmentation tasks with minimal user input.
• Use Case: Automatically segmenting tumors in MRI scans for radiotherapy planning, or identifying organs in CT scans for quantitative analysis, using a dataset of 50 annotated patient scans.

Quick Start

Train a 3D full-resolution segmentation model for dataset 7 using the command nnUNetv2_train 7 3d_fullres 0 --npz.