🐭 Mousetumorpy

A toolbox to segment and track murine lung tumor nodules in mice CT scans.

Hightlights

• Lungs segmentation and cropping: Run a pretrained YoloV8 model to segment the lungs and crop the CT scans around them.
• Tumor segmentation: Run a pretrained nnUNet model to segment tumor nodules in 3D.
• Tumor tracking: Track tumor nodules longitudinally across CT scans.

Installation

Install the mousetumorpy package with pip:

pip install git+https://github.com/EPFL-Center-for-Imaging/mousetumorpy.git

or clone the project and install the development version:

git clone https://github.com/EPFL-Center-for-Imaging/mousetumorpy.git
cd mousetumorpy
pip install -e .

Usage from Napari

To use Mousetumorpy's functions interactively in Napari, install the napari-mousetumorpy package.

Usage as a CLI

The command-line interface (CLI) provides several image processing functions.

Crop

Run a YoloV8 model to segment the lungs cavity and crop the image around the lungs.

mousetumorpy crop <image_file> <out_dir>

For more details, see mousetumorpy crop --help.

Predict

Run a nnUNet model to segment tumor nodules.

mousetumorpy predict <image_file> <out_dir>

For more details, see mousetumorpy predict --help.

Combine

Combine several 3D images (ZYX) into a single 4D image (TZYX).

mousetumorpy combine <image_1> <image_2> <image_3> <out_dir>

For more details, see mousetumorpy combine --help.

Track

Track tumor nodules across a 4D mask (TZYX) time series using trackpy or laptrack.

mousetumorpy track <labels_file> <image_file> <lungs_file> <out_dir>

For more details, see mousetumorpy track --help.

License

This project is licensed under the AGPL-3 license.

This project depends on the ultralytics package which is licensed under AGPL-3.