Under Development. The corresponding publication will be available soon.
Image processing tools for the segmentation of microglia cells in 3D mouse brain data.
Create a new conda environment.
conda create -n "netoglitazone3D" python=3.7
Activate the environment.
conda activate netoglitazone3D
Install python packages.
pip install -r requirements.txt
Tif stacks (3D image data) of mouse brain hemispheres, obtained with light-sheet microscopy (mesoSPIM), imaged across the sagittal plane.
The directories pre-processing, processing and plotting contain tools to detect microglia cells from 3D mouse hemispheres, perform statistical analysis and generate figures for the corresponding publication.
Tools for stack pre-processing.
- Conversion of image stacks from tif to raw/nrrd file formats.
- Cropping of image stacks to exclude large empty regions.
- Flipping of stacks (horizontal and/or stack flip) to match the Allen Brain Atlas orientation.
Main pipeline for image processing of 3D stacks.
- Cell segmentation: Detection of candidate microglia cells in 3D.
- Alignment: Hemisphere registration to Allen Brain Atlas Reference space.
- Removal of surface artefacts: To remove surface artifacts and perform voxelization of aligned cells.
- Voxelization**: Gaussian smoothing, with diameter 15 pixels.
Tools to perform statistical analysis and generate paper figures.
The pipeline was developed in the laboratories of Prof. Petros Koumoutsakos (Harvard University) and Prof. Adriano Aguzzi (University of Zurich) by
- Athena Economides
- Francesca Catto
- Sergey Litvinov
- Add python packages inside requirements.txt.
- Update voxelization script (inside processing/)
- Clean contents of
plotting.