Atlases are crucial to imaging statistics as they enable the standardization of inter-subject and inter-population analyses. While existing atlas estimation methods based on fluid/elastic/diffusion registration yield high-quality results for the human brain, these deformation models do not extend to a variety of other challenging areas of neuroscience such as the anatomy of C. elegans worms and fruit flies. To this end, this work presents a general probabilistic deep network-based framework for atlas estimation and registration which can flexibly incorporate various deformation models and levels of keypoint supervision that can be applied to a wide class of model organisms.
See our paper for further details:
@inproceedings{nejatbakhsh2023learning,
title={Learning Probabilistic Piecewise Rigid Atlases of Model Organisms via Generative Deep Networks},
author={Nejatbakhsh, Amin and Dey, Neel and Venkatachalam, Vivek and Yemini, Eviatar and Paninski, Liam and Varol, Erdem},
booktitle={International Conference on Information Processing in Medical Imaging},
pages={332--343},
year={2023},
organization={Springer}
}
Note: This research code remains a work-in-progress to some extent. It could use more documentation and examples. Please use at your own risk and reach out to us (anejatbakhsh@flatironinstitute.org) if you have questions. If you are using this code package, please cite our paper.
- Download and install anaconda
- Create a virtual environment using anaconda and activate it
conda create -n datlas python=3.8
conda activate datlas
-
Install Pytorch package
-
Install other requirements (pyro, matplotlib, scipy, sklearn, cv2, dipy, ray)
-
Run either using the demo file or the run script via the following commands
python run.py -c configs/test.yaml -o ../results/
Since the code is preliminary, you will be able to use git pull to get updates as we release them.