Deep learning based focus and stigmation correction for electron microscopes (https://doi.org/10.1038/s41467-024-45042-3).
To install the deepfocus package, run the following (requires conda):
git clone https://github.com/StructuralNeurobiologyLab/DeepFocus
cd DeepFocus
conda env create -f environment.yml -n deepfocus
conda activate deepfocus
pip install -e .
After installing the dependencies and the package you can run utils/infer.py to apply the pretrained model to the
example data:
python scripts/infer_examples.py
which results in:
Predicted `working distance` correction 13.03 +- 0.37 (mean +- s.d.) for a known aberration of -13.324 um.
Predicted `stigmator x` correction 0.32 +- 0.09 (mean +- s.d.) for a known aberration of 0.466 (a.u.).
Predicted `stigmator y` correction -0.15 +- 0.04 (mean +- s.d.) for a known aberration of 0.219 (a.u.).
The pretrained model is the baseline model used in the publication with two 512x512 input patches (symmetric perturbation of 5 µm). The example data is part of the training/validation data.
For trainings, we provide example scripts for the baseline model and the EfficientNet ({_efficientnet}.py suffix).
To start a training run the following command:
python scripts/train_deepfocus.py
The default root directory for the training data is data_root='~/DeepFocus/GT/' and for the training results
save_root='~/DeepFocus/trainings/'. Adjust the paths in the script(s) accordingly.
The DeepFocus project was developed at the Max Planck Institute for Biological Intelligence in Martinsried by Philipp Schubert under the supervision of Joergen Kornfeld. We would like to thank Rangoli Saxena for supporting us with MAPFoSt.
@article{Schubert2024,
title = {DeepFocus: fast focus and astigmatism correction for electron microscopy},
volume = {15},
ISSN = {2041-1723},
url = {http://dx.doi.org/10.1038/s41467-024-45042-3},
DOI = {10.1038/s41467-024-45042-3},
number = {1},
journal = {Nature Communications},
publisher = {Springer Science and Business Media LLC},
author = {Schubert, P. J. and Saxena, R. and Kornfeld, J.},
year = {2024},
month = jan
}