Deep Learning-Driven Automated High-Content dSTORM Imaging with a scalable Open Source Toolkit
Supplemental Software and test data sets for fully automated dSTORM acquistion using deep learning. The correspondig paper has been published .... . A. Python software for training and application of DeepLabV3 for segmentation of microscopy images (train_apply_NN.zip). Jupyter notebooks for training and applying deep neural netoworks (including explanations for input paramters) Training and Test data for nucleus segmentation B. Python software for fully automated acquisition of dSTORM images using networks trained with software A (automated_STORM_v1_0_0.zip). Jupyter notebook for acquistion of high-content images and fully automated dSTORM, w
Resource information
- Format
- zip
- Original repository
- 14690875
- DOI
- 10.5281/zenodo.14690875
- Credit basis
- Creators in Zenodo record metadata
- Access
- Open files available from the original Zenodo deposit
- Last checked
- Nov 5, 2026
- License
- cc-by-4.0
- Last record update
- Nov 5, 2026
About this resource
Supplemental Software and test data sets for fully automated dSTORM acquistion using deep learning. The correspondig paper has been published .... . A. Python software for training and application of DeepLabV3 for segmentation of microscopy images (train_apply_NN.zip). Jupyter notebooks for training and applying deep neural netoworks (including explanations for input paramters) Training and Test data for nucleus segmentation B. Python software for fully automated acquisition of dSTORM images using networks trained with software A (automated_STORM_v1_0_0.zip). Jupyter notebook for acquistion of high-content images and fully automated dSTORM, with explanations of the input parameters Jupyter notebook with a graphical user interface Configureation file used at our setup for µManager C. Python software for fully automated acquisition of dSTORM images using networks trained with software A (automated_STORM_easy_modules.zip). Jupyter notebook for acquistion of high-content images and fully automated dSTORM, using smaller, easier, and better separated modules for an introduction. Configureation file used at our setup for µManager D. Python software and FIJI macro for batch processing of STORM-films and FRC-resolution estimation (dSTORM_reconstruction_FRC.zip). Jupyter notebook for batch processing with the storm-analysis library and FRC calculation on these datasets FJI-macro for batch processing using ThunderSTORM E. Python software to retrieve periodicity from neural axon via pairwise distances. Jupyter notbooks with examples of simulatted and real axons Simulated and real image data as test data F. Installation (setup.zip) It is recommended to setup a new virtual environment with the required libraries. 1. To create a new environment with the required libraries (except stormanalysis and pytorch) run the following command: conda env create -f environment_automatedSTORM.yml nto an existing environment using conda and pip running the following two commands (run the conda command first and replace "requirements_conda.txt" by the acutal file location of the file): conda install --yes --file requirements_conda.txt pip install -r requirements_pip.txt 3. lnstall Pytorch and stormanalysis Pytorch can be retrieved following the instructions on https://pytorch.org ; the torchvison version should be 1.15. or higher Install stormanalysis (only needed if you want to use it for dSTORM reconstruction and resolution estimates). The package can be installed following the instructions on https://github.com/ZhuangLab/storm-analysis