Skip to content
Software for the paper "Fast and robust active neuron segmentation in two-photon calcium imaging using spatio-temporal deep learning," Proceedings of the National Academy of Sciences (PNAS), 2019.
Python MATLAB
Branch: master
Clone or download

README.md

STNeuroNet

STNeuroNet is 3-dimensional convolutional neural network (CNN) for segmenting "active" neurons from calcium imaging data. The network was implemented through NiftyNet, a TensorFlow-based open-source CNN platform. You can adapt the existing network to your imaging data.

Features

  • Pre- and post-processing steps for segmenting active neurons
  • A 3D CNN for batch-processing of calcium imaging data
  • MATLAB GUI for manual marking of calcium imaging data

System Requirements

  • Anaconda with Python 3.5
  • MATLAB 2017b and MATLAB Runtime version 9.3
    • Neural Network Toolbox, Image Processing Toolbox, and the GUI Layout Toolbox
    • MATLAB Runtime can be acquired from here
  • Tensorflow-gpu 1.4 (CUDA Toolkit 8.0 and cuDNN v7.0 required. Detailed instructions can be found here.)
  • NiftyNet version 0.2.0.post1

Documentation

The how-to guides are available on the Wiki.

Useful links

NiftyNet source code on GitHub

Link to Datasets:

Allen Brain Observatory dataset

Neurofinder Challenge website

Citing

If you use any part of this software in your work, please cite Soltanian-Zadeh et al. 2019:

  • S. Soltanian-Zadeh, K. Sahingur, S. Blau, Y. Gong, and S. Farsiu, "Fast and robust active neuron segmentation in two-photon calcium imaging using spatio-temporal deep-learning," Proceedings of the National Academy of Sciences (PNAS), 116(17), pp. 8554-8563, April 2019. DOI: 10.1073/pnas.1812995116

If you use NiftyNet in your work, please cite Gibson and Li, et al. 2018:

Licensing and Copyright

STNeuroNet is released under the GNU License, Version 2.0.

Acknowledgements

We thank David Feng and Jerome Lecoq from the Allen Institute for providing the ABO data, Saskia de Vries and David Feng from the Allen Institute for useful discussions, Hao Zhao for the initial implementation of the GUI, and Leon Kwark for the manual marking of the data.

You can’t perform that action at this time.