Introduction

Demonstration to run the AutoCUTS-LM-Analysis code with pictures. The full setup with scripts and some image examples are available at .

When using the code or data you can cite: Larsen NY, Li X, Tan X, Ji G, Lin J, Rajkowska G, Møller J, Vihrs N, Sporring J, Fei S, Nyengaard JR. "Cellular 3D-reconstruction and analysis in the human cerebral cortex using automatic serial sections". Commun Biol 4, 1030, Springer Nature,10.1038/s42003-021-02548-6, (2021). https://www.nature.com/articles/s42003-021-02548-6

Requirements

Before starting, you will need the following:

• Install MATLAB.
• Install Image processing toolbox and Computer vision toolbox before running the code.
• Install R.

Code Guideline

There are five folders in this repository.

1. BiopsyExtraction contains a userfriendly script to random sample biopsies from a block of tissue.

2. Section_segmentation script was used to detect focused sections and export them as individual files. The script was appled on a test image.

3. AutoCUTSAnalysis includes scripts implemented on a test collection of 50 images(will be available after publication of article) that involve the following steps:

• Align sections
• Crop aligned stack where only tissue appears
• Using the UNetDense architecture to segment cells, the region of interest could then be defined based on a density map of centroids.
• Filtering small cells away with k-means and reconstruct neurons in 3D.

4. ValidationDeeplearning will estimate the pixelwise and objectwise validation of the UNetDense model output.

• Pictures of objectwise validation will be attach after publication.
• Look at False-Negative plot in the article for further information.

5. CylindricalKfunction Demonstration to run the cylindrical K-function code with pictures. The function was introduced in 'The cylindrical K-function and Poisson line cluster point processe' by Jesper Møller et al. All credits go to the script writer, Ninna Vihrs.

• The code analyses the columnar structure in 3D point patterns for one test subject.
• Pictures will be attach after publication.

Contact

If you have any questions or suggestions, you can reach Nick via e-mail at nick.yin.larsen@gmail.com for the MATLAB scripts or Ninna via e-mail at nvihrs@math.aau.dk for the R-script.

Citation

If using this dataset, please cite:

@Article{NYLCOM,
  author    = {NY Larsen and X Li and X Tan and G Ji and J Lin and G Rajkowska and J Møller and N Vihrs and J Sporring and F Sun and JR Nyengaard},
  journal   = {Communication Biology (in press)},
  title     = {Cellular 3D-reconstruction and analysis in the human cerebral cortex using automatic serial sections},
  year      = {2021},
  number    = {4},
  volume    = {1030},
  doi       = {10.1038/s42003-021-02548-6},
  publisher = {Springer Nature},
}

Author: Nick Yin Larsen