Code for recreating results in the publication: Single cell metabolic imaging of tumor and immune cells in vivo in melanoma bearing mice. Features classical image processing based techniques for single cell segmentation. Specifically, this involves multiple thresholding methods for foreground/background differentiation, edge detection filters for border creation, and combination of methods through ensemble voting.
Dependencies
- scipy (1.6.2)
- numpy (1.20.1)
- pandas (1.2.4)
- opencv-python (4.5.5.64)
- tifffile (2021.4.8)
- matplotlib (3.3.4)
- ipywidgets (7.6.3)
System Specifications
- 2.50GHz CPU
- 16.0 GB RAM
- 64-bit operating system
- Windows 10
- No GPU dependencies required
- Clone the GitHub Repository
- Create an Anaconda environment.
- Run through the
Immune_Cell_Segmentation_Notebook.ipynb
. If this is your first time running the notebook, ensure that you run the first cell for installing all files fromrequirements.txt
base_path
is currently pointing to the demo data included above. To analyze your own data change thebase_path
to the location of your files.- Continue running through cells to generate file lists and read in all requisite files for recreating results.
- Update parameters in
adpative segmentation
and proceed with data generation. Voila! - Final spleen segmentation data was generated with the following parameters:
- Save Masks: True
- Minimum ROI: 40
- Include Lifetime: True
- Lifetime Minimum: 800
- Lifetime Maximum: 1500
- Final tumor segmentation data was generated with the following parameters:
- Save Masks: True
- Minimum ROI: 60
- Include Lifetime: True
- Lifetime Minimum: 1000
- Lifetime Maximum: 1500
author = {Alexa R. Heaton, Peter R. Rehani, Anna Hoefges, Angelica F. Lopez, Amy K. Erbe, Paul M. Sondel, and Melissa C. Skala},
title = {Single cell metabolic imaging of tumor and immune cells in vivo in melanoma bearing mice},
month = march,
year = 2023,
publisher = {Zenodo},
version = {v1.0},
doi = {10.5281/zenodo.7696762},
url = {https://github.com/skalalab/heaton_a-automated_immune_cell_segmentation}