This contains the code for our chromosome segmentation tool 'VaCe'. It is primarily used for finding chromosome centers in DAPI stained cancer cells.
| File | Description |
|---|---|
| config.ini | Configuration file to set parameters |
| environment.yml | File to create the environment |
| Makefile | Makefile to run the script |
...
| Folder | Description |
|---|---|
| sample | Folder containing sample images to use with our project |
| src | Contains our Python scripts |
To get started, run the following code to create the environment. Make sure to activate the environment every time before using the tool.
git clone
cd VaCe
conda env create -f environment.yml
conda activate vace
Input folder will only read .tif files. Images should be of cells stained with DAPI to highlight chromosomes blue.
Takes in cell images, processes with U-Net, and outputs valid chromosome centers for the images.
Set parameters in config.ini under Centers:
image_path : path to folder containing images
- centers/img_name_centers.npy - Raw values of centers from Torch tensor, stored in numpy format
This work implements code from ecSeg and NuSeT, and is inspired by Faster R-CNN.
If you liked this work, refer to our website and paper, linked on the website. https://sophialugo.github.io/Capstone/