Skip to content

mingrui/python-wsi-preprocessing

 
 

Repository files navigation

Python WSI Preprocessing

This project contains a variety of files for investigating image preprocessing using Python with the aim of using deep learning to perform histopathology image classification of whole slide images.

See main tutorial here.

See main project at https://github.com/CODAIT/deep-histopath for more information.

Mingrui's customization

file naming convention is changed to 'filename-filenumber-*'
add patch size adjustment

work in progress

magnification / objective_power settings
to handle different magnification(usually 10x, 20x, 40x), first generate the highest magnification tiles, for example: to generate 20x 512 x 512 patches, first generate 40x 1024 x 1024 patches, then half the patch to 512 x 512.

requirements:

jupyterlab pip install jupyterlab
matplotlib pip install matplotlib
PIL pip install Pillow
openslide pip install openslide-python
py_wsi: pull repo then python setup.py install
scipy: pip install scipy
scikit-image: pip install scikit-image
pandas: pip install pandas

About

Python Whole Slide Image Preprocessing

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 98.2%
  • Python 1.8%