This is a flask web app for an AI pathology application -- Deep learning Analysis of kidney frozen sections.
For one whole slide image of kidney frozen section, you can:
- Choose which histology primitives to segment by pre-trained deep learning models
- Have an overview statistic report for this kidney biopsy slide
- Click on one object detected and look at its detailed pathomic features
Flask - web framework
fabric.js - canvas
PyTorch - models
pyradiomics - pathomic features
(Nginx - cloud)
(Gunicorn - cloud)
conda create -n [name of enviroment] python=3.7
pip install -r requirements.txt
cd flask-pytorch-cloud
conda activate [name of enviroment]
python app.py
Then open the url and upload the sample image im0.png to test the model.