PathoSpotter is a computational tool built to help pathologists. This repository intends to improve the already operational, yet not optimal, tool entitled Classifier. The objective when concluding this project is to have a fast and secure web page where physicians can upload their digital images of kidney biopsies and obtain a diagnosis from our trained and published neural networks.
Imagine having a physician colleague, always available to give you his guesses and advices when it comes to diagnosis. This is what this tool is simulating. Our neural networks were trained to see the amyloidosis, hypercellularity, and sclerosis lesions in glomeruli. This tool by no means replaces the need of a trained physician, but intends to be a useful tool in the nephropathology context.
To get a local copy up and running follow these simple steps.
This tutorial assumes you already have a valid Python instalation. If not, please check the websites above for tutorials.
-
Clone the repo
git clone https://github.com/Karnezis/PS_Classifier.git
-
Create a virtual environment
python -m venv env
-
Activate your environment
env\\Scripts\\activate.bat
-
This project is built on top of some libraries. Install them with:
pip install -r requirements.txt
-
Now, run our project.
python .\\src\\main.py
-
Access the website in the link specified.
First, upload an image to the system. Then, click the button to submit it. Now wait for the results. Ta-da!
If you want to know more about this project, contact the authors to read the unpublished bachelor thesis in which this project is tested in latency, security, and maintainability.
Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.
- Fork the Project
- Create your Feature Branch (
git checkout -b feature/AmazingFeature) - Commit your Changes (
git commit -m 'Add some AmazingFeature') - Push to the Branch (
git push origin feature/AmazingFeature) - Open a Pull Request
Ellen C. Aguiar - chalegreaguiar@gmail.com
Ângelo A. Duarte - angeloduarte@uefs.br
Project Link: here.