DigiPathAI

A software application built on top of openslide for viewing whole slide images (WSI) and performing pathological analysis

Citation

If you find this reference implementation useful in your research, please consider citing:

@article{khened2020generalized,
  title={A Generalized Deep Learning Framework for Whole-Slide Image Segmentation and Analysis},
  author={Khened, Mahendra and Kori, Avinash and Rajkumar, Haran and Srinivasan, Balaji and Krishnamurthi, Ganapathy},
  journal={arXiv preprint arXiv:2001.00258},
  year={2020}
}

Features

• Responsive WSI image viewer
• State of the art cancer AI pipeline to segment and display the cancerous tissue regions

Application Overview

Results

Installation

Running of the AI pipeline requires a GPU and several deep learning modules. However, you can run just the UI as well.

Just the UI

Requirements

• openslide
• flask

The following command will install only the dependencies listed above.

pip install DigiPathAI

Entire AI pipeline

Requirements

• pytorch
• torchvision
• opencv-python
• imgaug
• matplotlib
• scikit-learn
• scikit-image
• tensorflow-gpu >=1.14,<2
• pydensecrf
• pandas
• wget

The following command will install the dependencies mentioned

pip install "DigiPathAI[gpu]"

Both installation methods install the same package, just different dependencies. Even if you had installed using the earlier command, you can install the rest of the dependencies manually.

Usage

Local server

Traverse to the directory containing the openslide images and run the following command.

digipathai <host: localhost (default)> <port: 8080 (default)>

Python API usage

The application also has an API which can be used within python to perform the segmentation.

from DigiPathAI.Segmentation import getSegmentation

prediction = getSegmentation(img_path, 
			patch_size  = 256, 
			stride_size = 128,
			batch_size  = 32,
			quick       = True,
			tta_list    = None,
			crf         = False,
			save_path   = None,
			status      = None)

Contact

• Avinash Kori ([email protected])
• Haran Rajkumar ([email protected])