Nuclei Grading of Clear Cell Renal Cell Carcinoma in Histopathological Image by Composite High-Resolution Network
A Composite High-Resolution Network for ccRCC nuclei grading.
The network has two parts:
- Propose a segmentation network called W-Net that can separate the clustered nuclei.
- Recast the fine-grained classification of nuclei to two cross-category classification tasks, based on two high-resolution feature extractors (HRFEs) which are proposed for learning these two tasks.
The two HRFEs share the same backbone encoder with W-Net by a composite connection so that meaningful features for the segmentation task can be inherited for the classification task. Finally, a head-fusion block is applied to generate the predicted label of each nucleus.
Link to MICCAI 2021 paper.
conda create --name hovernet python=3.6
conda activate hovernet
pip install -r requirements.txt
Download the ccRCC grading dataset as used in our paper from this link.
Ground truth files are in .mat format, refer to the website of this dataset for further information.
If any part of this code is used, please give appropriate citation to our paper.
Install the required libraries before using this code. Please refer to requirements.txt
- Zeyu Gao (betpotti@gmail.com)
- Jiangbo Shi (shijiangbo@stu.xjtu.edu.cn)
- Chen Li (cli@xjtu.edu.cn)
- Haichuan Zhang (haichuan@psu.edu)
BioMedical Semantic Understanding Group, Xi'an Jiaotong University
This project is licensed under the MIT License - see the LICENSE file for details
We have great thanks to the implementation of nuclei segmentation and classification framework HoVerNet. This code is modified from the tensorflow version of HoVerNet.
The datasets used are in whole or part based upon data generated by the TCGA Research Network.