Authors: Deng-Ping Fan, Tao Zhou, Ge-Peng Ji, Yi Zhou, Geng Chen, Huazhu Fu, Jianbing Shen, and Ling Shao.
-
This repository provides code for "Inf-Net: Automatic COVID-19 Lung Infection Segmentation from CT Scans" submit to TMI-2020. (arXiv Pre-print & medrXiv)
-
Any quesetions please contact to Deng-Ping Fan via E-mail.
Figure 1. Example of COVID-19 infected regions in CT axial slice, where the red and green masks denote the
ground-glass opacity (GGO) and consolidation, respectively. The images are collected from [1].
[1] COVID-19 CT segmentation dataset, link: https://medicalsegmentation.com/covid19/, accessed: 2020-04-11.
Our proposed methods consist of three individual components under different settings:
-
Inf-Net (Supervised Learning with segmentation and edge supervision).
-
Semi-Inf-Net (Semi-supervised learning with doctor label and pseudo label)
-
Semi-Inf-Net + FCN8s/UNet (Extension to Multi-class Segmentation, including Background, Ground-glass opacities, and Consolidation).
Figure 2. The architecture of our proposed Inf-Net model, which consists of three reverse attention
(RA) modules connected to the paralleled partial decoder (PPD).
-
Pre-trained Model
Coming soon ...
Figure 3. Overview of the proposed Semi-supervised Inf-Net framework.
-
Pre-trained Model
Coming soon ...
-
Pre-trained Model
Coming soon ...
We provide one-key evaluation toolbox for LungInfection Segmentation tasks, including Lung-Infection and Multi-Class-Infection.
Google Drive: https://drive.google.com/open?id=1BGUUmrRPOWPxdxnawFnG9TVZd8rwLqCF
-
Prerequisites: MATLAB Software (Windows/Linux OS is both work, however we suggest you test it in the Linux OS for convenience.)
-
run
cd ./Evaluation/
andmatlab
open the matlab software via terminal -
Just run
main.m
to get the overall evaluation results. -
Edit the parameters in the
main.m
to evaluate your custom methods. Please refer to the instructions in themain.m
We also build a semi-supervised COVID-19 infection segmentation (COVID-SemiSeg) dataset, with 100 labeled CT scans from the COVID-19 CT Segmentation dataset [1] and 1600 unlabeled images from the COVID-19 CT Collection dataset [2]. Our COVID-SemiSeg Dataset can be downloaded at Google Drive
[1]“COVID-19 CT segmentation dataset,” https://medicalsegmentation.com/covid19/, accessed: 2020-04-11. [2]J. P. Cohen, P. Morrison, and L. Dao, “COVID-19 image data collection,” arXiv, 2020.
-
Lung infection which consists of 50 labels by doctors (Doctor-label) and 1600 pesudo labels generated (Pesudo-label) by our Semi-Inf-Net model.
Download: http://dpfan.net/wp-content/uploads/LungInfection-Train.zip -
Multi-Class lung infection which also composed of 50 multi-class labels (GT) by doctors and 50 lung infection labels (Prior) generated by our Semi-Inf-Net model.
Download: http://dpfan.net/wp-content/uploads/MultiClassInfection-Train.zip
-
The Lung infection segmentation set contains 48 images associate with 48 GT. Download: http://dpfan.net/wp-content/uploads/LungInfection-Test.zip
-
The Multi-Class lung infection segmentation set has 48 images and 48 GT. Download: http://dpfan.net/wp-content/uploads/MultiClassInfection-Test.zip
To compare the infection regions segmentation performance, we consider the two state-of-the-art models U-Net and U-Net++. We also show the multi-class infection labeling results in Fig. 5. As can be observed, our model, Semi-Inf-Net & FCN8s, consistently performs the best among all methods. It is worth noting that both GGO and consolidation infections are accurately segmented by Semi-Inf-Net & FCN8s, which further demonstrates the advantage of our model. In contrast, the baseline methods, DeepLabV3+ with different strides and FCNs, all obtain unsatisfactory results, where neither GGO and consolidation infections can be accurately segmented.
Overall results: http://dpfan.net/wp-content/uploads/COVID-SemiSeg-Results.zip
Lung infection segmentation results: http://dpfan.net/wp-content/uploads/Lung-infection-segmentation.zip
Multi-class lung infection segmentation: http://dpfan.net/wp-content/uploads/Multi-class-lung-infection-segmentation.zip
Figure 4. Visual comparison of lung infection segmentation results.
Figure 5. Visual comparison of multi-class lung infection segmentation results, where the red and green labels
indicate the GGO and consolidation, respectively.
https://github.com/HzFu/COVID19_imaging_AI_paper_list
http://dpfan.net/wp-content/uploads/2020TMISubmissionInfNet.pdf
Please cite our paper if you find the work useful:
@article{fan2020InfNet,
title={Inf-Net: Automatic COVID-19 Lung Infection Segmentation from CT Scans},
author={Fan, Deng-Ping and Zhou, Tao and Ji, Ge-Peng and Zhou, Yi and Chen, Geng and Fu, Huazhu and Shen, Jianbing and Shao, Ling},
Journal = {arXiv},
year={2020}
}
We thanks xxx for their contributions.