iducvlab / cov19d Goto Github PK

• The dataset is called COV19-CT-DB obtained at AI-enabled Medical Image Analysis Workshop and Covid-19 Diagnosis Competition (ICCV 2021 Workshop: MIA-COV19D 2021). The project is conducted using chest CT-scan series of images aiming to develop an automated solution for Covid-19/non-Covid-19 diagnosis.
  
• The team (IDU-CVLab) is on the leaderboard here.
  

To replicate the codes, the following must be noted:

1. To run the code properly you would need a training set of images and a validation set of images.
2. The images must be put in the appropriate directories. With that, the directory of training and validation images included in the code should be changed to match the directory where your image datasets are located. This method is following the documentation for ‘imagedatagenerator’ and ‘flow_from_directory’ at https://www.tensorflow.org/api_docs/python/tf/keras/preprocessing/image/ImageDataGenerator
   

The algorithm was developed in two versions or steps as follows:

The Code 'COV19-CT-DB-CNN-Model.py'
The algorithm introduces a less hand-engineered CNN model Architecture for automated COVID-19 diagnosis.
The CNN model achitechture is:

The Code 'Static_Cropping_Deeplearning_Model_for_CT_images.py'
The work makes use of the above mentioned CNN model with images preprocessed before training. The preprocessing includes a static rectangular croping to the Region of Interest (ROI) in the CT sclices and statitical methods for uppermost and lowermost removal of the slcies in each CT scan image.
The code can be devided in two parts:

PartI. The code without slice processing and parameters tuning [as in version1].

PartII. The code with rectangle-shape cropping and hyperparameters tuning.
● Further dependencies used for this part are:
▪ CV2 == 4.5.4
▪ sklearn == 0.24.2

● If you find the method useful this method, please considering citing:
@article{Morani2021DeepLB,
title={Deep Learning Based Automated COVID-19 Classification from Computed Tomography Images},
author={Kenan Morani and Devrim {"U}nay},
journal={ArXiv},
year={2021},
volume={abs/2111.11191}
}

● Literature review, methedology, results and more can be found here. (~34min)