Skip to content

Extending our COVID-19 solution to Common Pneumonia cases

Notifications You must be signed in to change notification settings

IDU-CVLab/Pneumonia-Detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

47 Commits
 
 
 
 

Repository files navigation

Dataset:

  • COV19-CT Database was used for COVID and healthy cases. The first 51 CT scans in each class were used for training, and full validation set was used for validation.
  • For Pneumonia cases, a publicly available dataset at https://data.mendeley.com/datasets/3y55vgckg6/1 was used; 4 CT scans (test-59, 61, 62, and 63) were used for training, and 2 CT scans (test-64 and 65) were used for validation.
  • The dataset used in this method is a combination of the above-mentioned.

Methodology:

  • Augmentation were deployed on the Pneumonia class aiming at classes balance. Augmentation focused on zooming, flipping, and rotation; 2000 images from the existing 199 slices of Common Pneumonia were generated. using 'augmentator' at https://github.com/mdbloice/Augmentor.
  • Image processing the CNN model training used for Pnumonia cases is similar to the onle we applied during the first run of the MIA-COVID19 Workshop, 2021 here. Thus this work allow for extending our solution from just COVID-19 detection to Pnumonia cases.
  • For full details of the method and the results, refer to our paper imentioned below.
  • Kindly inform the organization owner if you wish to obtain the pretrained model in this study

Citation

This study is included as a full-text in the university's 2023 booklets, following the 5th International Medical Congress of Izmir Democracy University (IMCIDU) with ID-442/OP.

Releases

No releases published

Packages

No packages published

Languages