CCTCOVID: COVID-19 detection from chest X-ray images using Compact Convolutional Transformers.

Area of Science:

• Medical Imaging Analysis
• Artificial Intelligence in Healthcare
• Infectious Disease Diagnostics

Background:

• The rapid person-to-person transmissibility of COVID-19 necessitates urgent and accurate diagnostic methods.
• Screening medical images like X-rays is crucial for COVID-19 detection but faces challenges due to high case numbers and complex procedures.
• Deep learning shows significant promise for automating complex medical tasks, including image analysis.

Purpose of the Study:

• To develop and evaluate a novel transformer-based deep learning method for automated COVID-19 detection.
• To utilize Compact Convolutional Transformers (CCT) for analyzing X-ray images for COVID-19 diagnosis.
• To improve the efficiency and accuracy of COVID-19 screening in clinical settings.

Main Methods:

• Implementation of a transformer-based deep learning model utilizing Compact Convolutional Transformers (CCT).
• Training and validation of the CCT model on a dataset of X-ray images for COVID-19 classification.
• Comparative analysis against existing methods to demonstrate performance efficacy.

Main Results:

• The proposed CCT-based method achieved a high accuracy of 99.22% in detecting COVID-19 from X-ray images.
• The model demonstrated superior performance compared to previous state-of-the-art approaches.
• The results highlight the effectiveness of transformer architectures in medical image analysis for infectious diseases.

Conclusions:

• The developed transformer-based deep learning method offers a highly accurate and efficient solution for COVID-19 detection using X-ray images.
• Compact Convolutional Transformers show significant potential for advancing automated diagnostic tools in radiology.
• This approach can help alleviate the burden on medical practitioners and improve community-level screening capabilities.