COVID-19 classification using deep feature concatenation technique.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computer Science

Background:

• COVID-19 detection from medical images is a critical global health challenge.
• Chest X-ray and CT scans are primary diagnostic tools.
• Automated detection methods are needed to assist medical experts.

Purpose of the Study:

• To develop an automated deep learning system for COVID-19 detection.
• To improve the accuracy and speed of COVID-19 diagnosis using medical images.
• To evaluate a novel deep feature concatenation (DFC) mechanism.

Main Methods:

• Utilized a proposed Convolutional Neural Network (CNN) architecture with three deep layers.
• Implemented DFC to combine features from X-ray and CT scans, and with pre-trained CNNs (ResNet, GoogleNet).
• Optimized CNN performance using various algorithms, epochs, learning rates, and mini-batch sizes.

Main Results:

• The DFC mechanism proved effective in forming a definitive classification descriptor.
• The proposed method demonstrated superior performance compared to state-of-the-art techniques.
• Achieved high accuracy, precision, recall, and F-score in COVID-19 detection.

Conclusions:

• The developed deep learning approach with DFC offers a promising solution for rapid and accurate COVID-19 detection.
• The system can reduce the reliance on expert radiologists for initial screening.
• This method holds potential for widespread clinical application in pandemic scenarios.