CovFrameNet: An Enhanced Deep Learning Framework for COVID-19 Detection.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• The COVID-19 pandemic necessitates rapid diagnostic tools.
• Deep learning shows promise for medical image analysis.
• Existing methods require computational intelligence for efficient disease detection.

Purpose of the Study:

• To develop and evaluate a deep learning framework, CovFrameNet, for COVID-19 detection using chest X-rays.
• To integrate advanced image pre-processing techniques with a convolutional neural network (CNN) architecture.
• To characterize and detect novel coronavirus infection through computational methods.

Main Methods:

• A novel framework, CovFrameNet, combining image pre-processing and a CNN model was proposed.
• The CNN architecture featured an enhanced image pre-processing mechanism.
• The model was trained and validated on the NIH Chest X-Ray dataset and the COVID-19 Radiography database.

Main Results:

• The proposed deep learning model achieved high performance metrics.
• Specific results include an accuracy of 0.1, recall/precision of 0.85, F-measure of 0.9, and specificity of 1.0.
• The model demonstrated effectiveness in characterizing and detecting COVID-19 infection.

Conclusions:

• CNN-based methods with image pre-processing are effective for COVID-19 pre-screening.
• The proposed framework can aid in the confirmation of RT-PCR-detected COVID-19 cases.
• This approach offers a valuable computational intelligence solution for rapid disease detection.