A Hybrid Convolutional Neural Network Model for Diagnosis of COVID-19 Using Chest X-ray Images

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Infectious Disease Diagnostics

Background:

• The COVID-19 pandemic caused widespread disruption, necessitating rapid diagnostic methods.
• Reverse Transcription Polymerase Chain Reaction (RT-PCR) testing is time-consuming and faces global shortages.
• Accurate and timely detection of COVID-19 is crucial for patient management and disease control.

Purpose of the Study:

• To propose a novel image processing technique, C19D-Net, for detecting COVID-19 from chest X-ray images.
• To enhance the accuracy of COVID-19 detection for radiologists.
• To provide a faster diagnostic alternative in resource-limited settings.

Main Methods:

• Utilized deep learning (DL) features extracted via the InceptionV4 architecture.
• Employed a Multiclass Support Vector Machine (SVM) classifier for infection detection.
• Pre-processed a dataset of 1900 chest X-ray images from public databases.

Main Results:

• C19D-Net achieved high detection accuracies: 96.24% for 4-class, 95.51% for 3-class, and 98.1% for 2-class classification.
• The model demonstrated superior performance in precision, accuracy, F1-score, and recall compared to existing methods.
• The system effectively identified COVID-19 infection from chest X-ray images.

Conclusions:

• The C19D-Net model offers a highly accurate and efficient method for COVID-19 detection using chest X-rays.
• This AI-driven approach can supplement RT-PCR, particularly in areas with limited testing resources.
• C19D-Net has the potential to significantly aid radiologists in improving diagnostic accuracy for COVID-19.