Convolutional capsule network for COVID-19 detection using radiography images.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Deep Learning

Background:

• The COVID-19 pandemic necessitates rapid and accurate diagnostic tools.
• Existing deep learning models like Convolutional Neural Networks (CNNs) for chest X-ray analysis face limitations such as view-invariance and information loss.
• There is a critical need for advanced automated systems to aid in the diagnosis of COVID-19.

Purpose of the Study:

• To propose a novel deep learning-based decision support system for COVID-19 diagnosis using chest X-ray images.
• To address the limitations of CNNs by employing Capsule Networks (CapsNet).
• To evaluate the performance of the proposed Visual Geometry Group Capsule Network (VGG-CapsNet) model.

Main Methods:

• Development of a Visual Geometry Group Capsule Network (VGG-CapsNet) model.
• Utilizing chest radiography (CXR) images for training and validation.
• Comparative analysis against CNN-based models and CNN-CapsNet.

Main Results:

• The VGG-CapsNet model demonstrated superior performance compared to CNN-CapsNet.
• Achieved 97% accuracy in classifying COVID-19 versus non-COVID-19 cases.
• Attained 92% accuracy in classifying COVID-19 versus normal versus viral pneumonia.

Conclusions:

• The proposed VGG-CapsNet system effectively overcomes the limitations of traditional CNNs for COVID-19 detection.
• This deep learning approach provides a reliable and accurate automated solution for diagnosing COVID-19 from chest X-rays.
• The VGG-CapsNet system holds significant potential as a decision support tool in clinical settings.