An emerging network for COVID-19 CT-scan classification using an ensemble deep transfer learning model.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• The COVID-19 pandemic has caused millions of deaths globally, necessitating rapid and accurate diagnostic tools.
• Early detection of COVID-19 is critical for controlling viral spread and initiating prompt patient treatment.
• Computed Tomography (CT) imaging is a valuable resource for diagnosing COVID-19.

Purpose of the Study:

• To develop and evaluate artificial intelligence (AI) based models for the diagnosis of COVID-19 using CT scans.
• To compare the performance of a novel CNN model (CT6-CNN) against ensemble deep transfer learning models.

Main Methods:

• A Convolutional Neural Network (CNN) model named CT6-CNN was designed.
• Two ensemble deep transfer learning models were developed, integrating Xception, ResNet-101, DenseNet-169, and CT6-CNN.
• The models were trained and validated on the SARS-CoV-2 CT dataset comprising 2481 CT scans.

Main Results:

• The CT6-CNN model achieved an accuracy of 94.66%, precision of 94.67%, sensitivity of 94.67%, and F1-score of 94.65%.
• The ensemble deep learning models demonstrated superior performance, reaching an accuracy of 99.2%.
• Experimental results confirm the high effectiveness of the developed AI models, particularly the ensemble approaches.

Conclusions:

• AI-driven diagnostic tools, especially ensemble deep learning models, show significant promise for accurate and efficient COVID-19 detection from CT images.
• The developed models can aid clinicians in making timely diagnoses, contributing to better pandemic control and patient outcomes.