A Rapid Artificial Intelligence-Based Computer-Aided Diagnosis System for COVID-19 Classification from CT Images.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• High COVID-19 case numbers and limitations of PCR testing necessitate advanced diagnostic tools.
• Manual CT scan analysis is time-consuming and prone to errors due to similarities with pneumonia.
• Artificial intelligence offers a promising solution for accurate and efficient COVID-19 diagnosis from CT scans.

Purpose of the Study:

• To develop and evaluate an AI-based framework for classifying COVID-19, pneumonia, and normal cases using high-resolution computed tomography (CT) scans.
• To improve the accuracy and reduce the computation time of COVID-19 diagnosis compared to manual methods.
• To leverage deep learning and optimization algorithms for enhanced medical image analysis.

Main Methods:

• A database of COVID-19, pneumonia, and normal CT scans was prepared.
• Pretrained deep learning models (VGG16, ResNet50, ResNet101) were modified for classification.
• An improved firefly algorithm and a novel activation function were used for feature selection.
• Optimal features were fused and classified using multiclass supervised learning algorithms.

Main Results:

• The proposed AI framework achieved a high accuracy of 97.9% in classifying COVID-19 from CT scans.
• The system demonstrated a significantly reduced computational time of approximately 34 seconds.
• The method effectively distinguished between COVID-19, pneumonia, and normal CT scan patterns.

Conclusions:

• The developed AI framework provides an accurate and efficient tool for COVID-19 diagnosis using CT scans.
• This approach can assist radiologists, reduce diagnostic errors, and improve patient management.
• The study highlights the potential of AI in addressing challenges posed by infectious disease outbreaks.