Ftl-CoV19: A Transfer Learning Approach to Detect COVID-19

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Infectious Disease Research

Background:

• COVID-19, a contagious disease caused by a novel coronavirus, presents with symptoms like lung infection and breathlessness.
• Limited resources for testing and treatment exacerbate COVID-19 severity.
• Chest X-rays offer potential for rapid COVID-19 diagnosis using AI and machine learning.

Purpose of the Study:

• To propose a novel AI model for efficient COVID-19 detection using chest X-ray images.
• To address the limitations of existing models in terms of efficiency and complexity.

Main Methods:

• Development of the Fine-tuning Transfer Learning-Coronavirus 19 (Ftl-CoV19) model.
• Utilizing transfer learning with a pretrained VGG16 model.
• Incorporating convolution, max pooling, and dense layers within the model architecture.

Main Results:

• The Ftl-CoV19 model achieved high training (98.82%) and validation (99.27%) accuracy.
• Exceptional performance metrics include 100% precision, 98% recall, and 99% F1 score.
• Superior results compared to conventional models like CNN, ResNet50, InceptionV3, and Xception.

Conclusions:

• The Ftl-CoV19 model demonstrates significant potential for accurate and rapid COVID-19 detection from chest X-rays.
• This AI-driven approach offers a more efficient alternative to existing diagnostic methods.
• The model's high performance suggests its utility in clinical settings for early diagnosis and management of COVID-19.