A deep transfer learning-based convolution neural network model for COVID-19 detection using computed tomography scan...

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Machine Learning

Background:

• Coronavirus disease (COVID-19) poses a significant global health and economic threat due to its rapid international spread.
• Traditional diagnostic methods like RT-PCR can yield false negatives, necessitating alternative detection strategies.
• Medical imaging, particularly chest CT scans, offers a non-invasive approach for COVID-19 detection.

Purpose of the Study:

• To develop and evaluate deep learning models, specifically Convolutional Neural Network (CNN) frameworks enhanced with transfer learning, for detecting COVID-19 using CT scan images.
• To investigate the efficacy of various pre-trained CNN models for COVID-19 detection, even with limited datasets.
• To compare the performance of different transfer learning-based CNN architectures in identifying COVID-19 from chest CT scans.

Main Methods:

• Utilized transfer learning with established CNN architectures: VGG16, VGG19, Densenet121, InceptionV3, Xception, and Resnet50.
• Trained and evaluated these models on chest CT images for COVID-19 detection.
• Assessed model performance using metrics including accuracy, precision, recall, F1-score, loss, and ROC curves, alongside confusion matrices.

Main Results:

• The VGG16 model demonstrated superior performance among the evaluated architectures, achieving an accuracy of 98.00%.
• The study confirmed the effectiveness of deep transfer learning approaches for COVID-19 detection, even with constrained datasets.
• Experimental outcomes highlighted the potential of the proposed models in aiding the screening and monitoring of COVID-19 patients.

Conclusions:

• Deep transfer learning-based CNN models, particularly VGG16, show significant promise for accurate and efficient COVID-19 detection from chest CT scans.
• The developed framework can assist healthcare professionals in making timely therapeutic decisions for COVID-19 patients.
• This research contributes to the advancement of AI-driven tools for infectious disease diagnosis and management.