Deploying Machine and Deep Learning Models for Efficient Data-Augmented Detection of COVID-19 Infections.

Area of Science:

• Medical Informatics
• Artificial Intelligence
• Epidemiology

Background:

• The COVID-19 pandemic presents an unprecedented global health crisis, exacerbated by challenges in early detection and containment.
• Machine and deep learning models show promise in disease detection but are limited by data availability and reliability.
• Existing deep learning models (DLMs) struggle with accurate COVID-19 detection due to insufficient training data.

Purpose of the Study:

• To address the data scarcity issue hindering deep learning applications in COVID-19 detection.
• To propose and evaluate two novel data-augmentation models for enhancing Convolutional Neural Network (CNN) and Convolutional Long Short-Term Memory (ConvLSTM) models.
• To improve the accuracy and efficiency of COVID-19 detection using augmented deep learning models (DADLMs).

Main Methods:

• Development of two distinct data-augmentation models tailored for deep learning architectures.
• Integration of these models with CNN and ConvLSTM frameworks to create data-augmented deep learning models (DADLMs).
• Comparative analysis of DADLMs against traditional DLMs and machine learning techniques using key performance metrics.

Main Results:

• DADLMs demonstrated improved accuracy, reduced logarithmic loss, and faster testing times compared to DLMs without augmentation.
• The proposed models achieved an average increase of 4% to 11% in COVID-19 detection accuracy over conventional machine learning methods.
• Experimental results confirm the enhanced learnability and diagnostic performance of the augmented models.

Conclusions:

• The developed data-augmentation models are effective in boosting the performance of deep learning for COVID-19 detection.
• DADLMs offer a viable solution for rapid and consistent coronavirus diagnosis, supporting clinical decision-making.
• This approach enhances the contribution of AI to combating global health emergencies like the COVID-19 pandemic.