Design and Implementation of Optimized Feature Extraction method in Deep Learning Models for Enhanced COVID-19 Detection

Abstract

Nowadays, COVID-19 is a global pandemic, which was developed in 2019, and has posed enormous limitations to healthcare systems worldwide, emphasizing the need for rapid and accurate detection techniques. Early diagnosis plays an important role in controlling its spread and providing timely treatment. However, existing detection systems often face challenges in terms of accuracy, efficiency, and the capability to handle large datasets. This study offers a novel deep learning (DL) based approach to improve the detection of COVID-19 by focusing on optimized feature extraction and comparing its performance with existing systems. The research begins with an analysis of current systems for COVID-19 detection, identifying their strengths, limitations, and areas where improvements can be made. Building on this analysis, a new optimized feature extraction algorithm is proposed to extract global feature matrices from relevant datasets, aiming to enhance the efficiency of the detection procedure. A DL model is then developed to detect COVID-19, leveraging advanced neural network architectures to provide accurate results. Finally, the recital of the implemented system is associated with existing techniques using several evaluation parameters, like accuracy, precision, etc. The outcome of this study prove that the proposed system significantly outperforms current detection systems in terms of both accuracy and processing time. This research contributes to the field of AI-driven medical diagnostics and provides valuable insights for improving COVID-19 detection systems, offering a more well-organized and precise result for combating the ongoing pandemic. The proposed system also lays the foundation for future advancements in the application of DL in healthcare.