A Deep Learning Framework for Precise COVID-19 Classification from Chest X-Ray Images

A future where COVID-19 identification is significantly simplified is envisioned. The power of deep learning tools, specifically pre-trained Convolutional Neural Network (CNN) models, is explored to revolutionize COVID-19 detection. Enhanced speed and intelligence in healthcare responses are aimed in this study. COVID-19, caused by the SARS-CoV-2 virus, is presented as a complex and variable infectious condition. An uncertain prognosis often results from its pathological instability and symptom overlap with other diseases. While mild to severe respiratory illness is commonly experienced by the majority of virus-infected individuals, often resolving without specialized clinical intervention, severe illness requiring critical medical intervention is developed by some. A higher risk of severe outcomes is faced by vulnerable groups, including the elderly and those with underlying health issues like cancer, diabetes, cardiovascular, or chronic respiratory diseases. Early detection of virus carriers is considered crucial for effectively combating such viral diseases. A powerful method for classifying chest Xray images, thereby assisting in the pre-diagnostic identification of COVID-19, is offered by deep learning. In this study, pretrained CNN models were leveraged and compared with the MixMatch model. The CNN's significantly superior performance was observed, which led to its selection for further development. Exceptional accuracy was achieved through meticulous training and fine-tuning, including adjustments to learning rates and epochs, showcasing the advancements needed for superior diagnostic capability. The primary objective of this study is to demonstrate the accuracy of supervised deep learning in COVID-19 classification, particularly in terms of prognosis, with semi-supervised approaches being outperformed. This work contributes to the creation of robust tools for the early detection and classification of COVID-19 cases, ultimately facilitating more effective healthcare interventions.