IMPACTS OF COVID-19 VACCINATION ON CLINICAL SEVERITY AND LUNG DAMAGE BASED ON ARTIFICIAL INTELLIGENCE CLASSIFICATION

Abstract

Background: COVID-19 is an endemic outbreak that affects nearly all countries, thereby making vaccination crucial for the prevention and mitigation of infection severity. Variations in infection risk, hospitalization rates, infection severity, and mortality exist between vaccinated and unvaccinated individuals. The aim of this study to assess the impact of vaccination on the clinical severity and the degree of lung parenchymal damage according to artificial intelligence classification.

Methods: Using the medical records of COVID-19 patients at Dr. Zainoel Abidin Banda Aceh Hospital from January to December 2022, a retrospective design was used. Artificial intelligence model, constructed using visual geometry G architecture, was used to classify the clinical severity based on the symptoms and chest X-ray images, respectively.

Results: Out of the 153 patients recruited, 20% were unvaccinated, while 15% received a single dose and 65% received two doses of vaccination. According to the classification using the artificial intelligence model, 12.9% of the unvaccinated patients had critical conditions. While none were detected experiencing a severe damage, moderate level of lung parenchymal damage was found in 29%, 27.3%, and 7% of the unvaccinated, single-dose vaccination, and two-doses vaccination groups, respectively. The significant association between vaccination and the degree of lung parenchymal damage and clinical severity (p<0.05).

Conclusion: The administration of two vaccine doses correlated significantly with clinical severity and lung parenchymal damage. Therefore, vaccination provides protective effects against the disease progression. Policymakers should prioritize vaccination strategies due to the preventive benefits against the outbreak.