Comparative Analysis of Artificial Intelligence for Predicting COVID-19 using Diverse Chest X-ray Images

Abstract

COVID-19 prediction plays a crucial role in medical decision-making for respiratory health. Accurate and rapid prediction using advanced artificial intelligence (AI) techniques is particularly vital during pandemics. Recent research in the medical field has extensively explored and applied AI-based technologies such as deep learning, attention mechanisms, vision transformers, and explainable AI to various diseases, including COVID-19. However, the limited diversity of medical data poses challenges in accurately evaluating AI models' generalization capabilities for prediction. To address these limitations, we have gathered a comprehensive collection of diverse benchmarks on respiratory diseases, including COVID-19, from various sources. This dataset allows us to assess and compare the performance of AI models under consistent training and testing environments. Our findings aim to inspire researchers in the field and offer valuable insights into the future use of AI techniques in the medical domain. The best classification performance is achieved using ResNet152V2 recording overall accuracy of 96.17%.

Ähnliche Arbeiten