Research on pneumonia image analysis based on deep learning

Abstract

Pneumonia is a severe respiratory disease that can pose a significant threat to patients. Since the novel coronavirus (COVID-19) began to spread globally, the pneumonia it causes has led to millions of deaths worldwide. Early diagnosis of pneumonia is a critical step in its treatment, and pulmonary imaging examinations are the most essential tools for diagnosing the condition. While previous papers have summarized the utilization of deep learning in pneumonia image diagnosis, the rapid evolution of deep learning models and algorithms needs a more systematic review of both classical and contemporary models. In this paper, the author employs a systematic review to provide a comprehensive analysis and evaluation of deep learning models for pneumonia imagery, as well as a comparison of these models. It introduces classic models used for processing pneumonia images and also presents the latest research methods. A systematic summary of these deep learning models can help people better understand and learn about deep learning models related to pneumonia imagery, learn from past mistakes, and thereby enhance the precision of deep learning models for pneumonia diagnosis.