Classification of pneumonia caused by Covid-19 based on deep learning model

With the unexpected spread of Covid-19 in 2019, such disease took away millions of peoples lives. Therefore, investigating and curing Covid-19 become a very mandatory issue in different areas, such as biology, medicine, and statistics. This paper investigates different models of CNN in deep learning of computers in analyzing X-ray pictures of normal pneumonia and Covid-19 caused pneumonia patients. The database is from Kaggle and contains over 8000 images of X-rays of the chest. Besides, this paper discusses the imaging process technology, such as ConvNeXt, to edit X-ray images more convenient for computers to analyze and dispose of. According to the comparison of the sequential model and DenseNet model in CNN, the sequential model has better performance and accuracy. In the conclusion part, this paper also investigates whether better image processing work can improve the performance of models. Overall, these results shed light on guiding further exploration of both analyzing and distinguishing Covid-19 patients and normal pneumonia patients in order to decrease the work of hospitals and cure different patients in time.