Research progress in the application of deep learning for intelligent diagnosis in COVID-19
Corona virus disease 2019 (COVID-19) is highly contagious and poses a serious threat to the lives of people,and rapid screening can realize rapid treatment and prevent the progression of COVID-19.The current gold standard for COVID-19 detection is reverse transcription-polymerase chain reaction (RT-PCR) .However,RT-PCR is time-consuming and has a high false-negative rate,and radiologists' diagnosis of medical images is subjective and the workload is huge.So using artificial intelligence (AI) technology is essential to rapidly diagnose COVID-19.With the successful application of artificial intelligence in the medical field,deep learning techniques have become an effective method to assist in the diagnosis of COVID-19.In recent years,many scholars have used deep learning techniques to construct models for intelligent diagnosis of medical images,and the main content of this paper is to summarize and analyze such models,introducing models for segmenting lung regions,classification models for realizing binary classification or multi-classification as well as the clinical applications of the models.Meanwhile,in the article,the imaging features of COVID-19 patients are analyzed.COVID-19 patients tend to have bilateral lung involvement,in which ground glass shadow is the most common sign of images.The latest progress in COVID-19 research is also presented,mainly about the development of improving the accuracy of the AI model and the related research of"long COVID-19"syndrome.Therefore,under the normalized management of COVID-19,model accuracy can be improved by expanding the dataset or lightweighting the model structure,etc.As a new research field,scholars can further study the"long COVID-19"syndrome in terms of clinical symptoms,prognostic follow-up,and the combination of deep learning techniques.
deep learningCOVID-19artificial intelligencecomputed tomographysegmentation networkclassification network
万方数据
维普
