Computer Engineering and Applications ›› 2022, Vol. 58 ›› Issue (12): 51-65.DOI: 10.3778/j.issn.1002-8331.2109-0110

• Research Hotspots and Reviews • Previous Articles     Next Articles

Review of Deep Learning Diagnostic Methods for COVID-19

MA Jinlin, QIU Shuo, MA Ziping, CHEN Yong   

  1. 1.School of Computer Science and Engineering, North Minzu University, Yinchuan 750021, China
    2.Key Laboratory of Intelligent Information Processing of Image and Graphics, State Ethnic Affairs Commission, Yinchuan 750021, China
    3.School of Mathematics and Information Science, North Minzu University, Yinchuan 750021, China
    4.Department of Interventional Radiology, Ningxia Medical University General Hospital, Yinchuan 750004, China
  • Online:2022-06-15 Published:2022-06-15



  1. 1.北方民族大学 计算机科学与工程学院,银川 750021
    2.图像图形智能信息处理国家民委重点实验室,银川 750021
    3.北方民族大学 数学与信息科学学院,银川 750021
    4.宁夏医科大学总医院 放射介入科,银川 750004

Abstract: The high rate of infection has led to the rapid spread of COVID-19 across the globe, and the commonly used reverse transcription-polymerase reaction(RT-PCR) test has drawbacks such as time consuming, high false-negative rates, and inadequate medical equipment. Therefore, the development of efficient, accurate, and low-cost test techniques is critical for the diagnosis and treatment of COVID-19. With the successful application of Artificial intelligence in medicine, deep learning technology has become an effective method to assist detection and identification of COVID-19. This paper studies and summarizes the emerging deep learning diagnostic methods for COVID-19 in recent years. Two COVID-19 data sets used by deep learning methods are introduced. Six deep learning diagnostic methods based on VGGNet, Inception, ResNet, DenseNet, EfficientNet and CapsNet models are introduced. Three diagnostic methods combining deep learning with other machine learning methods are introduced. The research trend of COVID-19 diagnosis based on deep learning is prospected.

Key words: COVID-19, deep learning, X-ray, CT, lightweight

摘要: 新型冠状病毒肺炎的高感染率导致其在全球范围内迅速传播,常用的逆转录-聚合酶反应(RT-PCR)检测方法存在耗时、假阴性率偏高和医学用具不足的缺陷,因此开发高效、准确、低成本的影像检测技术对新型冠状病毒肺炎的诊断和治疗至关重要。随着人工智能在医学领域的成功应用,深度学习技术成为辅助检验和识别新型冠状病毒肺炎的有效方法。对近年来涌现的新型冠状病毒肺炎的深度学习诊断方法进行了研究和总结:介绍了深度学习方法使用的两种新型冠状病毒肺炎数据集;介绍了基于VGGNet、Inception、ResNet、DenseNet、EfficientNet和CapsNet模型的六种深度学习诊断方法;介绍了三种深度学习与其他机器学习方法结合的诊断方法;对基于深度学习的新型冠状病毒肺炎诊断方法的研究趋势进行了展望。

关键词: 新型冠状病毒肺炎, 深度学习, X射线, CT, 轻量化