白细胞介素-6与新型冠状病毒B.1.617.2变异株引起肺损伤相关性的人工智能定量分析  

Artificial intelligence quantitative analysis reveals the effect of interleukin-6 injury in 2019 novel coronavirus B.1.617.2

作　　者：甘清鑫[1] 刘晋新[1] 杨彦鸿 胡天丽 张烈光[1] 黄德扬[1] 李美瑜 Gan Qingxin;Liu Jinxin;Yang Yanhong;Hu Tianli;Zhang Lieguang;Huang Deyang;Li Meiyu(Department of Radiology,Guangzhou Eighth People's Hospital,Guangzhou Medical University,Guangzhou 510440,China;Cardiovascular Division,Guangzhou Eighth People's hospital,Guangzhou Medical University,Guangzhou 510440,China)

机构地区：[1]广州医科大学附属市八医院放射科,广州510440 [2]广州医科大学附属市八医院心血管内科,广州510440

出　　处：《新发传染病电子杂志》2021年第3期172-176,共5页Electronic Journal of Emerging Infectious Diseases

基　　金：广东省中医药局中医药科研项目(20211303)。

摘　　要：目的探讨白细胞介素-6(IL-6)与新型冠状病毒(2019-nCoV)B.1.617.2变异株引起肺损伤的相关性,提示临床应及时检测IL-6的水平和提前干预,以期降低重症肺炎发生的风险。方法收集2021年5月21日至2021年6月21日广州医科大学附属市八医院收治的2019-nCoV B.1.617.2变异株感染患者123例为研究对象进行回顾性研究,轻症患者105例,重症患者18例。比较两组患者的临床特点、炎性指标水平、IL-6及人工智能(AI)测量肺损伤程度。结果2019-nCoV B.1.617.2变异株感染的重症患者的年龄均值、血清降钙素原(PCT)峰值、IL-6峰值、AI测量肺损伤峰值占比均高于轻症组(P<0.05),重症组患者血清IL-6峰值均值为(197.28±105.70)pg/ml,AI测量肺损伤峰值占比均值(36.21±14.91)%,肺损伤以双侧肺部多发磨玻璃样变和实变为主要影像学特征。结论IL-6高表达是2019-nCoV B.1.617.2变异株引起肺损伤的潜在的免疫风险因素,应用AI定量分析有助于判断肺炎的严重程度,提前干预,以期降低患者发展为重症肺炎的风险。Objective To investigate the correlation of lung injury caused by interleukin-6(IL-6)in 2019 novel coronavirus(2019-nCoV)B.1.617.2.It is suggested that the level of IL-6 should be detected in time and intervention should be carried out in advance in order to reduce the risk of severe pneumonia.Methods A retrospective study was conducted to collect 123 patients infected with 2019-nCoV B.1.617.2 from the Guangzhou Eighth People's Hospital,Guangzhou Medical University,in May 21 to June 21,2021.There were 105 mild cases and 18 severe cases.The clinical characteristics,inflammatory index levels,IL-6 and artificial intelligence(AI)lung injury degree of the two groups were analyzed.Results The mean value of age,peak value of procalcitonin(PCT),peak value of serum IL-6 and peak value of pulmonary injury measured by AI in patients with 2019-nCoV B.1.617.2 strain were higher than those in patients with mild disease group(P<0.05).The mean value of serum IL-6 in patients with severe disease group was(197.28±105.70)pg/ml.The peak value of lung injury measured by AI accounted for(36.21±14.91)%,and the main imaging features of lung injury were multiple ground glass opacity and consolidation in bilateral lungs.Conclusion High expression of IL-6 is a potential immune risk factor for lung injury in patients with 2019-nCoV B.1.167.2 strain.AI quantitative analysis helps to determine the severity of pneumonia in patients,intervention as early as possible.In order to reduce the risk of severe pneumonia.

关 键 词：新型冠状病毒 病毒变异株 肺炎 白细胞介素-6 人工智能 

分 类 号：R563[医药卫生—呼吸系统]