天眼AI平台结合低mAs在COVID-19胸部CT筛查中对患者的双重保护价值研究

Study on Dual Protection Value of the AI Based uVision Technology Combined with Low mAs in Patients with Chest CT Screening for COVID-19

目的探讨天眼人工智能(Artificial Intelligence,AI)平台结合40 mAs低管电流量在新型冠状病毒肺炎(Coronavirus Disease 2019,COVID-19)胸部CT筛查中对患者的双重保护价值。方法前瞻性随机收集我院2020年2月25日至2020年3月6日行胸部CT筛查的患者共计155例。其中,A组为实验组73例,采用智能辅助摆位,定位框自适应勾画扫描框,参考管电流量40 mAs;B组为对照组82例,采用常规人工摆位,手动勾画扫描框,参考管电流量68 mAs。所有患者均采用自动管电流调制技术。计算两组图像的信噪比(Signal to Noise Ratio,SNR)和对比噪声比(Contrast Noise Ratio,CNR);由两位有经验的胸部影像诊断医生采用5分法对图像进行双盲评分;记录两组的容积剂量指数、剂量长度乘积、有效剂量;记录每例患者检查所需时间,每例患者与当班技师的接触次数。比较疫情之下两组患者做胸部CT扫描的心理差异。利用SPSS 26.0软件进行统计学分析。结果两组图像的SNR、CNR差异均有统计学意义,A组的SNR和CNR均小于B组(P<0.05);两组图像主观评分差异无统计学意义(P=0.132>0.05),均能满足临床诊断;两组辐射剂量差异有统计学意义(P<0.05),A组为(1.08±0.16)mSv,B组为(1.87±0.23)mSv,A组辐射剂量较B组下降约43%;A组患者的平均检查时间比B组节约大约23 s;A组患者与当班技师的接触次数比B组小,差异有统计学意义(P<0.05)。两组患者对院内感染和辐射的心理评分差异有统计学意义(P<0.05)。结论天眼AI平台结合40 mAs应用于COVID-19胸部CT筛查,不仅可以减少患者与当班技师的接触,减少患者的院内感染风险,而且在图像满足临床诊断的同时,可进一步减少患者接受的辐射剂量,形成双重保护。

Objective To investigate the dual protection value of the artificial intelligence(AI)based uVision technology combined with 40 mAs in chest CT screening of patients for coronavirus disease 2019(COVID-19).Methods A total of 155 cases were included prospectively in our hospital for chest CT screening from February 25,2020 to March 6,2020,and were randomly divided them into A group and B group.The 73 cases in A Group were performed by using Easy-Positioning empowered by uAI,auto plan box and 40 mAs low-dose scanning.The 82 cases in B Group were performed with conventional manual positioning,manual drawing of the plan box,and 68mAs low-dose scanning.All cases were adopted automatic tube current modulation technology(ATCM).The SNR and CNR of the images in two groups were calculated.Two experienced radiologists for chest used a 5-point,double blind manner to evaluate the image quality.The Volume dose index,dose length product and effective dose of the two groups were recorded.Also,the scan time of every patient and the contact times between the patient and the radiologic technologist on duty were recorded.The psychological difference between the two groups which performed chest CT scan under the epidemic situation was compared.The software of SPSS 26.0 was used to conduct statistical analysis.Results There were statistically significant differences in SNR and CNR between the two groups.SNR and CNR of A group were both lower than those of B group(P<0.05).There was no statistically significant difference in the subjective scores between the two groups(P=0.132>0.05),which could satisfy the clinical diagnosis.The difference of radiation dose between the two groups was statistically significant(P<0.05),with A group(1.08±0.16)mSv and B group(1.87±0.23)mSv.The scan time in A group was 23 seconds less than B group.The contact times between the patients and the radiologic technologist on duty in A group were smaller than those in B group,and the difference was statistically significant(P<0.05).The differences in mental scores of nosocomial infection and radiation between the two groups were statistically significant(P<0.05).Conclusion The application of the AI based uVision technology combined with 40 mAs in the chest CT screening of COVID-19 not only can reduce the contact between patient and the radiologic technologist,which can reduce the risk of nosocomial infection of patients,but also can reduce the radiation dose received by patients furtherly.While satisfying the clinical diagnosis,the AI based uVision technology has the dual protection value.