人工智能联合肺部低剂量CT在新型冠状病毒肺炎诊断中的应用评价

目的:探讨人工智能(AI)与肺部低剂量CT(LDCT)联合应用于新型冠状病毒肺炎(COVID-19)中有效降低辐射剂量的可行性方案.方法:选取在医院就诊的216例COVID-19住院患者,采用随机数表法将其分为常规剂量CareDose模式的80～150 mAs组、低剂量CareDose模式的21～50 mAs组和人工固...

Application of AI combined with low dose CT on lung in COVID-19 diagnosis

Objective:To discuss the feasible plan of the combination of artificial intelligence(AI)and low-dose computed tomography(LDCT)on lung in the application of effective reducing radiation dose for patients with Coronavirus Disease 2019(COVID-19).Methods:216 inpatients with COVID-19 were selected and were randomly divided into three groups:80-150 mAs group with conventional dose CareDose mode,21-50 mAs group with low dose Care Dose mode and 20 mAs group with artificial fixed low tube current mode,with 72 cases in each group.The scanning images of patients in three groups were read by AI,and the image quality and detection coincidence rate were judged as double blind method by three senior radiologists.And the CT dose index(CTDI_vol),dose-length product(DLP)and diagnostic coincidence rate of the 21-50 mAs group closed to those of 80-150 mAs group and were significantly better than those of 20 mAs group,and the differences of those between the two groups were significant.And the CTDI_vol,DLP and ED of 21-50 mAs group were significantly lower than those of 80-150 mAs group,and the differences of these indicators between 21-50 mAs group and 80-150 mAs,and between 21-50 mAs group and 20 mAs group(F=514.05,F=589.02,F=246.96,P<0.05),respectively.Conclusion:The combination of AI and low-dose automatic tube current modulation technique can significantly reduce the radiation dose of patients with COVID-19 in CT examination and does not affect the diagnostic detection ratel.