智能三维超微血管成像评估乳腺病变血流的观察者间一致性分析

An inter-observer consistency analysis for the evaluation of blood flow in breast lesions by super microvascular imaging

目的:对智能三维超微血流成像(3D-SMI)技术评估乳腺病变血流的观察者间一致性进行分析。方法:前瞻性研究2018年1-12月北京协和医院225例连续乳腺病变的智能3D-SMI图像,由两位医师对智能3D-SMI图像中病变血流架构进行定性评估,对病变内血管指数(VI)均测量三次取平均值。血流架构采用Weighted Kappa检验一致性,VI值采用Bland-Altman散点图计算测量数据的95%一致性界限。采用Spearman相关分析观察者之间的VI绝对差值与VI均值间是否存在相关性。结果:两位超声医师对智能3D-SMI评估血流架构的Weighted Kappa系数为0.839(95% CI为0.795~0.881, P<0.001),具有较强的观察者间一致性。VI值Bland-Altman图分析显示95.11%的值在临床允许范围之内,具有较好的一致性,观察者间VI值差值与VI均数大小间存在相关性( r s=0.639, P<0.001)。 结论:智能3D-SMI评估乳腺病变血流架构及定量测量血流丰富程度在观察者间具有较好的一致性。

Objective To analyze the inter-observer consistency of Smart 3D super microvascular imaging(Smart-3D SMI)in evaluating the blood flow of breast lesions.Methods Prospective study on Smart-3D SMI images of 225 cases of continuous breast lesions in Peking Union Medical College Hospital from January to December 2018.Two sonographers conducted qualitative evaluation of vascular architecture(VA)under Smart-3D SMI,and took the mean value of three measurements evaluations of vascular index(VI).Weighted Kappa was used to test the consistency of the vascular architecture,and Bland-Altman scatter plot was used to calculate the 95%consistency limit of the measured data for the VI value.Spearman correlation coefficient was used to analyze whether the absolute difference of VI between observers was correlated with the mean value of VI.Results The Weighted Kappa coefficient of Smart-3D SMI for the evaluation of VA by two sonographers was 0.839(95%CI was 0.795-0.881,P<0.001),showing strong inter-observer consistency.Bland-Altman plot analysis of VI value showed that the value of 95.11%was within the clinically acceptable range and had good consistency,and there was a correlation between the difference of VI value between observers and the value of VI mean(rs=0.639,P<0.001).Conclusions There is a good inter-observer consistency in evaluating the blood vascular architecture and quantitatively measuring the blood flow abundance by Smart-3D SMI.