医师资历和ROI设置对IDEAL-IQ评价肝脏脂肪含量的影响

目的 探讨医师资历和感兴趣区（ROI）设置对非对称回波三点法水脂分离-定量测定（IDEAL-IQ）评价肝脏脂肪含量的影响。 方法 选取2017年1月至3月于佛山市第一人民医院43例疑诊为脂肪肝患者[男性39例、女性4例，年龄（40.37±14.65）岁]的IDEAL-IQ数据，由3位医师以3种ROI面积（10、25、50 mm2）测定肝脏质子密度脂肪含量（PDFF）。由1位医师分别以距离肝脏包膜10、20、30 mm为中心勾画50 mm2 ROI测定PDFF。符合正态分布且方差齐的计量资料采用单因素方差分析。采用SPSS 17.0信度分析检验同一医师按不同面积ROI重复测定同组资料的一致性及不同医师按同一ROI面积重复测定同组资料的一致性。采用组内相关系数（ICC）比较检验结果的重复性。alpha系数>0.8且ICC>0.75表示可信度和可重复性均高。 结果 同一位医师按3种ROI面积测定的PDFF分别为（14.17±8.40）%、（13.49±8.42）%和（13.25±8.39）%，差异无统计学意义（F=0.138，P=0.871）。3位医师按同一ROI面积测定的PDFF分别为（14.10±8.81）%、（12.75±8.48）%和（14.06±8.22）%，差异无统计学意义（F=0.352，P=0.704）。同一位医师按3种ROI面积和3位医师按同一ROI面积测定的2组PDFF数据的alpha系数均>0.8且ICC均>0.75。距离肝脏包膜10、20、30 mm处测定的PDFF的差异无统计学意义。 结论 医师资历和ROI设置对IDEAL-IQ评价肝脏PDFF无影响，其具有极高的可信度和可重复性。

Radiologists' qualifications and regions of interest settings to evaluate liver fat by IDEAL-IQ sequence

Objective To explore the effect of radiologists' qualifications and region of interest (ROI) settings on the evaluation of liver proton density fat fraction (PDFF) by iterative decomposition of water and fat with echo asymmetry and least-squares estimation and image quantification (IDEAL-IQ). Methods A total of 43 patients(39 males and 4 females, 40.37±14.65 years old) underwent 3D IDEAL-IQ scanning by a GE Discovery MR750W 3.0 T MRI scanner at the First People's Hospital of Foshan. PDFFs were determined on fat fraction maps by three different radiologists with ROIs of 10, 25, and 50 mm2. When the measurement data showed a normal distribution and the variance was homogeneous, one-way ANOVA was used for statistical analysis. The consistency of the same group of data measured repeatedly by the same researcher according to the ROI of different areas with that measured by different radiologists was evaluated via reliability analysis by using SPSS 17.0. Intra-group correlation coefficients (ICC) were used to compare the repeatability of the test results. In this work, alpha>0.8 and ICC>0.75 were used to demonstrate high reliability and repeatability. Results The PDFFs measured with three different ROIs by the same doctor were (14.17±8.40)%, (13.49±8.42)%, and (13.25±8.39)%, respectively. There was not significant difference (F=0.138, P=0.871).The PDFFs measured by three doctors according to the same ROI respectively were (14.10±8.81)%, (12.75±8.48)% and (14.06±8.22)%, respectively. The difference was not statistically significant (F=0.352, P=0.704). The same radiologists determined the same group of data on the fat fraction maps with different ROIs, and reliability analysis indicated alpha>0.8 and ICC>0.75. These results indicate that the determination of PDFFs with different ROIs has high reliability and repeatability. The three radiologists then measured the same group of data on fat fraction maps with the same ROI, and reliability analysis indicated alpha>0.8 and ICC>0.75. Differences between the PDFFs measured at 10, 20, and 30 mm involved the vertical distance between the center of the ROI and the hepatic envelope. Conclusions Radiologists' qualifications and ROI settings have no effect on the IDEAL-IQ sequence evaluation of liver PDFF. Measurements made under different qualifications and ROIs by using the IDEAL-IQ technique have high reliability and repeatability.