ZooMit-DWI对肺癌的病理分型诊断价值:与常规DWI的对照研究

Diagnostic value of ZooMit-DWI for pathological classification of lung cancer:compared with conventional DWI

目的:探讨ZooMit-DWI在肺癌成像中的可行性及在肺癌不同病理类型及分化程度鉴别中的价值。方法:连续性收集76例肺癌患者的临床、影像及病理资料,所有患者均进行常规DWI(C-DWI)和ZooMit-DWI(Z-DWI),并分别测量其ADC值。比较两种成像序列的图像质量,评估不同观察者测量ADC值的一致性。分析两种DWI序列在不同肺癌病理类型及肺癌高中分化、低分化两组中的ADC值差异,并采用受试者操作特征(ROC)曲线评价其诊断效能。结果:最终入组63例患者,其中肺小细胞癌9例,鳞癌20例,腺癌34例。Z-DWI主观图像质量评分及图像信噪比、对比度、对比噪声比均较C-DWI图像高(P<0.05)。不同观察者测量C-DWI与Z-DWI图像ADC值的ICC分别为0.746(95%CI:0.61~0.84)、0.886(95%CI:0.82~0.93)。肺小细胞癌、鳞癌、腺癌的Z-ADC平均值两两比较差异均有统计学意义(P<0.05)。常规ADC平均值在小细胞癌组与鳞癌组(P<0.05)、腺癌组(P<0.05)间的差异有统计学意义,而在鳞癌组和腺癌组间的差异没有统计学意义(P>0.05)。Z-ADC值及C-ADC值在肿瘤中高分化组、低分化组差异均有统计学意义(P<0.05),ROC曲线显示Z-ADC值及C-ADC值诊断中高分化、低分化肺癌的ROC曲线下面积分别为0.897、0.776。结论:ZooMit-DWI序列在肺部肿瘤成像的图像质量及ADC值测量的可重复性均优于常规DWI序列,ZooMit-DWI序列有助于肺癌不同病理类型及组织分化程度的鉴别诊断。

Objective:To explore the feasibility of ZooMit-DWI in lung cancer imaging and its value in identifying pathological types and differentiation degree of lung cancer.Methods:Clinical,imaging,and pathological data of 76 lung cancer patients were collected consecutively.All patients underwent conventional DWI(C-DWI) and ZooMit DWI(Z-DWI),and ADC values were measured respectively.The image quality of the two sequences was compared and the inter-observers agreement of the ADC values was assessed.The differences in ADC values between two DWI sequences in different pathological types and different differentiation degrees of lung cancer were analyzed.Receiver operating characteristic(ROC) curves were used to evaluate the diagnostic efficiency.Results:63 patients were finally enrolled,including 9 small cell carcinoma,20 squamous carcinomas,and 34 adenocarcinomas.Subjective image quality score,image signal-to-noise ratio(SNR),contrast ratio,and contrast-to-noise ratio of Z-DWI were higher than that of C-DWI(P<0.05).The ICC of C-ADC and Z-ADC were 0.746(95%CI:0.61~0.84) and 0.886(95%CI:0.82~0.93),respectively.The mean values of Z-ADC for small cell carcinoma,squamous carcinoma,and adenocarcinoma of the lung were statistically significant in pairwise comparisons(P<0.05).The differences in C-ADC mean values were statistically significant between the small cell carcinoma group and the squamous carcinoma group(P<0.05),and the adenocarcinoma group(P<0.05),while the difference between the squamous carcinoma and the adenocarcinoma groups were not statistically significant(P>0.05).The differences in Z-ADC values and C-ADC values were significantly different in the highly and poorly differentiated groups(P<0.05).The ROC curves showed that the AUC values were 0.897 and 0.776,respectively,for the diagnosis of highly and poorly differentiated lung cancer.Conclusion:The image quality and repeatability of ADC value measured by ZooMit-DWI in lung tumor imaging are better than conventional DWI.ZooMit-DWI is useful for the differential diagnosis of different pathological types and differentiation degrees of lung cancer.