基于复合灵敏度编码直方图参数及表观扩散系数对乳腺良、恶性病变的鉴别诊断

Apparent Diffusion Coefficient and Its Histogram in Differentiating Benign and Malignant Breast Lesions via Multiplexed Sensitivity Encoding Diffusion Weighted Imaging

目的探讨基于复合灵敏度编码(MUSE)序列的全容积表观扩散系数(ADC)直方图参数及ADC值鉴别诊断乳腺良、恶性病变的价值。资料与方法回顾性分析2020年9月—2022年1月宁夏医科大学总医院经病理证实的112例乳腺病变患者,其中良性30例,恶性82例。术前均行MRI常规序列、动态对比增强MRI及MUSE高分辨率扩散成像(MUSE-DWI)。参照增强图像,对MUSE-DWI序列进行全瘤体勾画,获得病灶的15项ADC直方图参数;测量MUSE-DWI的ADC值(ADC_MUSE)。评估数据测量的可重复性,比较乳腺良、恶性病变ADC直方图参数及ADC值的差异。采用受试者工作特征曲线评估参数的诊断效能。结果除ADC偏度、ADC峰度及ADC熵外,其他ADC直方图参数在乳腺良、恶性病变组间差异均有统计学意义(P<0.05)。恶性组ADC均值、ADC 1%、5%、10%、25%、50%、75%、90%、95%、99%显著低于良性组,ADC标准差、ADC变异系数显著高于良性组。ADC 5%的鉴别诊断价值最高,曲线下面积为0.919,以0.94×10^(−3)mm^(2)/s为截断值,敏感度和特异度分别为95.12%、83.33%。ADC_MUSE在良、恶性组间差异有统计学意义(P<0.05),曲线下面积为0.896,以1.13×10^(−3)mm^(2)/s为截断值,敏感度和特异度分别为82.93%、86.67%。ADC 5%联合ADC_MUSE的曲线下面积为0.923,敏感度和特异度分别为90.24%、86.67%。结论ADC直方图参数鉴别乳腺良、恶性病变有一定价值,其中ADC 5%的鉴别诊断价值最高;MUSE-DWI直方图参数联合ADC值可进一步提高诊断效能。

Purpose To investigate the apparent diffusion coefficient(ADC)value and its histogram in the differential diagnosis of benign and malignant breast lesions based on multiplexed sensitivity encoding diffusion weighted imaging(MUSE-DWI).Materials and Methods 112 breast lesions confirmed by pathology from September 2020 to January 2022 in General Hospital of Ningxia Medical University were retrospectively analyzed,including 30 benign lesions and 82 malignant lesions.All patients underwent breast routine MRI,dynamic contrast�enhanced MRI and MUSE-DWI scans.According to the enhanced images,the Firevoxel software was used to delineate the whole tumor from MUSE-DWI sequences,including 15 ADC histogram parameters of the lesions.The ADC value of MUSE-DWI(ADC_MUSE)was measured.The inter-observer repeatability was evaluated,and the differences in ADC histogram parameters as well as ADC values between benign and malignant breast lesion groups were also compared.The receiver operating characteristic curve was used to evaluate the diagnostic performance.Results Except for ADC_skewness,ADC_kurtosis and ADC_entropy in ADC histogram parameters,the other parameters were significantly different between two groups(P<0.05).The ADC_mean,ADC 1%,ADC 5%,ADC 10%,ADC 25%,ADC 50%,ADC 75%,ADC 90%,ADC 95%,ADC 99%of malignant group were significantly lower than those of benign group,while ADC_StDev and ADC_CoeffOfVariation were significantly higher than those of benign group.ADC 5%had the highest differential diagnosis value,the area under the curve was 0.919,the cut-off value was 0.94×10^(−3)mm^(2)/s,the sensitivity and specificity were 95.12%and 83.33%,respectively.There was statistically significant difference in ADC_MUSE between the two groups(P<0.05),and the area under the curve of ADC_MUSE was 0.896,with a cut-off value of 1.13×10^(−3)mm^(2)/s,the sensitivity and specificity were 82.93%and 86.67%,respectively.The area under the curve of ADC 5%combined with ADC_MUSE value was 0.923,the sensitivity and specificity were 90.24%and 86.67%,respectively.Conclusion ADC histogram parameters may be meaningful in differentiating benign and malignant breast lesions,among which ADC 5%has the highest differential diagnostic value.MUSE-DWI histogram parameters combined with ADC values can further improve the diagnostic performance.