Xiao Xiao-Yun, Chen Xin, Guan Xiao-Feng, Wu Huan, Qin Wei, Luo Bao-Ming
Department of Ultrasound, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China.
Br J Radiol. 2016 Oct;89(1066):20160546. doi: 10.1259/bjr.20160546. Epub 2016 Aug 16.
To evaluate the diagnostic performance of superb microvascular imaging (SMI) in breast lesions, comparing with contrast-enhanced ultrasonographic microvascular imaging (MVI).
From April to November 2015, 132 patients (with 132 breast lesions) were enrolled in the retrospective study. All lesions were evaluated with colour Doppler flow imaging (CDFI), colour SMI (cSMI), monochrome SMI (mSMI) and contrast-enhanced ultrasonographic MVI. Receiver-operating characteristic curve analysis was performed to compare the diagnostic performance of SMI and MVI for discrimination between benign and malignant breast lesions.
Histological analysis showed 58 malignant and 74 benign lesions. mSMI was more sensitive in detecting blood flow signals in breast lesions than CDFI (p < 0.001) and cSMI (p < 0.001). Differences of vessels inside breast lesions and morphologic features of vessels between benign and malignant lesions were statistically significant on mSMI (p < 0.001). Using root hair-like and crab claw-like patterns as the criteria for malignant lesions, the sensitivity, specificity and accuracy for differentiation based on the microvascular architecture patterns were 77.6, 90.5 and 84.8% for mSMI and 89.6, 87.8 and 88.6% for MVI. Areas under curve of mSMI and MVI were not significantly different (p = 0.129).
mSMI can increase blood flow detection and depict the microvascular architecture of breast lesions. The diagnostic performance of mSMI was not significantly different from MVI. SMI has potential in the differential diagnosis of breast lesions.
mSMI is a non-invasive technique for vascularity evaluation of breast tumours and it is beneficial for breast tumour differentiation.
评估超微血管成像（SMI）在乳腺病变中的诊断性能，并与超声造影微血管成像（MVI）进行比较。
2015年4月至11月，132例患者（共132个乳腺病变）纳入回顾性研究。所有病变均采用彩色多普勒血流成像（CDFI）、彩色SMI（cSMI）、单色SMI（mSMI）及超声造影MVI进行评估。采用受试者操作特征曲线分析比较SMI和MVI鉴别乳腺良恶性病变的诊断性能。
组织学分析显示58个恶性病变和74个良性病变。mSMI在检测乳腺病变血流信号方面比CDFI（p＜0.001）和cSMI（p＜0.001）更敏感。mSMI显示乳腺病变内部血管差异及良恶性病变血管形态特征差异有统计学意义（p＜0.001）。以毛根样和蟹爪样模式作为恶性病变标准，基于微血管结构模式鉴别诊断的灵敏度、特异度和准确度，mSMI分别为77.6%、90.5%和84.8%，MVI分别为89.6%、87.8%和88.6%。mSMI和MVI的曲线下面积差异无统计学意义（p = 0.129）。
mSMI可提高血流检测并描绘乳腺病变的微血管结构。mSMI的诊断性能与MVI无显著差异。SMI在乳腺病变鉴别诊断中具有潜力。
mSMI是一种评估乳腺肿瘤血管的非侵入性技术，有助于乳腺肿瘤的鉴别诊断。
