Cong Rui, Li Jing, Guo Song
Department of Ultrasound, Shengjing Hospital of China Medical University, Shenyang, 110004, China.
Eur J Radiol. 2017 Feb;87:111-119. doi: 10.1016/j.ejrad.2016.12.021. Epub 2016 Dec 21.
To examine the efficacy of qualitative shear wave elastography (SWE) in the classification and evaluation of solid breast masses, and to compare this method with conventional ultrasonograghy (US), quantitative SWE parameters and qualitative SWE classification proposed before.
From April 2015 to March 2016, 314 consecutive females with 325 breast masses who decided to undergo core needle biopsy and/or surgical biopsy were enrolled. Conventional US and SWE were previously performed in all enrolled subjects. Each mass was classified by two different qualitative classifications. One was established in our study, herein named the Qual1. Qual1 could classify the SWE images into five color patterns by the visual evaluations: Color pattern 1 (homogeneous pattern); Color pattern 2 (comparative homogeneous pattern); Color pattern 3 (irregularly heterogeneous pattern); Color pattern 4 (intralesional echo pattern); and Color pattern 5 (the stiff rim sign pattern). The second qualitative classification was named Qual2 here, and included a four-color overlay pattern classification (Tozaki and Fukuma, Acta Radiologica, 2011). The Breast Imaging Reporting and Data System (BI-RADS) assessment and quantitative SWE parameters were recorded. Diagnostic performances of conventional US, SWE parameters, and combinations of US and SWE parameters were compared.
With pathological results as the gold standard, of the 325 examined breast masses, 139 (42.77%) samples were malignant and 186 (57.23%) were benign. The Qual1 showed a higher Az value than the Qual2 and quantitative SWE parameters (all P<0.05). When applying Qual1=Color pattern 1 for downgrading and Qual1=Color pattern 5 for upgrading the BI-RADS categories, we obtained the highest Az value (0.951), and achieved a significantly higher specificity (86.56%, P=0.002) than that of the US (81.18%) with the same sensitivity (94.96%).
The qualitative classification proposed in this study may be representative of SWE parameters and has potential to be relevant assistance in breast mass diagnoses.
探讨定性剪切波弹性成像（SWE）在乳腺实性肿块分类及评估中的效能，并将该方法与传统超声（US）、定量SWE参数及既往提出的定性SWE分类方法进行比较。
2015年4月至2016年3月，连续纳入314例有325个乳腺肿块且决定接受粗针活检和/或手术活检的女性。所有纳入研究的对象均先行传统US和SWE检查。每个肿块采用两种不同的定性分类方法进行分类。一种是本研究建立的，在此命名为Qual1。Qual1可通过视觉评估将SWE图像分为五种颜色模式：颜色模式1（均匀模式）；颜色模式2（相对均匀模式）；颜色模式3（不规则不均匀模式）；颜色模式4（病灶内回声模式）；颜色模式5（硬边征模式）。第二种定性分类在此命名为Qual2，包括一种四色叠加模式分类（Tozaki和Fukuma，《放射学报》，2011年）。记录乳腺影像报告和数据系统（BI-RADS）评估结果及定量SWE参数。比较传统US、SWE参数以及US和SWE参数组合的诊断效能。
以病理结果为金标准，在325个接受检查的乳腺肿块中，139个（42.77%）样本为恶性，186个（57.23%）为良性。Qual1的Az值高于Qual2和定量SWE参数（所有P<0.05）。当应用Qual1 =颜色模式1进行BI-RADS分类降级、Qual1 =颜色模式5进行BI-RADS分类升级时，获得了最高的Az值（0.951），且在相同灵敏度（94.96%）下，特异性（86.56%，P = 0.002）显著高于US（81.18%）。
本研究提出的定性分类可能代表SWE参数，有潜力为乳腺肿块诊断提供相关辅助。
