Pattern classification of ShearWaveâ„¢ Elastography images for differential diagnosis between benign and malignant solid breast masses
Pattern classification of ShearWaveâ„¢ Elastography images for differential diagnosis between benign and malignant solid breast masses
复制标题
DOI：
10.1258/ar.2011.110276
复制
发表时间：
2011-12-01
期刊：
影响因子：
1.3
通讯作者：
Fukuma, Eisuke
中科院分区：
文献类型：
--
作者：
Tozaki, Mitsuhiro;Fukuma, Eisuke
背景：SearWave(TM)弹性成像(SWE)提供了组织硬度的定量测量，并可能改善乳腺肿块的特征。目的：评价分型和杨氏模量法在乳腺实性肿块良恶性鉴别诊断中的价值。材料与方法：对96例100个乳腺实性肿块进行超声检查后的组织标本进行分析。我们尝试基于SWE图像创建视觉模式分类。结果：视觉评价可将病变图像分为4种类型：无发现(均匀编码蓝色；图案1)、垂直条纹伪影(图案2)、病变边缘局部着色区域(图案3)和病变内部不均匀着色区域(图案4)。型17个，型2 14个，型3 20个，型4 49个。当模式1、2为良性，模式3、4为恶性时，敏感性为91.3%(63/69)，特异性为80.6%(25/31)。良、恶性病变的平均杨氏模量值分别为42kPa和146kpa(P<；0.0001)。但在模式4中，良性病变的杨氏模数(50kPa.)低于恶性肿块的最小杨氏模数(61kPa.结论：视觉分型和增加杨氏模数测量有助于提高乳腺实性肿块的定性
Background: ShearWave (TM) Elastography (SWE) provides a quantitative measurement of tissue stiffness and may improve characterization of breast masses. However, the significance of Young's modulus measurements and appropriate SWE evaluation criteria has not been established yet.Purpose: To assess the usefulness of the pattern classification and Young's modulus measurements in the differential diagnosis between benign and malignant solid breast masses.Material and Methods: Ninety-six patients (age range 18-84 years, mean 54 years) with 100 solid breast masses who underwent tissue sampling after a US examination were analyzed. We tried to create a visual pattern classification based on the SWE images. After classifying the visual patterns, the Young's modulus of the lesions was measured in every case.Results: It was possible to classify the images into four patterns by the visual evaluation: no findings (coded blue homogeneously; Pattern 1), vertical stripe pattern artifacts (Pattern 2), a localized colored area at the margin of the lesion (Pattern 3), and heterogeneously colored areas in the interior of the lesion (Pattern 4). There were 17 Pattern 1 lesions, 14 Pattern 2 lesions, 20 Pattern 3 lesions, and 49 Pattern 4 lesions. When Patterns 1 and 2 were assumed to be benign, and Patterns 3 and 4 were assumed to be malignant, the sensitivity and specificity were 91.3% (63/69) and 80.6% (25/31), respectively. The mean Young's modulus measurements of the benign and the malignant lesions were 42 kPa and 146 kPa, respectively (P < 0.0001). No significant differences were found between benign and malignant lesions in Pattern 3. In Pattern 4, however, the Young's modulus of the benign lesions (50 kPa) was lower than the smallest Young's modulus of malignant lesions (61 kPa).Conclusion: The visual pattern classification and adding Young's modulus measurements may improve characterization of solid breast masses.