Search

The classification of the breast imaging-reporting and data system (BI-RADS) based on magnetic resonance imaging (MRI) refers to the classification of the degree of lesions according to the image signs of lesions, which is usually subjective. Moreover, the benign and malignant lesions of BI-RADS 3-5 are overlapping, which is prone to unnecessary invasive treatment due to high diagnostic categories in clinical diagnosis. To address these problems, this research applied radiomics for feature extraction and fusion of T1-weighted (T1W) and dynamic contrast-enhanced (DEC) MRI. The least absolute shrinkage and selection operator (LASSO) algorithm was used to screen out the optimal feature collection of each type of MR image. Support vector machine (SVM), random forest (RF), K-nearest neighbour (KNN) and logistic regression (LR) algorithms were applied for BI-RADS 3-5 classification, based on which the benign and malignant lesions were further classified. The results showed that the classification accuracy of breast BI-RADS 3-5 by four radiomics models based on feature fusion was 81.25%, 87.50%, 78.38%, and 81.25%, respectively. Their accuracy in distinguishing the benign and malignant breast lesions was 90.91%, 93.55%, 92.73%, and 94.55%, respectively. This indicates that the combination of radiomics and machine learning correlation algorithm has a good effect on breast MRI BI-RADS classification and benign and malignant differentiation, and feature fusion can further improve the accuracy of classification prediction.

Objective To explore the value of color pixel density （CPD） in differential diagnosis of Breast Imaging Reporting and Data System （BI-RADS） 4 breast masses by high-frequency ultrasound. Methods Relevant data of 81 pathologically-confirmed BI-RADS 4 breast masses were collected. All the masses were examined with color Super Microvascular Imaging （cSMI） before operation. CPD was calculated with the integrated software in the ultrasonic diagnostic instrument. Pathological biopsy results were regarded as the gold standard. The area under the receiver operating characteristic curve （ROC AUC） was created to evaluate the diagnostic efficiency of CPD. BI-RADS categories were corrected by CPD, and the sensitivity, specificity, accuracy, positive and negative predictive values of diagnosing breast masses were compared before and after correction. Results The ROC AUC of CPD was 0.854 （P < 0.05）. According to the cut-off values （9.55%） from ROC curve, the sensitivity, specificity, accuracy, positive and negative predictive values of CPD for differential diagnosis were 73%, 83%, 78%, 85% and 71%, respectively. After corrected by CPD, the sensitivity, specificity, accuracy, positive and negative predictive values of the BI-RADS were 87%, 83%, 85%, 87% and 83%, slightly higher than those before correction （80%, 67%, 74%, 75% and 73%）. The classification accuracy was improved by 23% （P < 0.05）. Conclusions CPD in high-frequency ultrasound is of certain value in the differential diagnosis of breast masses. In clinical practice, CPD can be employed to correct BI-RADS classification, aiming to improve the diagnostic efficiency of BI-RADS 4 breast masses.

Objective To investigate the application value of conventional ultrasound typing and contrastenhanced ultrasound (CEUS) in non-mass breast like lesions (NMLs). Methods A total of 50 patients (50 lesions) who were pathologically confirmed to have breast NMLs in Tianjin Fifth Central Hospital. The corresponding relationship between conventional ultrasound typing of NMLs and BI-RADS classification was analyzed, and CEUS was performed on breast NMLs to compare the diagnostic value of combined BIRADS and CEUS. Results The NMLs conventional ultrasound classified Ⅰ a and Ⅱ a lesions corresponded to BI-RADS classification 4a; Ⅲ and Ⅳ lesions corresponded to BI-RADS classification 4b; Ⅰ b and Ⅱ b lesions corresponded to BI-RADS classification 4c. A statistically significant correlation was observed between the expansion of lesion range and the presence of “crab foot sign” in benign and malignant lesions after the enhancement (P<0.05). Pearson correlation analysis showed that enlarged lesion area, “crab foot sign,” peak intensity, and area under the curve were associated with malignant breast NMLs. The diagnostic value of CEUS combined with BI-RADS was higher than that of conventional ultrasound BI-RADS. Conclusion NMLs conventional ultrasonographic classifications I a and Ⅱ a have the lowest malignant risk, whereas classifications Ⅰ b and Ⅱ b have the highest value. The expansion of lesion range and the presence of “crab foot sign” after enhancement are helpful for rapid differentiation of benign and malignant NMLs. [ABSTRACT FROM AUTHOR]

Objective To analyze the feature of breast complex cystic masses and to classify it at ultrasonography (US), which applied to the Breast Imaging Reporting and Data System (BI-RADS) categories 4a to 4c with pathological results as the golden standards. Methods The ultrasonographic data and clinical features of 78 patients with complex cystic masses confirmed by pathology in Cancer Hospital from July 2014 to June 2017 were retrospectively reviewed. The complex cystic breast masses were divided into four classes on the basis of their US features: type 1 [thick wall and (or) thick septa (> 0.5 mm)], type 2 (one or more intra-cystic masses), type 3 (mixed cystic and solid components with cystic components more than 50%) and type 4 (mixed cystic and solid components with solid components more than 50%). Positive values (PPVs) were calculated for each type. Multiple linear regression analysis was used to analyze the ultrasonographic features of the masses (lesion size, margins, blood flow resistance index, calcification, and axillary lymph nodes, etc.) with malignant correlation. Results There were 81 lesions in 78 patients. Among the 81 masses based on US appearance, 14 (17.3%) were classified as type Ⅰ, 18 (22.2%) as type Ⅱ, 18 (22.2%) as type Ⅲ, and 31 (38.3%) as type Ⅳ. The positive predictive values of the malignant lesions of type Ⅰ, type Ⅱ, Ⅲ and Ⅳ were 7.1%, 16.7%, 61.1% and 48.3%, respectively (P=0.040). In all the 81 masses, 14 were BI-RADS categories 4a, 18 were BI-RADS categories 4b and 49 were BI-RADS categories 4c. Masses with maximum diameter equal to or larger than 2.0 cm, unclear margins, RI≥0.7 and presence of abnormal axillary nodes assessment had a high probability of malignancy (P=0.030, 0.038, <0.001 and 0.025, respectively). Conclusion Ultrasound typing is helpful for differentiating benign and malignant breast complex cysts and classifying BI-AIDS 4a to 4c, thus providing clearer treatment for clinical practice. [ABSTRACT FROM AUTHOR]

【目的】评估2013版乳腺影像报告及资料系统（BI-RADS）分类在临床诊断性超声检查中的准确性。【方法】回 顾中山大学孙逸仙纪念医院2014年6月-2016年6月因临床症状至乳腺中心行乳腺超声检查的成年女性病例，筛选最终分 类3-5类且接受活检有最终病理诊断者进行分析，计算各类的阳性预测值（PPV），与美国放射学院（ACR）所提供的参考范 围进行比较分析。【结果】共有来自4428例女性病人的4 428个病灶纳入本研究。1198（27.1%）例被诊断为病理恶/交界性； 余3230（72.9%）例被诊断为病理良性。各BI-RADS分类的PPV均在2013年ACR所推荐的参考范围内。恶性病灶中，随着 BI-RADS分类的升高，腋窝淋巴结阳性的比例增加。病理学诊断为浸润性导管癌或浸润性小叶癌的病例，其在BI-RADS分 类中的分布呈现出从3类到5类的渐增趋势。【结论】本研究表明，2013版超声BI-RADS分类在临床工作中有着良好的诊断 效能，但仍需进一步的多中心研究来证实。

【目的】 初步评估声速测算技术与乳腺影像学报告及数据系统(BI-RADS)诊断标准相结合对乳腺肿物良恶性的诊断价值?【方法】回顾性分析经手术病理证实的75例女性患者共99个乳腺肿物?先进行常规超声检查,并用BI-RADS诊断标准判断其良恶性;接着应用声速匹配技术测量肿物的声速匹配值(ZSI),进而得到病灶的声速(SV);以病理结果作为金标准,构建受试者的工作特征曲线,比较两种方法以及二者结合后的诊断价值? 【结果】 SV判断乳腺良恶性肿物的最佳临界点为1 561 m/s,ROC曲线下面积为0.842;BI-RADS诊断标准ROC曲线下面积为0.818,与SV之间的差异没有统计学意义(P > 0.05)?BI-RADS诊断标准的敏感性高于SV,特异性?准确性较SV低(P < 0.05)?两种方法结合后诊断的特异性及准确性均较BI-RADS时增高(P < 0.05)?【结论】 BI-RADS诊断标准与SV相结合对乳腺肿物良恶性的判断有一定的帮助?