BI-RADS 3-5类乳腺微钙化灶X线与MRI影像学表现对比研究

发布时间：2018-05-27 03:10

本文选题：乳腺X线摄影 + 磁共振　；参考：《河北医科大学》2017年硕士论文

【摘要】：目的:本课题旨在于通过分析乳腺X线检出的BI-RADS分类3-5类微钙化病变的形态和分布特点以及对应部位的MRI表现特点,对比此两种检查手段对此类微钙化病变的诊断效能,从而指导选择合理的检查方法并提高早期癌的诊断率,对临床制定有效的检查方法及诊治方案提供帮助。方法:回顾性分析2015年3月-2016年12月在河北医科大学第四医院行乳腺X线摄影检查及乳腺DCE-MRI检查检出的BI-RADS3-5类微钙化性病变94例。依据美国放射学会(ACR)提出的乳腺影像报告和数据统计系统(BI-RADS)标准对X线和MRI图像进行分析和评估。X线图像的分析包括钙化的形态和分布,MRI图像分析依据BI-RADS-MRI标准:将病变分为肿块样强化和非肿块样强化,肿块样强化分析其肿块的形态、边缘及内部强化特征,非肿块样强化主要分析其形态学分布。依据病变大小,选取感兴趣区(ROI)绘制病变时间-信号强度曲线(TIC),并与相应的组织病理学结果进行对比分析。采用SPSS21软件进行统计学分析,分析内容包括:(1)乳腺X线摄影BI-RADS3-5类微钙化的形态和分布表现的阳性预测值;(2)乳腺X线摄影BI-RADS 3,4,5类钙化性病变的阳性预测值;(3)乳腺X线摄影BI-RADS 3-5类微钙化灶相对应的MRI强化表现及阳性预测值;(4)良、恶性病变动态增强曲线类型比较;(5)以BI-RADS 3,4a类病变为高度可能良性标准,以BI-RADS 4b,4c及5类为高度可能恶性标准,分别分析X线和MRI对BI-RADS3-5类微钙化的病变诊断的敏感性(sensibility)、特异性(specificity)、阳性预测值(PPV)、阴性预测值(NPV)和准确性(accuracy);(6)分析不同乳腺腺体类型的X线和MRI检查的敏感性(sensibility)和特异性(specificity)。结果:1病理结果94例X线表现为BI-RADS 3-5类微钙化性病变中,经术后病理证实86例。其中55例恶性,病理结果分别为:乳腺粘液腺癌2例、浸润性导管癌12例、低分化腺癌2例、浸润癌29例、导管原位癌1例、乳腺导管上皮不典型增生伴癌变1例、乳头Peget病伴大导管癌1例、非浸润性导管内癌1例、浸润性-小叶癌混合癌2例、重度不典型增生伴癌变1例、高核级导管内癌2例、高核级导管内癌peget病1例。良性病变31例,病理结果分别为:乳腺腺病伴上皮不典型增生1例、乳腺腺病10例、乳腺腺病伴柱状细胞改变及平坦上皮增生2例、正常腺体组织4例、乳腺腺病伴纤维脂肪组织3例、乳腺腺病伴导管扩张5例、柱状上皮不典型增生1例、乳腺导管扩张伴脂肪坏死及炎性改变1例、乳腺腺病伴纤维瘤1例、乳腺腺病伴钙化1例。另外8例经1年以上X线检查随访复查,经过三个月、六个月及一年复查,对比原片病变未见明显变化,故考虑为良性病变。2 X线形态学表现、分布与阳性预测值:94例病变中,X线上BI-RADS3类微钙化形态表现以圆点状为主(83%),分布以弥散和成簇为主(均41%);4a类微钙化形态表现以无定形模糊为主(56%),分布以成簇分布为主(64%);4b类微钙化形态以粗糙不均质为主(46%),分布以成簇分布为主(73%);4c类微钙化形态以细微多形性为主(67%),分布以段样分布为主(48%);5类微钙化形态以分枝状、蠕虫样为主(60%),分布以段样分布为主(45%)。钙化形态呈圆点状其PPV值为11.1%,无定形模糊其PPV值38.1%,粗糙不均质其PPV值为18.1%,细微多形性其PPV值为96.1%,分支状、蠕虫样其PPV值为100%;钙化分布呈弥散分布其PPV值为11.1%;区域分布其PPV值为60%;线样、导管样分布PPV值为70%;段样分布PPV值为82%;成簇分布PPV值为54%。通过以上数据对微钙化的形态及分布分析得出,高度恶性的钙化形态为细微多形性及分支状、蠕虫样状钙化,二者均高于点状、模糊不定形及粗糙不均质钙化,关于钙化的分布特点,段样分布的阳性预测值(PPV)均高于其他类型分布。3 MRI影像学表现与阳性预测值:(1)在MRI上BI-RADS3-5类微钙化性病变相对应的肿块型强化表现:其中BI-RADS 3类病变多表现为边缘光滑的肿块(100%),内部强化特点为环形强化(100%);BI-RADS 4a类病变多表现为边缘光滑(50%)或呈浅分叶的肿块(50%),内部强化特点为均匀强化(100%);BI-RADS 4b类病变多表现为边缘不光滑(60%)圆型、卵圆型的肿块(80%),内部强化特点为不均匀强化(80%);BI-RADS4c类钙化性病变多表现为边缘呈毛刺(40%)的圆型或椭圆形肿块(81.3%),内部强化特点多呈不均匀强化(60%);BI-RADS 5类病变多表现为边缘不光滑(44.4%)或呈毛刺状(44.4%)的圆型或类圆形肿块(52.5%),内部强化特点多为不均匀强化(100%)。肿块呈圆型或椭圆形PPV值为96.4%,分叶状PPV值为86.7%,不规则型PPV值为100%;肿块边缘光滑PPV值为0.0%,不光滑PPV值为100%,毛刺状PPV值为100%,浅分叶为90.9%。病变内部强化特点,均匀强化PPV值为66.7%,不均匀强化PPV值为96.8%,环形强化PPV值为0.0%,间隔强化PPV值为83.3%;(2)BI-RADS 3-5类钙化性病变表现为非肿块型病变强化表现:其中3类及4a类病变多表现为局灶样强化(63%)、(42.9%);4b类多表现为线样、导管样(66.7%);4c类病变多表现为线样、导管样及段样强化(50%);5类病变多表现为段样强化(66.7%)。在非肿块型强化类型中,点状强化PPV值为60%,局灶样强化PPV值为12.5%,线样、导管样强化PPV值为85.7%,段样强化PPV值为85.7%,区域样强化PPV值为25%,斑片样强化的PPV值为0.0%。4结果分析(1)呈肿块样强化的病变中,高度恶性的强化形式为边缘不光滑的不规则肿块,内部强化呈不均匀强化;非肿块样强化的病变中,高度恶性的强化形式为线样、导管样强化及段样强化。(2)在MRI动态增强扫描中,良性病变MRI强化曲线多表现为Ⅰ型(80.6%),恶性病变MRI强化曲线多表现为Ⅲ型(50.9%),良、恶性病变在MRI动态增强强化曲线之间差异具有统计学意义(P0.05)。(3)本研究中X线诊断为3类钙化性病变的PPV为0.0%,4a类钙化性病变的PPV值为32%,4b类钙化性病变的PPV值为63.7%,4c类钙化性病变的PPV值为95.2%,5类钙化性病变的PPV值为100%。(4)X线对BI-RADS 3-5类微钙化性病变诊断的敏感性、特异性、准确率、阳性预测值、阴性预测值、分别为:92.3%,38%,69.7%,67.6%,78%。(5)MRI对BI-RADS 3-5类微钙化性病变诊断的敏感性、特异性、准确率、阳性预测值、阴性预测值、分别为:92.3%,73%,84.3%,82.8%,87%。5乳腺腺体类型与钙化显示的关系关于钙化性病变在不同乳腺腺体类型中X线和MRI检查的敏感性和特异性,其中脂肪型腺体(a型)和散在纤维腺体型(b型)中X线和MRI检查敏感性和特异性均为100%;不均质型腺体(c型)中X线检查的敏感性为91%,特异性为31%,MRI检查的敏感性为91%,特异性为69.7%。两种检查方法对于腺体类型呈不均质致密型的BI-RADS 3-5类钙化性病变其敏感性相同,而MRI检查的特异性远高于乳腺X线检查。对于极度致密型(d型)腺体X线检查的敏感性为100%,特异性为33%,MRI检查的敏感性和特异性均为100%。因此对于不均匀致密型腺体及致密型腺体MRI检查占优势。结论:1 MRI难以显示乳腺的微小钙化,但能显示对应微小钙化区域病变的组织形态学及血流动力学改变。2 MRI检查的特异性、准确率、阳性预测值、阴性预测值均高于X线检查。3乳腺钙化性病变,特别是中间性(良、恶性鉴别困难)钙化是X线上诊断难点,虽然MRI不能直观地显示钙化,但对钙化性病变的良、恶性鉴别诊断具有较高的价值,可在一定程度上避免X线诊断上对钙化性病变的高估或低估4 MRI对典型恶微钙化灶的判定没有优势。此类钙化性病变若在MRI表现为线样导管样强化及节段性强化特征时,应行活检明确病理结果。5动态增强曲线对于含成簇钙化病变具有一定的诊断效能。6在MRI上表现为弥漫性强化的病灶仍存在恶性病灶的可能性。钙化性病变在MRI上无强化时,要求随访复查。由于X线摄影检查对微钙化病灶的高敏感性及简便易行是乳腺癌筛查的首选方法,MRI虽然难以显示乳腺的微小钙化,但根据其极高的分辨率、动态强化特点能很好的显示对应病变微钙化区域的组织形态学及血流动力学改变,对乳腺早期病灶的显示及诊断具有重要的意义。
[Abstract]:Objective: the purpose of this study is to analyze the morphological and distribution characteristics of 3-5 types of microcalcification in the BI-RADS classification of mammary glands and the characteristics of the MRI manifestations of the corresponding sites, and compare the diagnostic efficacy of these two methods for this kind of microcalcification, so as to guide the selection of reasonable examination methods and to improve the diagnosis rate of early cancer. Methods: a retrospective analysis of 94 cases of BI-RADS3-5 microcalcifications found in the fourth hospital of Hebei Medical University in December, March 2015, and breast DCE-MRI examination in fourth hospital of Hebei Medical University. The breast imaging report and data statistics based on the American Institute of Radiology (ACR) The X-ray and MRI images were analyzed and evaluated by the system (BI-RADS) standard. The analysis of X - ray images included the morphology and distribution of calcification. The MRI image analysis was based on the BI-RADS-MRI standard: the lesions were divided into lump like enhancement and non lump like enhancement. The mass sample enhancement was used to analyze the shape of the lumps, the edge and the internal intensification features, and the non lump like strengthening main points. According to the size of the lesion, we selected the region of interest (ROI) to draw the time signal intensity curve (TIC), and compared with the corresponding histopathological results. The statistical analysis was carried out by SPSS21 software. The contents included: (1) the positive precondition of the morphology and distribution of BI-RADS3-5 type microcalcification in mammography. Test values; (2) the positive predictive value of BI-RADS 3,4,5 calcified lesions in mammography; (3) MRI enhancement and positive predictive value of BI-RADS 3-5 microcalcifications in mammography; (4) good, malignant lesions dynamic enhancement curve type comparison; (5) BI-RADS 3,4 a lesion as a highly possible benign standard, BI-RADS 4b, 4C and 5 groups High possible malignant criteria, the sensitivity (sensibility), specificity (specificity), positive predictive value (PPV), negative predictive value (NPV) and accuracy (accuracy) of X-ray and MRI for the diagnosis of BI-RADS3-5 microcalcification (accuracy); (6) sensitivity (sensibility) and specificity (specificI) for the analysis of different mammary gland types (specificI) and specificity (specificI). Results: (TY) results: 1 the pathological results of 94 cases were found in the BI-RADS 3-5 type of microcalcification, and 86 cases were confirmed by postoperative pathology. 55 cases were malignant. The pathological results were 2 cases of mucous adenocarcinoma of the breast, 12 cases of invasive ductal carcinoma, 2 cases of low differentiated adenocarcinoma, 29 cases of invasive carcinoma, 1 cases of ductal carcinoma and 1 cases of ductal epithelial hyperplasia with canceration. 1 cases of head Peget disease, 1 cases of non invasive intraductal carcinoma, 2 cases of mixed carcinoma of infiltrative lobular carcinoma, 1 cases of severe atypical hyperplasia with canceration, 2 cases of high nuclear grade intraductal carcinoma, 1 cases of high nuclear grade intraductal carcinoma, 31 cases of benign lesions, 1 cases of atypical hyperplasia of mammary glands, 10 cases of mammary gland adenosis, mammary gland adenopathy, 2 cases of columnar cell change and flat epithelial hyperplasia, 4 cases of normal gland tissue, 3 cases of mammary gland disease with fibrous adipose tissue, 5 cases of mammary gland disease with catheter dilatation, 1 cases of columnar epitheliosis, 1 cases of mammary duct dilatation with necrosis and inflammatory changes, 1 cases of breast adenosis with fibrous tumor, 1 cases of mammary gland disease with calcification, and 8 cases after 1 years. After three months, six months and one year reexamination, there was no obvious change in the original lesion, so it was considered as the.2 X - ray morphology, distribution and positive predictive value: in 94 cases, the X-ray microcalcification was mainly (83%) on X - ray microcalcification, and distributed to diffuse and cluster mainly (41%); 4A The morphology of microcalcification is dominated by amorphous fuzzy (56%) and cluster distribution mainly (64%); 4B microcalcification is dominated by rough heterogeneity (46%), and the distribution is dominated by cluster distribution (73%); 4C microcalcification is mainly (67%) with microcalcification (67%), and the 5 type of microcalcification is branched and worm like Main (60%), distribution mainly in segment distribution (45%). Calcification is a circular shape with a PPV value of 11.1%, an amorphous fuzzy PPV value of 38.1%, a rough and uneven PPV value of 18.1%, a fine pleomorphic PPV value of 96.1%, a branch, and a worm like PPV value of 100%; the distribution of calcification is a dispersion and a PPV value of 11.1%; the distribution of its PPV value is 60%; line sample, guide The PPV value of the tube distribution is 70%, the PPV value of the segment distribution is 82%, and the cluster distribution PPV value is that 54%. through the above data analyses the morphology and distribution of microcalcification. The highly malignant calcification forms are fine polymorphic and branched, and worm like calcification, two are higher than the point, fuzzy and rough uneven calcification, and the distribution of calcification is special. Point, the positive predictive value (PPV) of the segment distribution (PPV) was higher than that of other types of.3 MRI imaging and positive predictive value: (1) the enhanced manifestation of the lump type corresponding to the BI-RADS3-5 microcalcification in MRI, among which, the BI-RADS 3 lesions were mostly marginal smooth masses (100%), and the internal intensification was characterized by annular enhancement (100%); BI-RADS 4a. Most of the lesions showed a smooth (50%) or shallow lobular mass (50%), and the internal enhancement was homogeneous (100%), and BI-RADS 4B lesions were characterized by irregular edge (60%) round, oval type mass (80%), inhomogeneous enhancement (80%), and BI-RADS4c calcified lesions as a circle of marginal burr (40%). Type or oval mass (81.3%) with inhomogeneous enhancement (60%); BI-RADS 5 lesions were mostly circular or circular mass (52.5%) with unsmooth edge (44.4%) or burr like (44.4%). The internal strengthening features were mostly uneven enhancement (100%). The mass was round or oval PPV value 96.4%, and the lobulated PPV value was 86.7%, no The regular PPV value was 100%, the smooth PPV value of the lump edge was 0%, the unsmooth PPV value was 100%, the burr shape PPV value was 100%, the superficial lobule was the internal strengthening characteristic of the 90.9%. lesion, the uniform strengthening PPV value was 66.7%, the unevenly strengthened PPV value was 96.8%, the annular enhanced PPV value was 0%, the interval strengthened PPV value 83.3%, and (2) BI-RADS 3-5 calcified pathological changes showed non The enhanced manifestation of lump type lesions: 3 and 4A lesions were characterized by focal enhancement (63%) and (42.9%); 4B was characterized by line samples and catheter like (66.7%); class 4C lesions were characterized by lines, catheter like and segmental enhancement (50%); the 5 types of lesions were characterized by segmental intensity (66.7%). In non lump type intensification types, punctate PPV value was 60%, focal lesion was found. The PPV value was 12.5%, the line sample, the catheter like enhancement PPV value was 85.7%, the segment enhancement PPV value was 85.7%, the regional sample enhanced PPV value was 25%, the PPV value of the patch sample intensification was 0.0%.4 result analysis (1) in the lump like enhancement lesion, the highly malignant enhancement form was the irregular edge not smooth mass, the internal strengthening was inhomogeneous strengthening; non - mass In the lesion of sample enhancement, the highly malignant form of enhancement was line sample, catheter like enhancement and segmental enhancement. (2) in MRI dynamic enhanced scan, the MRI enhancement curve of benign lesions was type I (80.6%), and the MRI intensification curve of malignant lesions was type III (50.9%), and the difference between good and malignant lesions was statistically significant in the dynamic enhancement curve of MRI. (P0.05) (3) (3) the X-ray diagnosis of 3 types of calcified lesions in this study was 0%, the PPV value of 4A calcified lesions was 32%, the PPV value of 4B calcified lesions was 63.7%, the PPV value of 4C calcified lesions was 95.2%, and the PPV value of 5 calcified lesions was 100%. (4) in the sensitivity and specificity of the diagnosis of BI-RADS 3-5 microcalcifications. Accuracy, positive predictive value, negative predictive value: 92.3%, 38%, 69.7%, 67.6%, 78%. (5) MRI for the diagnosis of BI-RADS 3-5 microcalcification, specificity, accuracy, positive predictive value, negative predictive value, respectively: 92.3%, 73%, 84.3%, 82.8%, 87%.5 mammary gland type and calcification display in different calcification lesions in different The sensitivity and specificity of X - ray and MRI examination in the mammary gland type were 100%. The sensitivity and specificity of the X - ray and MRI examination were 100% in the fatty gland (type A) and in the fibrous gland type (type b). The sensitivity of the X - ray examination in the heterogeneous gland (type C) was 91%, the specificity was 31%, the sensitivity of the MRI examination was 91%, and the specificity was 69.7%. two kinds of examination. The sensitivity of the method was the same to the BI-RADS 3-5 type calcified lesions of the inhomogeneous and dense glandular type, but the specificity of the MRI examination was much higher than that of the mammography. The sensitivity of the X-ray examination of the highly dense (D type) gland was 100%, the specificity was 33%, the sensitivity and specificity of the MRI examination were 100%., therefore, for the inhomogeneous density. Conclusion: 1 MRI can not show small calcification of the mammary gland, but it is difficult to show the Microcalcification of the breast, but it can show the specificity, accuracy, positive predictive value and negative predictive value of the.2 MRI examination for the changes of the lesions in the small calcification region, and the negative predictive values are higher than those of the X-ray examination of the.3 mammary calcification, especially in the middle of the.2. Intersex (benign, malignant differential difficulty) calcification is the difficult point in X-ray diagnosis. Although MRI can not display calcification intuitively, it is of high value for the differential diagnosis of calcified lesions. To some extent, it can avoid the overestimation or underestimation of calcified lesions by X - ray diagnosis. The 4 MRI has no advantage on the determination of typical malignant microcalcifications. Calcified lesions, if MRI appears as a line like catheter like enhancement and segmental enhancement, should undergo biopsy clear pathological results, the.5 dynamic enhancement curve has a certain diagnostic efficacy for cluster calcification, and the possibility of malignant Focus still exists in the diffuse enhancement of.6 on MRI. Calcified lesions have no enhancement on MRI. Gao Min's susceptibility to microcalcification is the first choice for screening of breast cancer. Although MRI is difficult to show small calcification of the breast, the dynamic strengthening features can show the histomorphology and hemodynamics of the corresponding microcalcification region, according to its high resolution. The change of learning is of great significance for the diagnosis and diagnosis of early breast lesions.
【学位授予单位】：河北医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R737.9;R445.2;R730.44

【参考文献】