汤 伟, 李瑞敏, 高 毅. A comparative study of diagnostic performance between digital breast tomosynthesis and conventional imaging methods[J]. China Oncology, 2017, 27(6): 487-495.
汤 伟, 李瑞敏, 高 毅. A comparative study of diagnostic performance between digital breast tomosynthesis and conventional imaging methods[J]. China Oncology, 2017, 27(6): 487-495. DOI: 10.19401/j.cnki.1007-3639.2017.06.015.
A comparative study of diagnostic performance between digital breast tomosynthesis and conventional imaging methods
背景与目的:乳腺癌已成为女性最常见的恶性肿瘤,乳腺X线摄影新技术的出现对乳腺癌诊断起重要作用;该研究旨在对比数字乳腺断层融合X线摄影(digital breast tomosynthesis,DBT)与常规影像学检查诊断乳腺良、恶性病变的效能。方法:收集2015年3月—2015年12月在复旦大学附属肿瘤医院就诊且经临床或超声检查怀疑乳腺病变的227例患者。所有患者均行B超、全屏数字乳腺X线摄影(full-field digital mammography,FFDM)、DBT和MRI检查。高年资影像医师双盲阅片,参照2013版BI-RADS分类标准,以病理结果为金标准,采用受试者工作特征(operating characteristic,ROC)曲线分析各种检查方法的诊断效能,并利用Z检验比较DBT与其他检查方法的统计学差异。结果:30例患者因图像不符合诊断要求而排除,剩余197例患者共发现205个病灶,其中良性病灶73例,恶性病灶132例。B超、FFDM、DBT、DBT+FFDM和MRI基于BI-RADS分类诊断乳腺良、恶性病变的ROC曲线的曲线下面积(area under the curve,AUC)分别为0.830 8、0.859 2、0.9167、0.919 8和0.935 4;两两比较结果显示,DBT的AUC高于B超和FFDM,且差异均有统计学意义(比B超相比,Z=7.36,P=0.006 7;与FFDM相比,Z=4.89,P=0.027 1);DBT与MRI、DBT+FFDM的AUC差异无统计学意义(与MRI相比,Z=0.02,P=0.900 2;与DBT+FFDM相比,Z=0.69,P=0.404 8)。结论:DBT较B超、FFDM能显著提高乳腺良、恶性病灶的诊断效能;并与MRI诊断效能相当。DBT作为乳腺X摄影新技术,具有良好的应用前景。
Abstract
Background and purpose: Breast cancer is the most common malignancy in women. The new technology of mammography is helpful in breast cancer diagnosis. This study aimed to compare the efficacy of digital breast tomosynthesis (DBT) with conventional imaging methods in the diagnosis of benign and malignant breast lesions. Methods: During the period from Mar. 2015 to Dec. 2015
227 patients with suspected lesions (by palpation or sonography) underwent further imaging exam in our hospital. The sonography
full-field digital mammography (FFDM)
DBT and breast MRI were performed on all the patients. A double-blind assessment was carried out according to BI-RADS (version 2013) by experienced radiologists. The receiver operating characteristic (ROC) curve was used to evaluate the diagnostic efficiency of all methods
referring to the pathological data as the golden standard; the difference in the efficiency of DBT from the other methods was determined by Z-test. Results: Thirty patients were excluded for the unsatisfactory images
and 205 lesions (132 malignant and 73 benign lesions) were detected in the remaining 197 patients. Area under the curve (AUC) of sonography
FFDM
DBT
DBT+FFDM and MRI based on the BI-RADS were 0.830 8
0.859 2
0.916 7
0.919 8
and 0.935 4
respectively. The AUC of DBT was significantly higher than those of sonography (Z=7.36
P=0.006 7) and FFDM (Z=4.89
P=0.027 1)
while there was no significantly difference between DBT and MRI (Z=0.02
P=0.900 2) or FFDM+DBT (Z=0.69
P=0.404 8). Conclusion: DBT could significantly improve the diagnostic performance for breast lesions compared with sonography and FFDM
providing a comparable efficiency to MRI. As a new mammography technology
Accuracy of tumour size assessment in the preoperative staging of breast lesions: a comparative study between digital breast tomosynthesis and conventional imaging methods
Value of MRI T2-weighted image texture analysis in evaluating the microvascular invasion for hepatocellular carcinoma
Accuracy of MRI for predicting shrinkage modes of primary breast tumor following neoadjuvant chemotherapy with three-dimensional reconstruction technique