Daniel B L, Yen Y F, Glover G H, Ikeda D M, Birdwell R L, Sawyer-Glover A M, Black J W, Plevritis S K, Jeffrey S S, Herfkens R J
Dept of Radiology, Stanford University School of Medicine, Lucas Magnetic Resonance Imaging Center, CA 94305-5488, USA.
Radiology. 1998 Nov;209(2):499-509. doi: 10.1148/radiology.209.2.9807580.
To compare various subjective, empiric, and pharmacokinetic methods for interpreting findings at dynamic magnetic resonance (MR) imaging of the breast.
Dynamic spiral breast MR imaging was performed in 52 women suspected of having or with known breast disease. Gadolinium-enhanced images were obtained at 12 locations through the whole breast every 7.8 seconds for 8.5 minutes after bolus injection of contrast material. Time-signal intensity curves from regions of interest corresponding to 57 pathologically proved lesions were analyzed by means of a two-compartment pharmacokinetic model, and the diagnostic performance of various parameters was analyzed.
Findings included invasive carcinoma in 17 patients, isolated ductal carcinoma in situ (DCIS) in six, and benign lesions in 34. Although some overlap between carcinomas and benign diagnoses was noted for all parameters, receiver operating characteristic analysis indicated that the exchange rate constant had the greatest overall ability to discriminate benign and malignant disease. The elimination rate constant and washout were the most specific parameters. The exchange rate constant, wash-in, and extrapolation point were the most sensitive parameters. DCIS was not consistently distinguished from benign disease with any method.
Dynamic spiral breast MR imaging proved an excellent method with which to collect contrast enhancement data rapidly enough that accurate comparisons can be made between many analytic methods.
比较各种主观、经验性和药代动力学方法,以解读乳腺动态磁共振(MR)成像的结果。
对52名怀疑患有或已知患有乳腺疾病的女性进行了乳腺动态螺旋MR成像检查。在静脉注射造影剂后,每7.8秒在整个乳腺的12个位置获取钆增强图像,持续8.5分钟。利用双室药代动力学模型分析了对应于57个经病理证实病变区域的时间-信号强度曲线,并分析了各种参数的诊断性能。
结果包括17例浸润性癌、6例孤立性导管原位癌(DCIS)和34例良性病变。尽管所有参数在癌和良性诊断之间均存在一定重叠,但受试者工作特征分析表明,交换率常数总体上最能区分良性和恶性疾病。消除率常数和廓清率是最具特异性的参数。交换率常数、流入率和外推点是最敏感的参数。用任何方法都不能始终如一地将DCIS与良性疾病区分开来。
乳腺动态螺旋MR成像被证明是一种很好的方法,能够快速收集对比增强数据,从而可以在多种分析方法之间进行准确比较。