摘要
研究提出一种新的定量分析临床常规采集的子宫动态增强磁共振成像数据的MRR-FCM方法,方法结合了提出的改进的参考区域模型用于估计组织的T1(0),模糊聚类分析方法和考虑血浆容积分数的参考区域模型。MRR-FCM方法包括5个步骤:1)使用模糊聚类分析方法自动分割子宫的图像区域;2)使用改进的参考区域模型估计基准的T1(0)值;3)把信号强度转化成对比剂浓度;4)使用模糊聚类分析方法自动地把对比剂浓度曲线分成预定数目的类;5)使用考虑血浆容积分数的参考区域模型逐像素估计子宫区域内的药物代谢动力学参数。通过用MRR-FCM方法分析了6位经病理证实的宫颈癌病人的图像数据,以验证该方法在分析临床动态增强磁共振成像数据的有效性和可行性。MRR-FCM方法给出的在体定量功能参数Ktrans和ve,揭示了宫颈癌病灶内部结构的异质性。Ktrans在宫颈癌病灶和正常子宫组织之间有显著性差异(p<0.01)。MRR-FCM方法能够定量分析在典型的临床环境下采集的子宫动态增强磁共振成像数据,并有可能扩展到其他器官的动态增强磁共振成像数据的定量分析之中。
This study proposes a new method called MRR-FCM for quantitative analysis of routinely acquired clinical uterine dynamic contrast-enhanced MRI,which combines our modified reference region model for estimating T1(0) of tissues,fuzzy C-means method and a reference region model including plasma volume fraction.The MRR-FCM method consists of five steps: 1)Automatically segmenting uterine image area using fuzzy C-means method;2)Estimating baseline T1(0) over uterine area using our modified reference region model;3)Converting signal intensities to contrast agent concentration;4)Automatically clustering contrast agent concentration curves into pre-defined number of clusters using fuzzy C-means method;5)Calculating pharmacokinetic parameter over uterine area pixel by pixel using a reference region model including plasma volume fraction.6 patients with pathologically proven cervix cancer were examined using the MRR-FCM method to demonstrate the feasibility and effectiveness of the MRR-FCM method in a typical clinical setting.The MRR-FCM method provides in vivo quantitative functional parameters of Ktrans and ve,which reveal the heterogeneity over lesion area.A significant difference(p0.01) in Ktrans between tumor area and normal tissue was found.The MRR-FCM method permits quantitative analysis of dynamic contrast-enhanced MRI data acquired in a typical clinical setting,which can be translated into analysis of dynamic contrast-enhanced MRI data from other organs.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2011年第3期501-507,共7页
Chinese Journal of Scientific Instrument
基金
973项目(2011CB707701)
北京市共建项目(JD100010609)资助