袁小东,张静,敖国昆,王培军,田嫒.CT灌注曲线形状的量化及对乳腺癌的诊断价值——计算机模拟和乳腺灌注[J].中国医学影像技术,2010,26(5):963~966
CT灌注曲线形状的量化及对乳腺癌的诊断价值——计算机模拟和乳腺灌注
Quantification of CT perfusion curve shape and its diagnosis of breast cancer: Computer simulation and breast perfusion
投稿时间:2009-09-21  修订日期:2009-11-05
DOI:
中文关键词:  灌注成像  乳腺肿瘤  时间-密度曲线  计算机模拟  去卷积
英文关键词:Perfusion imaging  Breast neoplasms  Time density curve  Computer simulation  Deconvolution
基金项目:
作者单位E-mail
袁小东 中国人民解放军第309医院放射科,北京 100091  
张静 同济大学附属同济医院放射科,上海 200065 zhangjingyxd@163.com 
敖国昆 中国人民解放军第309医院放射科,北京 100091  
王培军 同济大学附属同济医院放射科,上海 200065  
田嫒 中国人民解放军第309医院放射科,北京 100091  
摘要点击次数: 2293
全文下载次数: 947
中文摘要:
       目的 探讨血流动力学参数及参数组合对CT灌注时间-密度曲线(TDC)形状的影响,寻找定量评价TDC形状的方法,并应用于乳腺癌的前瞻性灌注研究。方法 在Excel 2003中构建算式,并编写相应程序代码以实现自动卷积、去卷积算法,其中推动剩余函数采用Weibull函数,在固定其他参数的情况下,分别以连续递增的血流量(BF)、平均通过时间(MTT)、血流量与平均通过时间之比(BF/MTT)、Weibull函数的形状参数和一个固定的动脉输入函数卷积,得到一系列TDC;考察上述指标对TDC形状的影响,筛选出能够定量描述TDC形状的变量。对28例乳腺肿块患者行CT灌注扫描,按TDC的形状分为速升速降型、持续上升型、平台型,以术后病理诊断为金标准,用受试者工作特性曲线(ROC)分析BF、MTT、血容量(BV)、BF/MTT、TDC形状分型在乳腺癌诊断中的价值。结果 ①Weibull函数的形状参数对TDC的影响轻微;BF影响TDC的高度,BF越大TDC高度越大;MTT影响TDC的高度和峰值时间,MTT越大TDC高度越大同时峰值时间推后;BF/MTT决定TDC的形状,取值偏大时TDC呈速升速降型,取值偏小时呈持续上升型,取值在两者之间时呈平台型;②乳腺灌注BF/MTT与TDC形状分型相关;③乳腺灌注各指标ROC曲线下面积由高到低依次为BF/MTT、BF、TDC形状分型、BV、MTT,前三者大于对角线下面积(P<0.05),可鉴别乳腺肿块的良恶性。结论 BF/MTT能够定量描述CT灌注TDC的形状,以非线性的方式整合BF和MTT的诊断信息,在乳腺癌的诊断中有重要的参考价值。
英文摘要:
      Objective To assess the influence of hemodynamic parameters and parameter combinations on shapes of CT perfusion time density curve (TDC), and to seek a method of quantitative evaluation on TDC shapes, as well as to use it in prospective research of CT perfusion in breast cancer. Methods Programming Excel 2003 to automate convolution and deconvolution calculations, Weibull function was used as impulse residue function (R(t)). When fixing the other parameters, progressive value of blood flow (BF), mean transit time (MTT), BF to MTT ratio (BF/MTT), parameters were taken respectively to obtain a series of virtual TDC, by which influence of the above parameters on TDC shape was verified and parameters that could describe TDC shape quantitatively were screened out. Twenty-eight patients with breast tumors underwent CT perfusion scan, and the presentations were classified as washout, persistent and plateau types according to TDC shape. Taking postoperative pathologic diagnosis as the golden standard, the value of BF, MTT, BV, BF/MTT and TDC shape type in the diagnosis of breast cancer was analyzed with receiver operating characteristic curve (ROC). Results ①The shape parameter of Weibull function had little influence on TDC. BF affected the height of TDC: the larger the BF was, the higher the TDC was. MTT affected the height and peak time of TDC: the larger the MTT was, the higher the TDC and at the same time peak time was delayed. BF/MTT determined TDC shape, TDC presenting as a washout type when relatively large values were taken, and as a persistent type when relatively small values were taken. ②BF/MTT of breast tumor was correlated with the TDC shape type. ③The area under ROC of all parameters of breast perfusion decreased in sequence as: BF/MTT, BF, TDC shape type, blood volume (BV) and MTT. The three former were larger than the area under the diagonal line (P<0.05), and had significance in differential diagnosis of malignant and benign breast tumors. Conclusion BF/MTT can describe TDC shape of CT perfusion quantitatively. It integrates the diagnostic information of BF and MTT in a non-linear way, and is therefore of significant reference value in the diagnosis of breast cancer.
查看全文  查看/发表评论  下载PDF阅读器