Reddy Kasireddy V, Mitra Abhishek, Yalavarthy Phaneendra K
Annu Int Conf IEEE Eng Med Biol Soc. 2016 Aug;2016:1224-1227. doi: 10.1109/EMBC.2016.7590926.
The deconvolution in the perfusion weighted imaging (PWI) plays an important role in quantifying the MR perfusion parameters. The PWI application to stroke and brain tumor studies has become a standard clinical practice. The standard approach for this deconvolution is oscillatory-limited singular value decomposition (oSVD) and frequency domain deconvolution (FDD). The FDD is widely recognized as the fastest approach currently available for deconvolution of MR perfusion data. In this work, two fast deconvolution methods (namely analytical fourier filtering and analytical showalter spectral filtering) are proposed. Through systematic evaluation, the proposed methods are shown to be computationally efficient and quantitatively accurate compared to FDD and oSVD.
灌注加权成像（PWI）中的去卷积在量化磁共振灌注参数方面起着重要作用。PWI在中风和脑肿瘤研究中的应用已成为标准的临床实践。这种去卷积的标准方法是振荡受限奇异值分解（oSVD）和频域去卷积（FDD）。FDD被广泛认为是目前可用于磁共振灌注数据去卷积的最快方法。在这项工作中，提出了两种快速去卷积方法（即解析傅里叶滤波和解析肖沃尔特谱滤波）。通过系统评估，与FDD和oSVD相比，所提出的方法在计算上更高效且在定量上更准确。
