Voxel-based correlation between coregistered and DSC MRI in subjects with suspected Alzheimer disease.

 
Acta Radiol. 2008 Dec;49(10):1154-61.

Voxel-based correlation between coregistered single-photon emission computed tomography and dynamic susceptibility contrast magnetic resonance imaging in subjects with suspected Alzheimer disease.

Division of Radiology, Karolinska Institute, Karolinska University Hospital, Stockholm, Sweden. lena.cavallin@karolinska.se
BACKGROUND: Current diagnosis of Alzheimer disease is made by clinical, neuropsychologic, and neuroimaging assessments. Neuroimaging techniques such as magnetic resonance imaging (MRI) and single-photon emission computed tomography (SPECT) could be valuable in the differential diagnosis of Alzheimer disease, as well as in assessing prognosis. PURPOSE: To compare SPECT and MRI in a cohort of patients examined for suspected dementia, including patients with no objective cognitive impairment (control group), mild cognitive impairment (MCI), and Alzheimer disease (AD). MATERIAL AND METHODS: 24 patients, eight with AD, 10 with MCI, and six controls, were investigated with SPECT using (99m)Tc-hexamethylpropyleneamine oxime (HMPAO, Ceretec; GE Healthcare Ltd., Little Chalsont UK) and dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) with a contrast-enhancing gadobutrol formula (Gadovist; Bayer Schering Pharma, Berlin, Germany). Voxel-based correlation between coregistered SPECT and DSC-MR images was calculated. Region-of-interest (ROI) analyses were then performed in 24 different brain areas using brain registration and analysis of SPECT studies (BRASS; Nuclear Diagnostics AB, Stockholm, Sweden) on both SPECT and DSC-MRI. RESULTS: Voxel-based correlation between coregistered SPECT and DSC-MR showed a high correlation, with a mean correlation coefficient of 0.94. ROI analyses of 24 regions showed significant differences between the control group and AD patients in 10 regions using SPECT and five regions in DSC-MR. CONCLUSION: SPECT remains superior to DSC-MRI in differentiating normal from pathological perfusion, and DSC-MRI could not replace SPECT in the diagnosis of patients with Alzheimer disease.

PMID:  18855165  [Click to show in PubMed]

在脑功能组分析里面,单样本t检验,有独立样本t检验,配对t检验,avova,协变量回归分析,但是能否进行基于体素的相关分析,譬如看SPECT和PET图像相关关系等?

这篇文章关于方法部分说的过于简单,而且涉及一些不是常见软件,不好重复。

这篇文章我没看。“譬如看SPECT和PET图像相关关系”,是完全可以的,即可以看某两个图的spatial correlation,也可以在组水平上(如20个人,每个人都有两个图)voxel-by-voxel的相关(即across subjects)。

 I have written a program to compute the correlation map between two groups. Please see the GroupCorr.m.