Imran M B, Kawashima R, Awata S, Sato K, Kinomura S, Ono S, Yoshioka S, Sato M, Fukuda H
Department of Nuclear Medicine and Radiology, Aoba Brain Research Imaging Center, Institute of Development, Aging and Cancer, Tohoku University, Tohoku University Hospital, Sendai, Japan.
J Nucl Med. 1999 Feb;40(2):244-9.
This study assessed the accuracy and reliability of Automated Image Registration (AIR) for standardization of brain SPECT images of patients with Alzheimer's disease (AD). Standardized cerebral blood flow (CBF) images of patients with AD and control subjects were then used for group comparison and covariance analyses.
Thirteen patients with AD at an early stage (age 69.8+/-7.1 y, Clinical Dementia Rating Score 0.5-1.0, Mini-Mental State Examination score 19-23) and 20 age-matched normal subjects (age 69.5+/-8.3 y) participated in this study. 99mTc-hexamethyl propylenamine oxime (HMPAO) brain SPECT and CT scans were acquired for each subject. SPECT images were transformed to a standard size and shape with the help of AIR. Accuracy of AIR for spatial normalization was evaluated by an index calculated on SPECT images. Anatomical variability of standardized target images was evaluated by measurements on corresponding CT scans, spatially normalized using transformations established by the SPECT images. Realigned brain SPECT images of patients and controls were used for group comparison with the help of statistical parameter mapping. Significant differences were displayed on the respective voxel to generate three-dimensional Z maps. CT scans of individual subjects were evaluated by a computer program for brain atrophy. Voxel-based covariance analysis was performed on standardized images with ages and atrophy indices as independent variables.
Inaccuracy assessed by functional data was 2.3%. The maximum anatomical variability was 4.9 mm after standardization. Z maps showed significantly decreased regional CBF (rCBF) in the frontal, parietal and temporal regions in the patient group (P < 0.001). Covariance analysis revealed that the effects of aging on rCBF were more pronounced compared with atrophy, especially in intact cortical areas at an early stage of AD. Decrease in rCBF was partly due to senility and atrophy, however these two factors cannot explain all the deficits.
AIR can transform SPECT images of AD patients with acceptable accuracy without any need for corresponding structural images. The frontal regions of the brain, in addition to parietal and temporal lobes, may show reduced CBF in patients with AD even at an early stage of dementia. The reduced rCBF in the cortical regions cannot be explained entirely by advanced atrophy and fast aging process.
本研究评估了自动图像配准（AIR）用于阿尔茨海默病（AD）患者脑SPECT图像标准化的准确性和可靠性。然后将AD患者和对照受试者的标准化脑血流量（CBF）图像用于组间比较和协方差分析。
13例早期AD患者（年龄69.8±7.1岁，临床痴呆评定量表评分0.5 - 1.0，简易精神状态检查表评分19 - 23）和20例年龄匹配的正常受试者（年龄69.5±8.3岁）参与了本研究。为每位受试者进行了99mTc - 六甲基丙烯胺肟（HMPAO）脑SPECT和CT扫描。SPECT图像借助AIR转换为标准尺寸和形状。通过在SPECT图像上计算的指标评估AIR进行空间标准化的准确性。通过对相应CT扫描进行测量来评估标准化目标图像的解剖学变异性，使用由SPECT图像建立的变换对其进行空间归一化。借助统计参数映射，将患者和对照的重新对齐的脑SPECT图像用于组间比较。在各个体素上显示显著差异以生成三维Z图。通过计算机程序评估个体受试者的CT扫描以检测脑萎缩。以年龄和萎缩指数作为自变量，对标准化图像进行基于体素的协方差分析。
通过功能数据评估的不准确性为2.3%。标准化后最大解剖学变异性为4.9毫米。Z图显示患者组额叶、顶叶和颞叶区域的局部脑血流量（rCBF）显著降低（P < 0.001）。协方差分析显示，与萎缩相比，衰老对rCBF的影响更明显，尤其是在AD早期的完整皮质区域。rCBF降低部分归因于衰老和萎缩，然而这两个因素不能解释所有的缺陷。
AIR能够以可接受的准确性转换AD患者的SPECT图像，而无需任何相应的结构图像。即使在痴呆早期，AD患者除顶叶和颞叶外，脑的额叶区域也可能显示CBF降低。皮质区域rCBF降低不能完全由严重萎缩和快速衰老过程来解释。
