全脑定量结构MRI和DTI对阿尔茨海默病的实验和临床研究

发布时间：2018-06-04 06:13

本文选题：中枢神经系统 + 扩散张量成像　；参考：《华中科技大学》2013年博士论文

【摘要】：第一部分APP/PS1转基因小鼠活体全脑DTI定量研究目的：以往的研究已将扩散张量成像(diffusion tensor imaging, DTI)应用于阿尔茨海默病(Alzheimer's disease, AD)动物模型的组织病理学研究中,但是少有关于结构特异性方面的报道。基于体素的分析方法(voxel-based analysis, VBA)和基于解剖图谱的分析方法(atlas-based analysis, ABA)是DTI全脑分析方法中两种互补的方法。本研究的目的在于采用全脑DTI的分析方法,明确AD动物模型病理变化的空间位置分布特征。材料与方法：同时采用VBA和ABA的方法,对APP/PS1转基因小鼠(n=9)和野生型对照(n=9)进行全脑的DTI对比分析。采用多种度量指标,如各向异性分数(fractional anisotropy, FA)、扩散轨迹(total diffusivity, trace)、轴向弥散(axial diffusivity, DA)和放射弥散(radial diffusivity, DR)对阿尔茨海默病小鼠不同类型的病理变化进行量化分析。采用Kappa分析的方法对手动描绘的感兴趣区(region of interest, ROI)和基于解剖图谱方法所勾画的ROI进行比较,以评估图像配准的准确性。MR检查之后,对APP/PS1转基因小鼠和野生型对照进行组织学检查分析。结果：结果显示,APP/PS1转基因小鼠存在广泛的脑结构异常,包括灰质区域如新皮层、海马、纹状体、丘脑、下丘脑、屏状核、杏仁核及梨状皮层,和白质区域如胼胝体/外囊、扣带束、隔、内囊、海马伞及视束,均表现为FA值或DA值升高,或者FA值和DA值同时升高(p0.05,FDR校正)。手动描绘的ROI与ABA方法所描绘的ROI之间的平均Kappa值均接近0.8,且在APP/PS1转基因小鼠组和野生型对照组之间无显著性差异(p0.05)。组织病理学分析证实了灰质区域如新皮层和海马区微结构的DTI变化。DTI同时也发现了广泛的白质区域的弥散改变,但这种差异仅靠单层的组织学定性观察难以准确评估。结论：本研究报道了APP/PS1转基因小鼠脑结构特异性的病理变化,同时也证实了全脑DTI定量分析方法在AD动物模型中的可行性。第二部分AD、MCI和健康人群脑白质差异的空间分布模式探讨目的：近年来大量研究均发现阿尔茨海默病(AD)患者、轻度认知障碍(MCI)患者和健康人群的脑白质完整性存在显著差异,然而AD和MCI患者脑白质损害的空间分布模式少有报道。本研究旨在通过全脑的DTI定量分析,探讨AD、MCI和健康人群脑白质差异的空间分布模式,找到疾病早期诊断和监测疾病进展的可靠指标。材料与方法：依据NINCDS-ADRDA可能AD的标准纳入AD患者21例(M/F=11/10,平均年龄66.8岁)；依据Petersen的标准纳入MCI患者8例(M/F=3/5,平均年龄64.4岁)；及无神经系统疾病的健康对照15例(M/F=8/7,平均年龄65.3岁)。采用GE公司signa HDxt3.0Tesla超导磁共振扫描仪行扩散张量成像(diffusion tensor imaging,DTI),扫描参数如下：TR/TE=10000/83ms, FA=90°, Matrix=256x256, FOV=240mmx240mm, Phase FOV=1,层厚3.0mm无间隔,NEX=1,42层覆盖全脑,b值为1000s/mm2,30个方向。得到DTI原始图像之后,利用DTIstudio软件进行FA图重建,利用DiffeoMap软件对图像进行基于解剖图谱的分析,测量深部灰质和深部白质共58个脑区结构的FA值。AD、MCI和健康对照组58个脑区结构的FA值首先采用单因素方差分析并进行事后检验,两两比较组间差异；然后对相关脑区FA值与简易精神状态量表(mini-mental state examination, MMSE)评分做相关分析。结果：与健康人群相比,AD患者深部灰质和深部白质结构存在广泛的FA值降低(p0.05,FDR校正)。其中,胼胝体压部和丘脑的FA值在MCI组和健康对照组间存在显著差异(p0.05,FDR校正),但在AD组和MCI组间无差异(p0.05)；扣带束和上纵束等8个结构的FA值在AD组和MCI组间有显著差异(p0.05,FDR校正),但在MCI组和健康对照组间无差异(p0.05)。相关分析显示,扣带束和上纵束的FA值与MMSE评分存在显著的正相关关系,以右侧扣带束的相关系数值最高(r=0.606,p=0.001)；而胼胝体压部和丘脑区域FA值与MMSE不存在相关关系(p0.05)。结论：AD和MCI患者脑白质损害的空间分布模式存在显著差异。胼胝体压部和丘脑显微结构病变是早期事件,与认知功能下降关系不大。而扣带束和上纵束白质病变与疾病进展有关,与认知功能下降显著相关。第三部分定量结构MRI对阿尔茨海默病的鉴别诊断研究目的：提出一种全新的方法,可将脑部T1加权磁共振(magnetic resonance, MR)图像转变为特征矢量,应用于基于内容的图像检索(content-based image retrieval, CBIR)。为了克服临床中同一人群的解剖学个体差异及成像参数的不一致性,我们提出了一种基于目标图像与解剖图谱之间差异的图像分析方法(Gap between an Atlas and a target Image Analysis, GAIA),利用基于解剖图谱的图像分割方法(atlas-based analysis, ABA),寻找目标图像与解剖图谱之间差异的大小,从中提取目标图像的解剖学特征,用于阿尔茨海默病的鉴别诊断研究。材料与方法：选取阿尔茨海默病(Alzheimer's disease, AD)、亨廷顿病(Huntington's disease, HD)、脊髓小脑性共济失调6型(Spinocerebral ataxia type6, SCA6)、原发性进行性失语症(primary progressive aphasia, PPA)患者及正常人的T1加权MR图像共102例,作为训练数据。另外随机选取AD、HD、SCA6、PPA患者及正常人的T1加权MR图像共170例作为测试数据。采用GAIA的方法对训练数据进行模式分类,分别提取AD、HD、SCA6、PPA患者及正常人的神经解剖学特征作为特征矢量；随后将这些特征矢量应用到测试数据中,每一个测试数据分别得到一个判别得分(discriminant score),利用判别得分对其进行病种的判别,并评估GAIA判别不同种类疾病的准确性。结果：从训练数据中提取出来的特征矢量,与我们所选取的各神经变性疾病所对应的病理学标志完全一致。大部分测试数据的判别得分能够准确的将其分类至各自对应的疾病种类中去。不具备该疾病典型相关解剖学特征的数据不能被准确分类。GAIA可将阿尔茨海默病从其它类型的神经变性疾病中区分开来。结论：我们提出的GAIA方法,是基于疾病相关的解剖学特征的提取方法,在图像的特征提取与模式识别中有着广阔的应用前景。在未来,可使得放射科医生只需要提交一名患者的图像,就能够将具有类似解剖学特征的相关临床病例全部检索出来,从而对某种疾病的诊断、治疗、预后及随访预测进行大样本的人口学普查及统计分析。
[Abstract]:Part one quantitative study of whole brain DTI in APP/PS1 transgenic mice
Objective: Previous studies have applied diffusion tensor imaging (DTI) to the histopathological study of the animal model of Alzheimer's disease (AD), but there are few reports on structural specificity. The voxel based analysis (voxel-based analysis, VBA) and anatomic map based Atlas-based analysis (ABA) is the two complementary method in the DTI whole brain analysis. The purpose of this study is to identify the spatial distribution characteristics of the pathological changes in the AD animal model by using the whole brain DTI analysis method.
Materials and methods: at the same time, VBA and ABA were used to compare the whole brain DTI of APP/PS1 transgenic mice (n=9) and wild type control (n=9). A variety of metrics, such as the anisotropy fraction (fractional anisotropy, FA), the diffusion trajectory (total diffusivity, trace), axial dispersion, and radiation diffusion were used. Dial diffusivity, DR) quantified the pathological changes of different types of Alzheimer's disease mice. Kappa analysis was used to compare the manually depicted region of interest (region of interest, ROI) and ROI based on the anatomic mapping method to evaluate the accuracy of the image registration by.MR, and to APP/PS1 GM The mice and wild type control were examined histologically.
Results: the results showed that the APP/PS1 transgenic mice had extensive brain structural abnormalities, including the gray matter regions such as the new cortex, the hippocampus, the striatum, the thalamus, the hypothalamus, the screen nucleus, the amygdala and the pyriform cortex, and the white matter areas such as the corpus callosum / outer capsule, the buckle band, the septum, the internal capsule, the hippocampal umbrella and the optic tract, or the value of the FA and the DA, or the value of the FA and DA. The average value of the value increased simultaneously (P0.05, FDR correction). The average Kappa value between the manual depicted ROI and the ABA method was close to 0.8, and there was no significant difference between the APP/PS1 transgenic mice and the wild type control group (P0.05). The histopathological analysis confirmed that the DTI change.DTI of the gray matter region, such as the neocortex and the hippocampus microstructures, was also at the same time Extensive changes in the white matter area were found, but the difference was difficult to accurately assess by single layer histological observation.
Conclusion: This study reported the pathological changes in the specific brain structure of APP/PS1 transgenic mice, and also confirmed the feasibility of DTI quantitative analysis in the AD animal model.
The second part is the spatial distribution pattern of white matter difference between AD, MCI and healthy people.
Objective: in recent years, a large number of studies have found significant differences in white matter integrity between patients with Alzheimer's disease (AD), mild cognitive impairment (MCI) and healthy people. However, there are few reports of spatial distribution patterns of brain white matter damage in AD and MCI patients. This study aims to explore AD, MCI and healthy brain by quantitative analysis of DTI in the whole brain. The spatial distribution pattern of white matter is a reliable index for early diagnosis and monitoring of disease progression.
Materials and methods: 21 cases of AD patients (M/F=11/10, mean age 66.8 years) were included according to the standard of NINCDS-ADRDA possible AD; 8 cases of MCI patients (M/F=3/5, average age 64.4 years old) were incorporated according to the Petersen standard; 15 healthy controls (M/F=8/7, mean age 65.3) with no nervous system disease (M/F=8/7, 65.3 years old). Diffusion tensor imaging (DTI), the scanning parameters are as follows: TR/TE=10000/83ms, FA=90, Matrix=256x256, FOV=240mmx240mm, Phase FOV=1, layer thickness 3.0mm spacer, NEX=1,42 layer covering the whole brain. The DiffeoMap software was used to analyze the image based on the anatomic map, to measure the FA value.AD of 58 brain regions in deep gray matter and deep white matter. The FA values of 58 brain regions in the MCI and the healthy control group were first used for single factor analysis of variance and after the post test, and 22 of the differences were compared. Then the FA value and simple spirit in the related brain regions were compared. Mini-Mental State Examination (MMSE) score was used for correlation analysis.
Results: compared with the healthy population, the deep gray matter and deep white matter structure of the AD patients had extensive FA values (P0.05, FDR correction). There were significant differences between the corpus callosum pressure and the FA value of the thalamus between the MCI group and the healthy control group (P0.05, FDR correction), but there was no difference between the AD group and the MCI group (P0.05); the 8 structures of the cingulate bundle and the upper longitudinal bundle were found. The FA values were significantly different between the AD group and the MCI group (P0.05, FDR correction), but there was no difference between the MCI group and the healthy control group (P0.05). The correlation analysis showed that there was a significant positive correlation between the FA value of the cingulate bundle and the upper longitudinal beam and the MMSE score, which was the highest (r=0.606, p=0.001) of the right cingulate band (r=0.606, p=0.001), and the corpus callosum pressure and the thalamus region F. There is no correlation between the A value and the MMSE (P0.05).
Conclusion: there are significant differences in spatial distribution patterns of brain white matter damage in AD and MCI patients. The lesions of the corpus callosum and thalamus are early events and have little to do with the decline of cognitive function. The buckle and upper longitudinal bundle white matter is related to the disease progression, which is significantly related to the decline of cognitive function.
The third part is quantitative structure MRI in the differential diagnosis of Alzheimer's disease.
Objective: to propose a new method to transform the T1 weighted magnetic resonance (MR) image into the feature vector and apply it to the content based image retrieval (content-based image retrieval, CBIR). In order to overcome the inconsistency of the individual differences and the imaging parameters of the same population in the clinic, we put forward a new method. Gap between an Atlas and a target Image Analysis, GAIA, based on the difference between the target image and the anatomical map, and using the image segmentation method based on the anatomic map (atlas-based analysis) to find the difference between the target image and the anatomic map, and extract the anatomical features of the target image from the image analysis method. A study on the differential diagnosis of Alzheimer's disease.
Materials and methods: Alzheimer's disease (AD), Huntington's disease (Huntington's disease, HD), spinal cerebellar ataxia type 6 (Spinocerebral ataxia type6, SCA6), and 102 cases of primary progressive aphasia (primary progressive) and normal people were used as training data. In addition, a total of 170 cases of T1 weighted MR images of AD, HD, SCA6, PPA and normal people were selected as test data. The training data were classified by GAIA method, and the neuroanatomical features of AD, HD, SCA6, PPA patients and normal people were extracted respectively as feature vectors, and then these feature vectors were applied to the test data, each of which was applied to each of the test data, each of which was applied to the test data, each of which was applied to each of the test data. A discriminant score (discriminant score) was obtained for the test data, and the discriminant score was used to discriminate the disease and evaluate the accuracy of GAIA to distinguish the different kinds of diseases.
Results: the feature vectors extracted from the training data are in complete agreement with the pathological signs corresponding to the neurodegenerative diseases we selected. Most of the test data can be accurately classified into their respective disease types. Data that does not possess the typical anatomical characteristics of the disease can not be found. Accurate classification of.GAIA can distinguish Alzheimer's disease from other types of neurodegenerative diseases.
Conclusion: the GAIA method, which we propose, is based on the extraction of disease-related anatomical features, and has a broad application prospect in image feature extraction and pattern recognition. In the future, the radiologist can only submit one patient's image to all related clinical cases with similar anatomical characteristics. A large population census and statistical analysis of the diagnosis, treatment, prognosis and follow-up prediction of a disease are conducted.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：R445.2;R749.16

【相似文献】