【作者】 高飞；

【导师】 柳澄； 王光彬；

【作者基本信息】 山东大学 ， 影像医学与核医学， 2012， 博士

【摘要】 GABA (γ-amino butyric acid, γ-氨基丁酸)是中枢神经系统中主要的抑制性神经递质,并在人类大脑的发育中发挥重要作用。GABA由前体谷氨酸(glutamate, Glu)通过谷氨酸脱羧酶(glutamic acid decarboxylase, GAD)的作用合成,并参与GABA/谷氨酸-谷氨酰胺循环。在这一体内重要的代谢物循环中,氨基丁酸能神经元首先释放GABA,随之被星形细胞接收后合成为谷氨酰胺并释放,接着谷氨酰胺被氨基丁酸能神经元吸收生成谷氨酸,最后谷氨酸被重新合成为GABA。研究表明,许多神经疾病和精神疾病,比如抑郁症、精神分裂症、阿尔茨海默病(Alzheimer’s disease, AD)和多发性硬化(multiple sclerosis, MS)等都会直接影响GABA/谷氨酸-谷氨酰胺循环。所以,测量活体人脑内的GABA含量具有重要的临床意义。磁共振波谱(magnetic resonance spectroscopy, MRS)技术主要利用不同化学环境下原子核的化学位移作用来检测活体组织内代谢物的含量。根据研究原子核的不同,磁共振波谱技术可以分为氢谱(1H)、磷谱(31P)、炭谱(13C)和氟谱(19F)等。但是,由于更高的分辨率和更敏感的信号,氢谱在临床上最常用。磁共振氢谱的发展使我们可以检测和量化脑内的重要代谢产物,包括天门冬氨酸(NAA),肌酸(Cr),胆碱(Cho)和肌醇(mI)等。目前,MEGA-PRESS (MEscher-GArwood Point RESolved Spectroscopy)波谱序列被广泛应用于测量脑内GABA含量。但是,据我们所知,未见有文献利用磁共振波谱,研究正常人群脑内GABA含量与年龄的相关性。所以,本研究在3T磁共振上,利用MEGA-PRESS序列结合波谱编辑技术检测正常人群脑内GABA含量,并探讨GABA含量与年龄、性别和大脑区域的关系。论文包含三个部分,摘要如下：第一部分：MEGA-PRESS序列分离并量化脑内GABA含量的可重复性研究目的：在3T磁共振上,利用新的波谱编辑技术MEGA-PRESS序列,检测健康志愿者脑内额叶感兴趣区(volume of interest, VOI)的GABA含量,并在数据量化过程中,探讨GABA简单量化和相对量化的组间和组内的可重复性。材料和方法：选取10名健康志愿者(5男,5女)作为研究对象,年龄范围为20-30岁,平均年龄25.20±2.66岁。先利用飞利浦3.0T磁共振(Philips, Achieva, TX, Best, The Netherlands) T1加权TFE序列进行扫描,获取MR高分辨率三维大脑影像。然后,利用磁共振波谱MEGA-PRESS序列扫描10名健康志愿者的额叶感兴趣区。所有志愿者都至少扫描两次,间隔时间至少24小时。其中3名志愿者在3星期内连续扫描4次,每次间隔时间至少3天。应用5分量表对共振频率3.02ppm处的GABA谱线质量进行评估。利用jMRUI v.4.0软件对所有波谱数据中的GABA峰值进行简单量化(GABA+)和相对量化(GABA+/Cr; GABA+/NAA)结果：MEGA-PRESS序列对10名健康志愿者的额叶感兴趣区进行了26次扫描。所有数据经软件处理后,都在共振频率3.02ppm处得到显示良好的GABA峰(量表评估均值为4.98)。对于GABA量化方法的组间和组内可重复性,GABA+/Cr的变异系数最低,GABA+/NAA其次,GABA+的变异系数最高。结论：MEGA-PRESS序列在分离脑内GABA上具有较高的稳定性。对于GABA量化方法的组间和组内可重复性,GABA含量的相对量化优于简单量化,而且GABA+/Cr的可重复性优于GABA+/NAA。第二部分：重现并分割磁共振波谱扫描中的脑内感兴趣区目的：利用后处理软件重现磁共振波谱扫描前划定的感兴趣区,并分割感兴趣区的组织成分,以获得其内脑白质、脑灰质和脑脊液三种成分的比例。材料和方法：选取5名健康志愿者(3男,2女)作为研究对象,年龄范围为24-30岁,平均年龄26.20±2.28岁。先利用飞利浦3.0T磁共振T1加权TFE序列进行扫描,获取MR高分辨率三维大脑影像,并手动划定位于顶叶的感兴趣区进行MEGA-PRESS序列波谱扫描。然后,利用MRIcron软件把扫描后获得的所有二维DICOM格式影像数据转换为三维ANALYZE格式,应用FSL软件包,先把ANALYZE格式的大脑影像自动去除颅骨和其它非脑组织,然后,应用软件包的FAST (FMRIB’s automated segmentation tool)功能,获得分割为脑灰质、脑白质和脑脊液三种成分的大脑影像。最后,我们与瑞典哥德堡大学合作,利用他们提供的Re-create软件重现脑内感兴趣区,并自动计算出其内脑灰质、脑白质和脑脊液三种成分的比例。结果：所有后处理软件在PC机(联想ThinkPad T410)上工作稳定,5名志愿者的脑内感兴趣区都重现在全脑影像上。感兴趣区重现的位置与扫描前的定位一致。5名志愿者感兴趣区内三种成分比例的均值为：脑灰质(51.6%),脑白质(39.7%)和脑脊液(8.7%)。结论：FSL软件和瑞典哥德堡大学的Re-create软件相结合,可以稳定地分割并重现脑内的感兴趣区。第三部分：正常人群脑内GABA含量与年龄的相关性研究目的：利用磁共振波谱编辑技术—MEGA-PRESS,探讨年龄因素对脑内额叶和顶叶GABA含量的影响,以及分析GABA含量是否有性别差异和脑内区域差异。材料和方法：当地社区100名(49男,51女)健康的志愿者参加本次研究。年龄范围为20岁-76岁(男：24岁-76岁,平均年龄=46.12±14.48岁；女：20岁-70岁,平均年龄=44.98±14.75岁)。100名志愿者按年龄均匀分布：20岁-30岁：20名(9男,11女)；31岁-40岁：20名(10男,10女)41岁-50岁：20名(10男,10女)；51岁-60岁：20名(11男,9女)61岁-76岁：20名(9男,11女)先利用飞利浦3.0T磁共振T1加权TFE序列进行扫描,获取MR高分辨率三维大脑影像,并手动划定位于额叶和顶叶的感兴趣区进行MEGA-PRESS序列波谱扫描。然后,利用jMRUI v.4.0软件的波谱处理和拟合算法功能对所有波谱数据中3.02ppm处的GABA进行相对量化(GABA+/Cr)。最后,采用FSL软件和瑞典哥德堡大学的Re-create软件相结合的方法,分割并重现所有脑内的感兴趣区。结果：对于所有男女志愿者,额叶和顶叶的GABA相对值与年龄都呈明显负相关(额叶：相关系数为-0.68,p<0.001；顶叶：相关系数为-0.54,p<0.001)。方差分析的结果表明：在额叶感兴趣区内,GABA相对值随年龄下降的趋势具有性别差异,也就是说,年龄与性别存在交互作用(p=0.025)。协方差分析的结果显示,额叶GABA含量的相对值明显高于顶叶(p=0.036)。对于两个脑内区域,男、女志愿者脑内GABA含量的相对值都没有显著差异(额叶：p=0.797；顶叶：p=0.482)。额叶和顶叶感兴趣区内脑灰质比例,即脑灰质/(脑灰质+脑白质),分别为57.85%和57.95%。在额叶和顶叶的感兴趣区,脑灰质比例与年龄没有明显相关性(额叶：p=0.14；顶叶：p=0.11),并且,脑实质比例与GABA相对值(GABA+/Cr)也没有明显相关性(额叶：p=0.30；顶叶：p=0.32)。感兴趣区内脑灰质比例没有性别差异(p=0.37)和脑内区域差异(p=0.82)结论：在额叶和顶叶的感兴趣区内,GABA含量的相对值随年龄增加而减少,而且,在额叶,女性感兴趣区内GABA含量相对值下降的速度比男性快。额叶感兴趣区内GABA含量的相对值明显高于顶叶。此外,男、女志愿者脑内GABA含量的相对值没有显著差异。目前,GABA在大脑中的抑制作用被神经行为学和认知神经学领域广泛关注,所以,本论文对GABA与年龄相关性的结果将为这些研究提供重要的参考标准和理论基础。更多还原

【Abstract】 Gamma-aminobutyric acid (GABA) is the main inhibitory neurotransmitter in the human brain and plays an important role during brain development. GABA is primarily localized in GABAergic neurons, which synthesize GABA from glutamate by glutamic acid decarboxylase (GAD). The GABA/glutamate-glutamine cycle is a critical metabolite shuttle in the human brain:The released GABA is taken up into astrocytes, where it is further metabolized to glutamine. Glutamine released can be taken up by GABAergic neurons, where it is synthesized to glutamate and then new GABA. Numerous reports have been published indicating that the GABA/glutamate-glutamine cycle is affected in a variety of neurological and psychiatric disorders, such as depression, schizophrenia, Alzheimer’s disease and multiple sclerosis.Magnetic resonance spectroscopy (MRS) of the human brain is a non-invasive and powerful technique that permits detection and quantification of key metabolites of the pathological and healthy brain tissue in vivo. Depending on the nucleus, there are many forms in MRS, including phosphorus (31P), carbon (13C), and fluorine (19F), but proton MRS (1H MRS) became more prevalent in the clinical fields because of its higher signal sensitivity and better spatial resolution. The important metabolites detectable with1H MRS include N-acetylaspartate (NAA), creatine (Cr), choline (Cho) and mI.Recently, MEGA-PRESS (MEscher-GArwood Point RESolved Spectroscopy) of J-difference edited method is the most widely used technique for measuring GABA. However, to the best of our knowledge, no previous1H-MRS study has explored the normal age-related variations in regional GABA levels using MEGA-PRESS sequence. Therefore, the purpose of this study was to determine age-related effect of GABA concentrations, and clarify if there are regional and gender differences of GABA concentrations in two brain regions:(a) frontal lobe;(b) parietal lobe in normal human using the J-difference edited MRS at3T.There are three parts in the study:Part1. Repeatability of vivo quantification of GABA by MEGA-PRESS sequence of MRSObjective:To apply edited MRS to measure GABA concentrations in the frontal region at3T, and investigate the intra-individual and inter-individual repeatability of GABA simple and relative quantification.Material and methods:Ten healthy volunteers (5men and5women) aged between20and30years were recruited from the local community (mean=25.20±2.66years). All subjects were scanned on a3T scanner (Philips ’Achieva’ TX, Best, The Netherlands) and T1-weighted three-dimensional TFE images were used as a localizer. The GABA concentration in the frontal region was measured using a MEGA-PRESS sequence.All subjects were scanned at least twice within a period of24hours. In three subjects, four continuous scans were conducted within a period of3weeks. All the metabolite quantification was performed with jMRUI v.4.0software. For simple quantification, no concentration reference was chosen (GABA+), For relative quantification, the internal concentration reference was chosen Cr or NAA (GABA+/Cr or GABA+/NAA).Results:Edited spectra from the frontal region were successfully collected in all10participants. For both the intra-individual and inter-individual repeatability, coefficient of variation of GABA+/Cr was lower than GABA+/NAA and GABA+Conclusions:The MEGA-PRESS technique of MRS is stable to measure GABA in the frontal region at3T. For both the intra-individual and inter-individual repeatability, quantification with GABA+/Cr was much more reliable than GABA+/NAA and GABA+alone.Part2. Segmentation and re-creation of VOIs in the brainObjective:To segment3D T1-weighted brain images and re-create VOIs (volume of interest) for obtaining the ratio of gray matter (GM) volume to the gray matter+white matter (GM+WM) volumes of VOIs in the brain.Material and methods:Five healthy volunteers (3men and2women) aged between24and30years were recruited from the local community (mean=26.20±2.28years). All subjects were scanned on a3T scanner and T1-weighted three-dimensional TFE images were used as a localizer. The GABA concentration in the parietal region was measured using a MEGA-PRESS sequence.Each pixel in the3D T1-weighted brain images was segmented as GM, WM, or cerebrospinal fluid (CSF) using an automatic brain segmentation program, FAST (FMRIB’s automated segmentation tool) in the FSL package. VOIs were co-registered to the anatomical images using the "Re-creation of VOI" Matlab tool. Tissue GM fractions were obtained by calculating the ratio of GM volume to the GM+WM volumes in the VOIs.Results:All post-processing software stably run in the personal computer (Lenovo ThinkPad T410). The VOIs of5subjects were re-created in the brain, which is consistent with the localization of VOI before MRS scanning. The mean of three tissue fractions in the5VOIs are GM (51.6%), WM (39.7%) and CSF (8.7%)。Conclusions:It is stable to segment3D T1-weighted brain images and re-create VOIs using FSL package and the "Re-creation of VOI" Matlab tool.Part3. Age-related GABA concentration changes in normal human brainObjective:To determine age-related effect of GABA concentrations, and clarify if there are regional and gender differences of GABA concentrations in normal human using the MEGA-PRESS sequence of MRS at3T. Material and methods:One hundred healthy volunteers (49men and51women) aged between20and76years were recruited from the local community (men:range24-76, mean=46.1±14.5years; women:range20-70, mean=45.0±14.7years). All subjects were scanned on a3T scanner using T1-weighted three-dimensional TFE sequence. The GABA concentrations in the frontal and parietal region were measured using a MEGA-PRESS sequence.All the metabolite quantification was performed with jMRUI v.4.0software. The brain images was segmented as gray matter (GM), white matter (WM), or cerebrospinal fluid (CSF) using the FSL package. VOIs were co-registered to the anatomical images using the "Re-creation of VOI" Matlab tool.Results:There were significant negative correlations between age and GABA+/Cr in both regions studied (frontal region:r=-0.68, p<0.001; parietal region:r=-0.54, p<0.001) from all the subjects. An analysis of variance showed a significant gender difference in the age-related reduction in GABA+/Cr ratio in the frontal region (Gender×age, p=0.025). The GABA+/Cr ratio is significantly higher in the frontal region than in the parietal region (p=0.036). There was no significant difference between men and women for the GABA+/Cr ratio in the frontal region (p=0.797) and parietal region (p=0.482).The mean gray matter tissue fraction GM/(GM+WM) was57.85%and57.95%in the frontal region and parietal region, and did not change significantly with age (p=0.14, p=0.11), and did not correlate with GABA+/Cr ratios (p=0.30, p=0.32) There were no significant gender-related and region-related difference in the fraction of GM in the VOI (p=0.37, p=0.82, respectively).Conclusions:Brain GABA+/Cr ratios decrease with age in the frontal region and parietal region, with the frontal decline is more rapid in women. GABA+/Cr ratio is significantly higher in our frontal region than our parietal region. Evidence of a measureable decline in GABA is important in considering the neurochemical basis of the cognitive decline that is associated with normal aging.更多还原