基于MD-CGAN的脑部肿瘤图像生成方法研究

深度学习已广泛用于脑部磁共振（MR）图像分析中,但脑部肿瘤MR图像样本不足会严重影响深度学习模型的性能.提出基于多鉴别器循环一致性生成对抗网络（MD-CGAN）的样本生成方法 .利用所提出的MD-CGAN生成脑部肿瘤病理区域图像,将生成的脑部肿瘤病理区域图像覆盖脑部正常图像子区域,合成得到脑部肿瘤MR图像. MD-CG...     

Finer discrimination of brain activation with local multivariate distance

The organization of human brain function is diverse on different spatial scales. Various cognitive states are always represented as distinct activity patterns across the specific brain region on fine scales. Conventional univariate analysis of functional MRI data seeks to determine how a particular cognitive state is encoded in brain activity by analyzing each voxel separately without considering the fine-scale patterns information contained in the local brain regions. In this paper, a local multivariate distance mapping (LMDM) technique is proposed to detect the brain activation and to map the fine-scale brain activity patterns. LMDM directly represents the local brain activity with the patterns across multiple voxels rather than individual voxels, and it employs the multivariate distance between different patterns to discriminate the brain state on fine scales. Experiments with simulated and real fMRI data demonstrate that LMDM technique can dramatically increase the sensitivity of the detection for the fine-scale brain activity patterns which contain the subtle information of the experimental conditions.     

Finer discrimination of brain activation with local multivariate distance

The organization of human brain function is diverse on different spatial scales. Various cognitive states are always represented as distinct activity patterns across the specific brain region on fine scales. Conventional univariate analysis of functional MRI data seeks to determine how a particular cognitive state is encoded in brain activity by analyzing each voxel separately without considering the fine-scale patterns information contained in the local brain regions. In this paper, a local multivariate distance mapping (LMDM) technique is proposed to detect the brain activation and to map the fine-scale brain activity patterns. LMDM directly represents the local brain activity with the patterns across multiple voxels rather than individual voxels, and it employs the multivariate distance between different patterns to discriminate the brain state on fine scales. Experiments with simulated and real fMRI data demonstrate that LMDM technique can dramatically increase the sensitivity of the detection for the fine-scale brain activity patterns which contain the subtle information of the experimental conditions.     

Two different gene loci related to the spatial patterning of brain ventricle in vertebrate

Observations on living embryonic brains and the microstructure of brain ventricle of goldfish revealed that there are two brain ventricle phenotypes in gynogenetic haploid embryos. One phenotype is as normal as that of the control inbreeding diploid embryos, which has normal differentiated forebrain, midbrain and hindbrain. Another phenotype is obviously abnormal, the brain patterning is irregular, and no distinct brain ventricle can be observed. The ratio of haploid embryos with normal brain pattern to that with abnormal brain pattern is 1:3. This ratio indicates that there are two gene loci involved in the spatial patterning of the brain ventricle. Since the possibility that deleterious recessive mutant alleles exist on both of the two gene loci had been excluded in this experiment, the phenotype represented the expressional state rather than the genotype of these two genes. Therefore, the ratio of 1:3 suggests that the expressing probability for each copy of the two genes is 50%, and the regulatory mechanism of the expression is based on two sets of chromosomes, controlled by the rule of the diploid-dependent regulatory mechanism.     

Two different gene loci related to the spatial patterning of brain ventricle in vertebrate

Observations on living embryonic brains and the microstructure of brain ventricle of goldfish revealed that there are two brain ventricle phenotypes in gynogenetic haploid embryos. One phenotype is as normal as that of the control inbreeding diploid embryos, which has normal differentiated forebrain, midbrain and hindbrain. Another phenotype is obviously abnormal, the brain patterning is irregular, and no distinct brain ventricle can be observed. The ratio of haploid embryos with normal brain pattern to that with abnormal brain pattern is 1∶3. This ratio indicates that there are two gene loci involved in the spatial patterning of the brain ventricle. Since the possibility that deleterious recessive mutant alleles exist on both of the two gene loci had been excluded in this experiment, the phenotype represented the expressional state rather than the genotype of these two genes. Therefore, the ratio of 1∶3 suggests that the expressing probability for each copy of the two genes is 50%, and the regulatory mechanism of the expression is based on two sets of chromosomes, controlled by the rule of the diploid-dependent regulatory mechanism.     

自闭症患者的光学脑成像研究

阐述了新近发展起来的无损光学脑功能成像技术. 通过比较几种常用的脑成像技术, 论述了光学脑成像研究自闭儿童皮层活动的可行性和便利性. 总结了国内外迄今为止利用光学脑成像研究自闭症的初步进展. 最新实验研究表明具有较高自闭特质的儿童，在共同注意任务中其前额叶的血液动力学响应与正常儿童不同. 进一步利用光学脑成像研究脑皮层活动规律及其固有的特征，有望为儿童自闭症的早期诊断提供可靠的客观依据.     

How did brains evolve?

Three reports on mammalian brain evolution analyse the same comparative data on brain component volumes but come to partially conflicting conclusions. Clark et al. conclude from their analysis of volumetric brain proportions ("cerebro-types") that cerebellum size is invariant across mammalian taxonomic groups, the neocortex and cerebellum do not co-vary in size (in contradiction to ref. 1), and cerebrotype-based measures identify directional changes in brain architecture. Here I provide evidence that calls each of these conclusions into question. The failure of the cerebrotype measure to identify species differences in brain architecture that are independent of gross brain size undermines the proposal by Clark et al. that it could be useful for detecting evolutionary patterns and phylogenetic relationships.     

三维多相头部脑组织形变特性仿真

基于头部脑组织多相结构解剖图,建立多相非线性脑组织形变仿真计算模型,包括大脑、小脑、脑干、胼胝体以及脑脊液等材料属性各异的6个组成部分.小脑组织本构方程为非线性,其它各个相组织的本构方程均为线性.计算了转动惯性载荷下多相非线性脑组织的变形状况.计算结果表明,脑组织各相交界处应变较大,最大剪应力出现在大脑额叶侧,为257 Pa,脑组织各部分变形差别显著,这会导致脑组织的撕裂和弥漫性轴索损伤(DAI).     

基于对称导联脑电信号分析方法的脑损伤区域的判别

为研究采用脑电信号判别脑损伤的部位,提出了基于对称导联的信号分析方法。首先,获取安静状态和唤名刺激下脑电数据,提取脑电信号的近似熵和频谱特征;然后,分别计算对称导联的EEG特征,并对损伤区和非损伤区两组的特征值进行单因素方差分析;最后,选取一个病例,分析不同部位的特征参数比值进行损伤部位的判别,并同CT图像诊断结果进行匹配对比。结果表明,严重意识障碍患者的脑损伤和未损伤部位对称电极之间脑电特征参数比值具有显著性的差异(P0.05)。单一病例的分析结果表明,采用本文提出的方法所得出的判断结果是正确的。因此,该方法能够简便可靠地实现脑损伤部位的判别,在临床辅助诊断中有一定的应用推广价值。     

基于混合神经网络的脑部MRI图像语义分割算法

针对传统医学图像对缺乏标注的数据进行自动分割时存在分割精度不高、边缘模糊等问题,提出了一种利用混合神经网络对脑部核磁共振成像(magnetic resonance imaging,MRI)的图像进行语义分割的算法。利用仿射网络对脑部MRI图像进行线性几何变换,基于卷积神经网络进行3D医学图像仿射变换,加入稠密模块减轻梯度消失和加强特征传递问题; 通过空间转换网络对脑部MRI进行空间转换,基于图谱的分割法获得脑部图像的分割结果。采用MICCAI的公共数据集BraTs2019进行实验验证,结果表明,算法可由脑部肿瘤MRI图像获得较好的分割精度和分割效率,为脑部MRI图像语义分割的研究提供一种新的实验方案。     

榆紫叶甲 Ambrostoma quadriim-pressum脑及咽下神经节GABA阳性神经元分布模式的构建

采用树脂石蜡(CP)组织包埋切片技术,结合免疫组化链霉菌抗生物素蛋白-过氧化物酶(SP)双染法,初步构建了榆紫叶甲脑及咽下神经节GABA阳性神经元的分布模式.结果显示,榆紫叶甲脑结构比较特殊:视叶不发达,但具内髓板;嗅叶不发达;蕈形体冠高度退化,而根叶相对发达;中央复合体结构明显;后脑经由围咽神经束与咽下神经结相连,咽下神经结非常发达.榆紫叶甲的所有髓区几乎都有GABA免疫阳性纤维分布.一侧脑半球及咽下神经节共存在约2 800个GABA阳性神经元,分布广泛,且趋向于集群分布.该研究表明榆紫叶甲的脑结构及GABA阳性神经元的分布模式与榆紫叶甲的生活习性、进化地位等密切相关.     

三维颅内血肿动态成像

基于电流场"异物扰动"成像方法,在三维球模型上对颅内血肿动态成像进行了仿真研究.研究了颅内血肿体积大小的改变对头颅表面电位改变的规律,还研究了导电性能差的颅骨的存在对头颅表面电位改变规律的影响.研究结果表明:从头颅表面两个不同的位置注入电流后,从头颅表面电位的变化规律可以反映出颅内血肿体积大小的改变情况,而且电流可以穿透颅骨.这将有助于医务人员根据头颅表面边界电位的变化情况来推测颅内血肿的变化情况.     

基于解剖距离的功能脑网络建模方法研究

为了分析解剖距离对静息态功能脑网络建模的影响,在解剖距离的基础上将节点共同邻居引入到功能脑网络的建模中。设计三个模型,分别是基于解剖距离,节点共同邻居个数和解剖距离与节点共同邻居个数。借鉴残差的方法创建最优静息态功能模型网络评价标准。在每种模型网络中选出最优的模型网络。然后将每种模型中的最优网络与真实数据fMRI网络对比选出能够最大程度模拟真实网络的模型网络。结果证明在AAL模板下单独基于解剖距离建立的功能脑网络能最大化的模拟真实的网络。     

空气污染对儿童大脑发育的影响

目的综述空气污染对儿童身体健康的影响,特别是对于儿童大脑发育的影响,为控制污染和保护儿童健康成长提供参考依据。方法利用文献法对国内外40余篇关于空气污染对儿童大脑发育影响的参考文献进行概述。结果 (1)生活在世界各地大中城市的一些儿童群体样本出现的一定程度神经、行为和认知变化与空气污染暴露有关;(2)怀孕动物产前暴露于单一或复合标准污染物,其子代会产生神经递质永久性变化,同时改变大脑发育;(3)天然屏障(如血脑屏障、鼻腔、肠和肺上皮屏障)的健康发展对儿童健康发育结果至关重要:(4)空气污染会导致大脑中广泛性神经炎症,进而造成大脑各个部位神经组织的损伤和弥漫性损失,最终影响认知功能的发展。(5)接触空气污染物可能会对脊髓的髓鞘和中枢神经系统中神经元的功能产生有害影响,也会使促炎症介质中神经炎症与细胞损失明显增加。结论面对严重的空气污染,对青少年健康特别是对儿童大脑发育的影响日益受到人们的关注。未来研究不仅需要严格控制污染物排放标准,保护人类生存环境;而且需要采用多种技术,进行跨学科协作研究,以实现对暴露于空气污染物环境条件下的儿童进行有效保护,同时也需要制定有效策略,从健康教育和流行病学角度进行深入研究,监测空气污染物对儿童健康的影响,确保儿童健康成长。     

一种新的脑白质分割方法

基于扩散张量成像提出一种新的脑白质分割方法.首先,计算扩散张量成像的各向异性参数和扩散参数,并得到各个参数下的脑部图像;然后,通过期望值最大化(expectation maximization,EM)模型求得各个各向异性参数图像的脑白质和非脑白质区域;最后,通过STAPLE(simultaneous truth and performancel evel estimation)模型融合各个DTI参数图像分割结果,得到脑白质分割结果.实验结果表明,该方法具有较好的分割效果,能有效地从脑组织中分割出脑白质.     

脑信息处理动态特征研究

为提取脑信息处理过程中的动态特征参数,提出运用基于相空间重构思想的时间序列分维算法(G-P算法). 讨论了G-P算法的3个重要参数(即无标度域、嵌入维数和延时)的确定规则,记录大脑在不同状态下的EEG信号并计算其关联维数. 实验结果表明,EEG关联维数能够反映脑信息处理过程中的神经元群活动状态,可作为脑信息处理的非线性特征参数.     

结合光学脑成像及机器学习分类算法对自闭症大脑活动特征的研究进展

自闭症的诊断一直以来是基于行为层面的观察。近年来脑成像研究表明，自闭症大脑存在结构及功能的改变，这些发现为基于成像的诊断提供了新的途径。准确的诊断要基于对大脑异常特征的准确描述，多个脑生理参量的成像可以为自闭症的大脑特征提供更全面的信息。在现有的无损脑功能成像中，光学脑成像具有较高的时间及可接受的空间分辨率，测量时对头部晃动相对不敏感，适用于对儿童，特别是自闭症儿童的研究。近红外光谱技术提供皮层的血氧代谢，而漫射相关谱技术测量皮层的血流。这两类参量可以提供互补的皮层血液动力学信息，结合更为高效的分类算法，有望取得对自闭症大脑皮层活动特征的更为准确的描述和区分。在广泛文献调研及结合自身研究成果的基础上，对利用光学脑成像研究自闭症的大脑活动特征；结合特征参量利用机器学习分类算法对自闭症的预测；以及利用多模态光学脑成像技术（即结合近红外光谱及漫射相关谱）研究自闭症的前景等问题做了系统的回顾与展望，希望为自闭症相关领域的科研人员提供参考和借鉴。     

基于改进BET的MRI脑组织自动提取算法

提出一种基于改进BET(brain extraction tool)的MRI脑组织自动提取算法.首先，该算法结合图像梯度信息能够估计出更为准确的脑重心(center of gravity，COG)；其次，该算法构建了新的脑表面形变力，在垂直于脑表面切线的扩张力中引入了边缘力，该力很好地抑制了脑组织的边界泄漏和过度分割问题.使用本文方法对MRI脑影像进行了自动脑组织提取，实验结果表明，本文算法能够自动获得更加准确的脑组织提取结果，特别是在脑组织边缘处，本文算法与BET算法相比，提取结果更准确.     

人类合作行为的神经生物基础及脑模型

已有的脑成像研究表明,个体在社会博弈中的合作决策主要激活与认知控制、社会认知以及奖励加工有关的三类大脑神经网络.首先对上述三类大脑系统的功能及其相应的大脑结构和神经回路进行阐述,并对相关的脑成像研究进行回顾.对激素水平的研究发现,催产素能通过增加人们之间的信任来促进合作.在此基础上,介绍新近提出的关于合作的脑机制模型,最后指出这一领域已有研究存在的一些问题以及未来的研究方向.