【作者】 白红民；

【导师】 卢亦成； 王伟民； 江涛；

【作者基本信息】 第二军医大学 ， 外科学， 2010， 博士

【摘要】 手术治疗脑内肿瘤的的理想目标是最大程度切除肿瘤体积,但不损伤运动、感觉、语言及其他重要认知功能。这类手术的关键就是如何术中准确实时“脑功能区”定位。准确的定位才能避免因过多切除而造成永久性神经功能障碍,也可避免因过分“胆小”而造成病灶切除不彻底,难以达到理想的治疗效果。对于运动、感觉等脑的初级功能定位,无论是术前功能磁共振定位还是术中直接电刺激定位,目前均比较成熟。而语言、记忆等高级认知功能的定位目前临床上仍存在问题。语言功能是人类所特有的高级认知功能,相对于运动、感觉、视觉等,语言相关皮质在人脑分布更广也更复杂。不同个体的语言皮质定位存在很大差异,尤其在颅内存在病变的病人,语言皮质的不典型分布则更为普遍[4]。如何定位语言区皮质,是优势半球,尤其是接近语言区病变手术的关键。积极地进行“语言区”脑内病变手术切除而不发生术后语言障碍,从而保障病人的生存质量,已逐渐成为当前神经外科特别关注的问题。术中皮质直接电刺激被视为术中确定语言区皮质和皮质下通路的“黄金方法”,用来提高手术切除效果。但我们实践中发现：术中直接电刺激会产生“假阳性”和“假阴性”刺激结果,“假阳性”刺激可导致肿瘤切除程度降低,而“假阴性”刺激则可导致永久性神经功能障碍,备受关注。“假阴性”刺激的原因有多种,包括刺激强度太低,刺激时间太短及癫痫发作后不应期刺激等。这些“假阴性”刺激可通过严格按照一定的流程和规则刺激而避免[9]。另外一种引起“假阴性”刺激的因素是术中任务选择不当。目前定位术中直接电刺激定位语言区时应用的术中任务主要有“数数”,“阅读”及“命名”任务。命名被认为是语言功能中的一个核心功能,可用来监测语言区。但目前术中所用的命名图片主要包括动物、鸟类、水果、蔬菜以及工具等,且混合在一个命名任务当中。而不同类型的命名区存在不同,造成这种“假阴性”刺激发生。因此研究这种类别特异性命名区脑定位有助于选择合理的术中任务,进一步避免术后语言功能障碍。本研究利用血氧水平依赖功能磁共振(Blood oxygen level dependent-functional magnetic resonance imaging, BOLD-fMRI)技术对健康志愿者名人面孔、动物和工具三种不同类型的命名脑区进行定位；进而使用静息态功能磁共振技术(Rest fMRI)研究这些不同类型命名任务激活脑区的功能连接(Functional connectivity, FC),最后,利用BOLD-fMRI技术和术中皮质直接电刺激技术(Direct electrical stimulation, DES)对脑功能区病变患者进行类别特异性命名区定位,以期发现国人的名人面孔、动物和工具三种不同类型命名脑功能区的分布特点。该研究将有助于神经外科医生进一步了解一些脑区的特殊功能,有助于优化唤醒手术中使用的语言任务,进一步避免“假阴性”刺激和术后语言功能障碍,提高神经外科脑内病变术后病人的生存质量。第一部分健康志愿者类别特异性命名脑区的功能磁共振研究目的：研究健康志愿者名人面孔、动物和工具三种不同类型命名任务的脑定位特点。方法：21名健康志愿者行三种不同类型命名任务的BOLD-fMRI扫描,采用SPM8软件进行脑功能图像分析,获取各激活体素的位置和强度,并用Marsbar0.42软件对其进行感兴趣区分析。结果：名人面孔命名区主要位于双侧前颞叶内侧(包括海马和杏仁核,且有明显的左侧优势)、双侧额下回三角部和盖部；命名动物主要激活左侧辅助运动区；而命名工具主要激活左侧运动前区和右侧辅助运动区。三种任务均有双侧梭状回激活,但命名工具激活范围比其余两种任务都大。57个感兴趣区经过单因素方差分析,有统计学意义的共有14个(P<0.05),面孔特异性激活脑区主要包括双侧前颞叶内侧,右侧岛叶前部、右侧额下回三角部和盖部,工具激活的特异性脑区主要位于右枕上回、左额下回眶部和双侧梭状回。动物激活的脑区在三组间无明显统计学意义。小结：本研究结果提示国人脑内可能存在类别特异性命名区,但分别比较广泛。提示在术中定位功能区过程中,可能需要选择类别特异性命名任务,以减少阴性刺激的发生率。第二部分健康志愿者类别特异性命名区静息态功能连接研究目的：研究健康志愿者名人面孔、动物和工具命名任务激活脑区的相关网络分布特点。方法：使用RESTV1.3软件分析21名健康志愿者的Rest-fMRI数据,根据第一部分研究结果,设计不同类型命名任务激活脑区的“种子点”,与全脑其他脑区进行功能连接(FC)分析,获取与各个“种子点”密切相关的脑区。结果：三种类型命名任务的激活脑区的网络分布比较广泛,名人面孔命名任务激活的脑区网络主要包括双侧前颞叶内侧,右侧额下回三角部和盖部,此外,还包括双侧额中回后部、颞极和左侧顶枕叶外侧面,左颞中回和颞下回。动物命名激活脑区网络则主要位于辅助运动区,双侧颞中回后部、额中回(左侧明显优势)、额下回、顶上小叶和左侧顶下小叶。工具命名激活脑区网络主要有左侧运动前区,右侧枕上回前部,双侧顶叶、额中回后部、颞极和左前颞叶内侧,左额上回后部。小结：名人面孔、动物和工具三种不同类型命名任务的脑定位网络有一定重叠,也存在差异,分布且比较广泛,提示这些部位的手术需要完善相关术前检查,以明确命名区的脑内定位。第三部分类别特异性命名区脑定位的临床研究目的：研究脑功能区病变病人上述三种不同类型命名任务的脑定位特点。方法：18例脑功能区病变病人进行BOLD-fMRI扫描和分析处理,在全麻术中唤醒状态下,使用类别特异性命名任务,术中皮质直接电刺激(DES)等来定位名人面孔、动物和工具三种不同类型的命名脑区。结果：①BOLD-fMRI发现名人面孔、动物和工具三种不同类型命名任务有共同的激活脑区：左额中回中后部,左额下回三角部,辅助运动区和双侧梭状回。同时,也存在类别特异性脑区：在双侧颞极,双侧海马前部和右侧前岛叶,名人面孔命名激活阳性率较高,但这种差别在三组间无统计学意义(P>0.05)。工具命名阳性激活率在双侧顶上小叶和右侧额中回后部相对较高。②DES也证实存在名人面孔特异性命名脑区(刺激过程中只有名人面孔命名受到抑制,而动物命名和工具命名均未受到抑制),但这些脑区分布比较弥散,主要分布在：左颞极、左额下回后部、左额上回后部和左缘上回；刺激过程中名人面孔和动物均出现命名障碍的脑区主要分布左颞叶中部；而动物和工具都命名不能的脑区主要有左额下回后部、左额中回后部和左颞后。三种类型均出现命名不能或语言紊乱的脑区主要位于左缘上回,左额中回后部和左额下回后部以及左颞上回后部；此外电刺激过程中,没有发现单独工具命名、单独动物命名或名人面孔和工具同时命名不能的脑区。小结：国人大脑内存在名人面孔特异性的命名区,动物命名脑区可被名人面孔名人任务或工具命名任务所激活。提示在全麻术中唤醒状态下,利用DES定位命名区的过程中,需要增加名人面孔命名任务,以减少阴性刺激的发生率,避免术后严重功能障碍,提高病人的术后生存质量。结论：该研究明确了国人脑内存在名人面孔类别特异性的命名区,动物命名脑区可以被名人面孔或工具命名任务所激活。这就要求在全麻术中唤醒状态下,利用皮质电刺激定位命名区过程中,需要增加名人面孔命名任务,以减少阴性刺激的发生率。同时,该研究还提示没有必要增加单独动物类型命名任务,避免延长定位时间。此外,该研究还拓宽了“功能区”的范围,使以往我们认为的“非功能区”成为功能区,如“左颞极”等,提示脑功能区手术的范围需要进一步认识,需引起临床的高度注意。更多还原

【Abstract】 The goal of surgery in the treatment of intrinsic cerebral tumors is to resect the maximum neurologically permissible tumor volume, and to spare the eloquent areas that are associated with the control of motor, sense, language, or memory, and other cognitive functions. The key of this kind of surgery is an accurate real-time brain mapping for eloquent areas. Accurate mapping of eloquent areas can avoid both post-operative permanent neurological sequelae caused by intra-operatively damage to the eloquent areas, and a premature interruption of the resection by a "coward" surgical strategy. Brain mapping for motor and/or sensory can be accurately achieved not only by preoperative functional magnetic resonance imaging, but also by direct electrical stimulation. However, mapping of the language, memory, or other cognitive functions remains a problem for current surgical treatment of lesions harbored in the eloquent areas.Language is one of higher cognitive functions unique to the human beings. Language-related cortices are much broader and more complex than that of motor and/or sensory function. The language cortices vary among individuals, and this variation is much more apparent when patients have brain lesions near language cortices. How to map cortical areas for language is critical to the surgery of lesions located in the dominant hemisphere, particularly in areas close to language-related cortices. It gradually becomes a neurosurgical research hot spot for maximally resection of lesions in brain language areas, and meanwhile, minimally damage to the eloquent areas.Intra-operative direct electrical stimulations have been regarded as the "gold standard" in defining cortical and subcortical pathways of the language areas. This technique minimizes definitive post-operative neurological deficit while concurrently improves the quality of resection. However, it is necessary to stress that the technique can also result in "false positives" and "false negatives" even when direct electrical stimulations are properly used. Consequently, "false positives" could wrongly lead to a premature interruption of the resection, whereas "false negatives" could result in permanent neurological sequelae.There are several causes for "false negatives", such as an intensity of stimulation that is too low and stimulation that lasts too short a time or is performed during a transient post-epileptic refractory phase, which may lead to an erroneous technically "negative mapping". Nevertheless, such errors can be avoided by strictly following the practical rules of stimulation.Another cause for false negatives is an inappropriate intra-operative task for functional mapping. Counting, reading and naming are the three mostly used intra-operative tasks for mapping of language. Naming of pictures of objects is the basic task used in mapping procedures because naming is a core component in language abilities and is supposed to be a reliable method for identifying essential language sites. Patients are asked to name a set of line drawing of objects when the stimulation probes are placed at a cortical site. The pictures shown intra-operatively include pictures of animals, birds, fruits, vegetables, man-made objects, etc. They are blended in one naming task. However, cortical sites for naming may vary among different categories. This may results in a "false negative" mapping. Therefore, study on mapping of the category-specific naming cortices can help us to choose a more appropriate intra-operative task for functional mapping which can decrease the postoperative permanent morbidity.In this essay, we firstly conducted a blood oxygen level dependent-functional magnetic resonance imaging (BOLD-fMRI) study to map the category-specific sites associated with naming of famous faces, animals and tools in health Chinese volunteers. Then, the functional connectivity of the activated clusters was calculated by rest-state fMRI. In the third part, we investigated the category-specific naming sites in patient with brain lesions near eloquent areas by both BOLD-fMRI and direct electrical stimulation. The goal of this study was to test the hypothesis of dissociation of category specific naming area in cerebral cortices. The study will contribute to a better understanding of special functions in certain cortices, to help neurosurgeons choose more appropriate intra-operative tasks for functional mapping, and further to avoid postoperative permanent aphasia, and to improve the postoperative quality of life in patients with lesions near eloquent area.Part I A functional MRI study of mapping category-specific sites associated with naming of famous faces, animals and tools.Objective:To study the category specific cortices for naming famous faces, animals and tools in health volunteers.Methods:21 Chinese health volunteers were recruited to study the specific invoked areas involved in naming pictures of these three categories using BOLD-fMR. Functional images were analyzed using statistical parametric mapping (SPM8), and the region of interest (ROIs) was analyzed using Marsbar0.42.Results:Naming famous faces caused more activation in the bilateral medial parts of anterior temporal lobe (including head of hippocampi and amygdala with a significant left dominance). Bilateral activation of pars triangularis and pars opercularis in the naming of famous faces was also discovered by this study. Naming animals invoked greater responses in the left supplementary motor areas, while naming tools invoked more in left premotor areas and right supplementary motor areas. The extent of activated bilateral fusiform gyri by naming of tools was much larger than that by naming of famous faces or animals.57 regions of interest were analyzed by one-way ANOVA, and 14 of 57 activated clusters were shown statistically significant among three categories (P<0.05). They included bilateral medial parts of anterior temporal lobe, right anterior insula, right pars triangularis and pars opercularis in the naming of famous faces, and the right superior occipital gyrus and left pars orbitalis and bilateral fusiform gyri in naming of tools. There were no significant specific activated clusters in naming of animal.Conclusion:The brain cortices involved in the naming process will vary from naming of famous faces, animals and tools. And the category specific naming sites are extensive. This finding suggests that various categories of pictures should be adopted during intra-operative language mapping to get a broader map of language function, in order to minimize the incidence of false-negative stimulations and permanent post-operative deficits.PartⅡRest state functional connectivity of category-specific naming areas in health volunteers.Objective:To study the network distributed cortices related to the activated clusters specific for naming famous faces, animals and tools.Methods:Resting-state fMRI data from 21 healthy volunteers were analyzed the functional connectivity by software REST v1.3 to obtain with network distributed cortices closely related to the "seed voxles" that were picked based on activated clusters specific for naming famous faces, animals or tools.Results:Three types of naming tasks activated a broader network of brain regions Naming famous faces network mainly included bilateral medial parts of anterior temporal lobe and right pars triangularis and pars opercularis. It also included bilateral posterior middle frontal gyrus, bilateral temporal pole, left lateral side of pariet-occipital lobe, left middle and inferior temporal gyrus. Naming animals’network are mainly located in supplementary motor areas, bilateral posterior middle temporal gyrus, and bilateral posterior middle frontal gyrus with a significant left dominance, bilateral inferior frontal gyrus and superior parietal lobule, and left inferior parietal lobule. While naming tools’ network was mainly consisted of left premotor area, right anterior superior occipital gyrus, bilateral parietal lobe, posterior middle frontal gyrus and temporal pole, and left medial temporal lobe, left posterior of superior frontal gyrus.Conclusion:The cortical network involved in different types of naming process overlaps in some cortices. There are also certain differences in activated areas among the three categories. And the category specific naming networks are rather extensive. This finding suggests that some more preoperative examinations are needed to clarify the naming cortices when surgeries are involved in certain areas.PartⅢClinical study on brain mapping of category-specific naming corticesObjective:To study the category specific cortices for naming famous faces, animals and tools in patients with lesions near eloquent areas.Methods:Naming sites for different categories of naming tasks were evaluated by the BOLD-fMRI in 18 patients with brain lesions in cerebral hemisphere. Then during awake procedures, the category specific naming cortices were detected by intraoperative direct electrical stimulation using a category specific naming task.Results:①Using BOLD-fMRI, we found that the common regions activated by naming the three categories of picture were the posterior parts of left middle frontal gyri, pars triangularis of left inferior frontal gyri, supplementary motor areas and bilateral fusiform gyri. There were also category-specific brain regions:Naming famous faces caused more activation in the bilateral temporal poles and head of hippocampi, and right anterior insula. But the differences among three groups are not statistically significant (P> 0.05). Naming tools invoked greater responses in bilateral superior parietal lobule and posterior parts of right middle frontal gyrus.②Intraoperative direct electrical stimulation has also confirmed the existence of famous face-specific naming regions that are distributed extensively, mainly in the left temporal pole, posterior parts of left inferior or superior frontal gyrus and left supramarginal gyrus. The common areas interfered by stimulation when perform naming famous faces, animals and tools were left supramarginal gyrus, posterior parts of left middle frontal gyrus and left inferior frontal gyrus, and the posterior parts of left superior temporal gyrus. There was no cortex specific for naming of animals, or naming of tools, or naming of both famous faces and tools. The middle parts of left temporal lobe were found for naming both famous faces and animals. Other regions for both naming animals and naming tools are mainly located in the posterior parts of left inferior frontal gyrus, middle frontal gyrus and left temporal lobe.Conclusion:The brain cortices involved in the naming process varies from naming of famous faces, animals and tools. There are some cortices exclusively for naming of famous faces. Cortices for naming of animals can be activated by naming of famous faces or tools. This finding suggests pictures of famous faces should be adopted during intra-operative language mapping to get a broader map of language function, in order to minimize the incidence of "false-negative" stimulations and permanent post-operative deficits, and to improve the patient’s postoperative quality of life.更多还原