认知与发展心理学研究室知识库

内侧颞叶（medial temporal lobe，MTL）是执行记忆功能的核心区域之一，也是人脑中最具可塑性的组成部分。MTL跨越皮层下和皮层，包括从古皮质到新皮质的复杂神经结构，其各区域不仅在空间结构上互相毗邻又相互区分，在细胞构筑和信号通道上均是互相联系而又有差别。因其与老化及神经退行性疾病的密切关系，很多研究聚焦 MTL，以期揭示对老化或疾病具有特异性的形态变化，但MTL各区域在健康人群毕生发展过程中的变化规律，仍未可知。
随着磁共振成像技术的不断完善，探究 MTL形态的毕生发展规律成为可能。本研究首先基于多中心横向数据：美国南西克莱恩研究所洛克兰样本（200 人，6-80岁）、中国彩巢计划（51人，7-17 岁）、中国成人毕生发展数据（149 人，18-80岁），构建了大样本高精度人脑结构形态影像毕生发展（共计 400 人，6-80 岁）样本。其次，结合 FreeSurfer 皮层下脑图谱和美国人脑连接组计划高清皮层图谱，分离出 MTL 分区的七个子区域：皮层下的海马（hippocampus，HP），以及皮层上的内嗅皮层（entorhinal cortex，EC）、围嗅皮层（perirhinal cortex，PRC）、前下托（presubiculum，PreS）、 三 个 海 马 旁 回 区 域 （parahippocampal area，  PHA）（PHA1、PHA2、PHA3）。再次，计算个体MTL上述子区域的多维形态学指标：体积、皮层面积、皮层厚度、分形维度，在此基础上控制颅内体积、性别、扫᧿参数等协变量，对各区域形态指标进行年龄回归拟合，经多重比较校正后筛选出
显著随年龄变化的指标，建模各指标最佳随龄变化模型，绘制 MTL 各区域形态学指标的毕生发展轨线。最后，为进一步考察 MTL 随龄纵向发展模式，本研究基于 198 名中国学龄儿童的 429 个纵向追踪样本，利用广义混合可加模型建模绘制出 MTL各形态学指标在 6-20 岁期间的生长发育曲线。  
上述毕生发展建模分析发现： MTL各区域毕生发展遵循其基本的空间和功能分布；皮层下区域的毕生变化呈现倒 U 形，学龄期的发育保持稳定或缓慢上升，在 50 岁左右加速下降；皮层区域的毕生变化均以线性下降为主，但是不同子区域存在其局部变化特点；这些毕生发展模式在不同的样本和中心之间具备可重复性，在特定区域（如 PHA）东西方样本之间的差异明显，尤其表现在皮层厚度上。上述结果不仅是首次在中国人群上绘制出 MTL 的毕生发展轨线和生长发育曲线，而且也是国际上第一个系统的 MTL 高精度分区的毕生发展图谱，这不仅将推进发展认知神经科学的 MTL 知识更新，更可为从发育到老化全过程的临床监测、
预防、诊断ᨀ供形态量化参考。

Medial temporal lobe (MTL) is one of the key areas of memory function, demonstrating the most plastic component of the human brain. It expands from the subcortex to the cortex,  including the complex nerve structure from the allocortex to the neocortex. Not only the spatial structures are distinguished from each other among the adjacent subdivisions of MTL, but also their cell constructions and positions on the signaling pathway are different. Because of its close relationship with aging and neurodegenerative diseases, many studies focus on MTL to reveal its specific morphological changes in aging or disease. However, developmental changes of its subdivisions across the human lifespan are still unknown.
Along with the continuous improvement of magnetic resonance imaging technology, it is now possible to explore how the MTL changes across lifespan. First, based on a multi-center cross-sectional data from the Nathan Kline Institute – Rockland Sample (N = 200, 6-80 years), Chinese Color Nest Project (N = 51, 7-17 years), Chinese Adult Lifespan Dataset (N = 149, 18-80 years), this study constructed a large sample of high-resolution structural MRI of human brain morphology across lifespan (N = 400, 6-80 years). Second, the MTL were divided into seven sub-divisions including the hippocampus (HP), the entorhinal cortex (EC), the perirhinal cortex (PRC), the presubiculum (PreS), and three parahippocampal areas (PHA) (PHA1, PHA2, PHA3), according to the parcellations from FreeSurfer and the Human Connectome Project. The multiple morphological metrics of the sub-divisions were derived: volume, surface area, cortical thickness and fractal dimensionality. By fitting linear regression model of each metric to age-related changes with controlling the covariates such as intracranial volume, sex and imaging parameters, significant lifespan changes were detected after Bonferroni correction. Estimation of the lifespan trajectory percentiles and prediction of these lifespan trajectories were achieved for each metric. Finally, in order to further investigate the longitudinal development model of MTL and its sub-divisions, we collected 429 longitudinal data from 198 Chinese typically developing school-age children and depicted their growth curves of each metrics during 6-20 years old, using the Generalized Addictive Mixed Models (GAMMs).
Our analyses demonstrated that: 1) the lifespan development of the MTL follows its basic spatial and functional distribution; 2) Subcortical area showed an inverted U-shaped across lifespan, keeping stable or increasing slowing during school age and decreasing progressively after 50 years old; 3) The lifespan changes of the cortical region are dominated by linear decreases, but there are local variations in the different sub-divisions. These lifespan development models are reproducible across different samples and centers. The obvious difference in PHA was from the cortical thickness. The present findings are using a high-precision parcellation, not only for the first time in the Chinese population to depict the MTL lifespan trajectory and growth curves, but also the first international systematic lifespan trajectory map of the MTL lifespan development. It will promote updating  the knowledge of the MTL in the field of cognitive developmental neuroscience, as well as providing quantitative reference to the clinical monitoring, prevention, diagnosis of the MTL related brain diseases during the whole process form growth to aging.