Assignment of functional activations to probabilistic cytoarchitectonic areas revisited
Assignment of functional activations to probabilistic cytoarchitectonic areas revisited
复制标题
DOI：
10.1016/j.neuroimage.2007.03.060
复制
发表时间：
2007-07-01
期刊：
影响因子：
5.7
通讯作者：
Amunts, Katrin
中科院分区：
文献类型：
--
作者：
Eickhoff, Simon B.;Paus, Tomas;Amunts, Katrin
标准参考空间的概率细胞结构图为分析人脑结构-功能关系提供了有力的工具。虽然这些微观结构定义的图谱已经成功地用于分析体感、运动或语言功能，但在分析结构-功能关系中的几个概念问题仍需要进一步澄清。在本文中，我们通过对视觉手势引起的前顶叶激活的示例性分析，展示了基于概率细胞结构图的功能激活解剖定位的主要方法。在考虑了体积或局部最大值标记的概念基础和实施之后，我们对可能遇到的一些潜在的解释困难、限制和警告进行了评论。在此基础上，提出了一种基于分布分析的结构-功能对应量化方法。这种方法将在特定功能定义位置观察到的细胞结构概率与整个大脑的概率的区域特定零分布（即完整的概率图）联系起来。重要的是，该方法避免了对单个皮质区域的体素进行独特分类的需要，并可能增加不同区域获得的结果之间的可比性。此外，由于基于分布的标记量化了相对于解剖区域的激活的“集中趋势”，它将与已建立的方法相结合，允许对功能激活的解剖底物进行高级表征。最后，讨论了各种方法的优点和缺点，重点是哪种方法最适合特定情况的问题。(c) 2007爱思唯尔公司版权所有。
Probabilistic cytoarchitectonic maps in standard reference space provide a powerful tool for the analysis of structure-function relationships in the human brain. While these microstructurally defined maps have already been successfully used in the analysis of somatosensory, motor or language functions, several conceptual issues in the analysis of structure-function relationships still demand further clarification. In this paper, we demonstrate the principle approaches for anatomical localisation of functional activations based on probabilistic cytoarchitectonic maps by exemplary analysis of an anterior parietal activation evoked by visual presentation of hand gestures. After consideration of the conceptual basis and implementation of volume or local maxima labelling, we comment on some potential interpretational difficulties, limitations and caveats that could be encountered. Extending and supplementing these methods, we then propose a supplementary approach for quantification of structure-function correspondences based on distribution analysis. This approach relates the cytoarchitectonic probabilities observed at a particular functionally defined location to the areal specific null distribution of probabilities across the whole brain (i.e., the full probability map). Importantly, this method avoids the need for a unique classification of voxels to a single cortical area and may increase the comparability between results obtained for different areas. Moreover, as distribution-based labelling quantifies the "central tendency" of an activation with respect to anatomical areas, it will, in combination with the established methods, allow an advanced characterisation of the anatomical substrates of functional activations. Finally, the advantages and disadvantages of the various methods are discussed, focussing on the question of which approach is most appropriate for a particular situation. (c) 2007 Elsevier Inc. All rights reserved.