State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, 100875, China.
School of Psychology, Beijing Normal University, Beijing, 100875, China.
Hum Brain Mapp. 2018 Jul;39(7):2786-2799. doi: 10.1002/hbm.24040. Epub 2018 Mar 25.
The functional profiles of regions in the ventral occipital-temporal cortex (VTC), a critical region for object visual recognition, are associated with the VTC connectivity patterns to nonvisual regions relevant to the corresponding object domain. However, whether and how whole-brain connections affect recognition behavior remains untested. We directly examined the necessity of VTC connectivity in object recognition behavior by testing 82 patients whose lesion spared relevant VTC regions but affected various white matter (WM) tracts and other regions. In these patients, we extracted the whole-brain anatomical connections of two VTC domain-selective (large manmade objects and animals) clusters with probabilistic tractography, and examined whether such connectivity pattern can predict recognition performance of the corresponding domains with support vector regression (SVR) analysis. We found that the whole-brain anatomical connectivity of large manmade object-specific cluster successfully predicted patients' large object recognition performance but not animal recognition or control tasks, even after we excluded connections with early visual regions. The contributing connections to large object recognition included tracts between VTC-large object cluster and distributed regions both within and beyond the visual cortex (e.g., putamen, superior, and middle temporal gyrus). These results provide causal evidence that the VTC whole-brain anatomical connectivity is necessary for at least certain domains of object recognition behavior.
腹侧枕颞叶皮层（VTC）中区域的功能特征与非视觉区域的 VTC 连接模式相关，这些区域与相应的对象域相关。然而，整个大脑的连接是否以及如何影响识别行为仍未得到检验。我们通过测试 82 名患者，直接检查了 VTC 连接在物体识别行为中的必要性。这些患者的病变区域未影响到相关的 VTC 区域，但影响了各种白质（WM）束和其他区域。在这些患者中，我们使用概率追踪技术提取了两个 VTC 域选择性（大型人造物体和动物）集群的全脑解剖连接，并使用支持向量回归（SVR）分析来检查这种连接模式是否可以预测相应域的识别性能。我们发现，特定于大型人造物体的 VTC 集群的全脑解剖连接成功预测了患者对大型物体的识别性能，但不能预测动物识别或控制任务，即使我们排除了与早期视觉区域的连接。对大型物体识别有贡献的连接包括 VTC-大型物体集群与视觉皮层内外分布区域之间的束（例如，壳核、上颞叶和中颞叶）。这些结果提供了因果证据，表明 VTC 的全脑解剖连接对于至少某些物体识别行为的域是必要的。
