The Clinical Hospital of Chengdu Brain Science Institute, MOE Key Lab for Neuroinformation, Center for Information in Medicine, High-Field Magnetic Resonance Brain Imaging Key Laboratory of Sichuan Province, School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China.
Department of Biomedical Engineering, New Jersey Institute of Technology, 607 Fenster Hall, University Height, Newark, NJ, 07102, USA.
Brain Struct Funct. 2020 Jun;225(5):1601-1613. doi: 10.1007/s00429-020-02078-7. Epub 2020 Apr 30.
Schizophrenia subjects have shown deficits of inhibition in conditions such as a stop signal task. The stop signal response time (SSRT) is consistently longer compared with healthy controls, and is accompanied by decreased brain activations in the right inferior frontal gyrus. However, as to how the response inhibition function is supported by distributed brain networks, and whether such networks are altered in schizophrenia are largely unknown. We analyzed functional MRI data of a stop signal task from 44 schizophrenia patients and 44 matched controls, and performed whole-brain psychophysiological interaction analysis to obtain task-modulated connectivity (TMC). Support vector classification was used to classify schizophrenia, and support vector regression was applied to explore the relationships between TMC and behavior indexes, such as SSRT. Schizophrenia group showed a decreased TMC pattern which mainly involved the fronto-parietal network, and increased TMC related to the sensorimotor network. Moreover, TMC could only successfully predict SSRT in the control group, further suggesting an abnormal task modulation in schizophrenia. Lastly, we compared the classification and prediction results from different types of measures, i.e., TMC, task-independent connectivity (TIC), task-functional connectivity (TFC), and resting-state functional connectivity (RSFC). TMC performed better in the behavior predictions, while TIC performed better in the classification. TFC and RSFC had similar classification and prediction performance as TIC. The current results provide new insights into the altered brain functional integration underlying response inhibition in schizophrenia, and suggest that different types of connectivity measures are complementary for a better understanding of brain networks and their alterations.
精神分裂症患者在停止信号任务等情况下表现出抑制缺陷。与健康对照组相比，停止信号反应时间（SSRT）明显更长，并且右侧额下回的大脑活动减少。然而，关于分布式脑网络如何支持反应抑制功能，以及这种网络在精神分裂症中是否发生改变，在很大程度上尚不清楚。我们分析了 44 名精神分裂症患者和 44 名匹配对照者的停止信号任务的功能磁共振成像数据，并进行了全脑心理生理交互分析以获得任务调节连接（TMC）。支持向量分类用于对精神分裂症进行分类，支持向量回归用于探索 TMC 与 SSRT 等行为指标之间的关系。精神分裂症组显示出 TMC 模式减少，主要涉及额顶网络，并且与感觉运动网络相关的 TMC 增加。此外，TMC 仅能成功预测对照组中的 SSRT，进一步表明精神分裂症中存在异常的任务调节。最后，我们比较了不同类型的测量指标（即 TMC、任务独立连接（TIC）、任务功能连接（TFC）和静息状态功能连接（RSFC）的分类和预测结果。TMC 在行为预测中表现更好，而 TIC 在分类中表现更好。TFC 和 RSFC 的分类和预测性能与 TIC 相似。目前的结果为反应抑制中大脑功能整合的改变提供了新的见解，并表明不同类型的连接测量指标对于更好地理解大脑网络及其改变是互补的。
