Defitech Chair in Clinical Neuroengineering, Center for Neuroprosthetics and Brain Mind Institute, EPFL, Geneva CH-1201, Switzerland.
Defitech Chair in Clinical Neuroengineering, Center for Neuroprosthetics and Brain Mind Institute, Clinique Romande de Readaptation (CRR), EPFL Valais, Sion CH-1950, Switzerland.
Cereb Cortex. 2022 Jul 21;32(15):3187-3205. doi: 10.1093/cercor/bhab409.
Discrimination and integration of motion direction requires the interplay of multiple brain areas. Theoretical accounts of perception suggest that stimulus-related (i.e., exogenous) and decision-related (i.e., endogenous) factors affect distributed neuronal processing at different levels of the visual hierarchy. To test these predictions, we measured brain activity of healthy participants during a motion discrimination task, using electroencephalography (EEG) and functional magnetic resonance imaging (fMRI). We independently modeled the impact of exogenous factors (task demand) and endogenous factors (perceptual decision-making) on the activity of the motion discrimination network and applied Dynamic Causal Modeling (DCM) to both modalities. DCM for event-related potentials (DCM-ERP) revealed that task demand impacted the reciprocal connections between the primary visual cortex (V1) and medial temporal areas (V5). With practice, higher visual areas were increasingly involved, as revealed by DCM-fMRI. Perceptual decision-making modulated higher levels (e.g., V5-to-Frontal Eye Fields, FEF), in a manner predictive of performance. Our data suggest that lower levels of the visual network support early, feature-based selection of responses, especially when learning strategies have not been implemented. In contrast, perceptual decision-making operates at higher levels of the visual hierarchy by integrating sensory information with the internal state of the subject.
运动方向的辨别和整合需要多个脑区的相互作用。感知的理论解释表明,刺激相关的(即外源性的)和决策相关的(即内源性的)因素会影响视觉层次结构的不同水平上分布式的神经元处理。为了检验这些预测,我们使用脑电图(EEG)和功能磁共振成像(fMRI)在一项运动辨别任务中测量了健康参与者的大脑活动。我们分别对与外源性因素(任务需求)和内源性因素(知觉决策)相关的活动对运动辨别网络的影响进行建模,并将动态因果建模(DCM)应用于两种模式。事件相关电位的 DCM(DCM-ERP)揭示了任务需求影响初级视觉皮层(V1)和内侧颞叶区域(V5)之间的互惠连接。随着练习,更高的视觉区域被越来越多地涉及,这一点通过 DCM-fMRI 揭示出来。知觉决策调节更高的水平(例如,V5 到额眼区,FEF),以一种可预测性能的方式。我们的数据表明,视觉网络的较低层次支持早期基于特征的反应选择,特别是在尚未实施学习策略的情况下。相比之下,知觉决策通过将感觉信息与主体的内部状态相结合,在视觉层次结构的较高层次上运作。