Wiltshire Travis J, Euler Matthew J, McKinney Ty L, Butner Jonathan E
Department of Psychology, University of UtahSalt Lake City, UT, United States.
Department of Language and Communication, Centre for Human Interactivity, University of Southern DenmarkOdense, Denmark.
Front Physiol. 2017 Sep 1;8:633. doi: 10.3389/fphys.2017.00633. eCollection 2017.
Humans are high-dimensional, complex systems consisting of many components that must coordinate in order to perform even the simplest of activities. Many behavioral studies, especially in the movement sciences, have advanced the notion of to describe how systems with many components coordinate to perform specific functions while also exhibiting the potential to re-structure and then perform other functions as task demands change. Consistent with this notion, within cognitive neuroscience it is increasingly accepted that the brain flexibly coordinates the networks needed to cope with changing task demands. However, evaluation of various indices of soft-assembly has so far been absent from neurophysiological research. To begin addressing this gap, we investigated task-related changes in two distinct indices of soft-assembly using the established phenomenon of EEG repetition suppression. In a repetition priming task, we assessed evidence for changes in the correlation dimension and fractal scaling exponents during stimulus-locked event-related potentials, as a function of stimulus onset and familiarity, and relative to spontaneous non-task-related activity. Consistent with predictions derived from soft-assembly, results indicated decreases in dimensionality and increases in fractal scaling exponents from resting to pre-stimulus states and following stimulus onset. However, contrary to predictions, familiarity tended to increase dimensionality estimates. Overall, the findings support the view from soft-assembly that neural dynamics should become increasingly ordered as external task demands increase, and support the broader application of soft-assembly logic in understanding human behavior and electrophysiology.
人类是高维复杂系统，由许多组件组成，这些组件必须协同工作才能完成哪怕是最简单的活动。许多行为研究，尤其是运动科学领域的研究，提出了“软组装”的概念，以描述具有许多组件的系统如何协同执行特定功能，同时还展现出随着任务需求的变化而重新组织并执行其他功能的潜力。与此概念一致，在认知神经科学领域，人们越来越认可大脑能灵活地协调应对不断变化的任务需求所需的网络。然而，迄今为止，神经生理学研究尚未对软组装的各种指标进行评估。为了开始填补这一空白，我们利用已确立的脑电图重复抑制现象，研究了软组装的两个不同指标中与任务相关的变化。在一项重复启动任务中，我们评估了刺激锁定事件相关电位期间关联维数和分形标度指数变化的证据，作为刺激起始和熟悉程度的函数，并相对于自发的非任务相关活动进行评估。与从软组装得出的预测一致，结果表明从静息状态到刺激前状态以及刺激开始后，维数降低，分形标度指数增加。然而，与预测相反，熟悉程度往往会提高维数估计值。总体而言，这些发现支持了软组装的观点，即随着外部任务需求的增加，神经动力学应变得越来越有序，并支持软组装逻辑在理解人类行为和电生理学方面的更广泛应用。
