Crowdsourcing neuroscience: Inter-brain coupling during face-to-face interactions outside the laboratory
Crowdsourcing neuroscience: Inter-brain coupling during face-to-face interactions outside the laboratory
复制标题
DOI：
10.1016/j.neuroimage.2020.117436
复制
发表时间：
2021-02-15
期刊：
影响因子：
5.7
通讯作者：
Poeppel, David
中科院分区：
文献类型：
--
作者：
Dikker, Suzanne;Michalareas, Georgios;Poeppel, David
当我们在社交行为中感到联系或参与时，我们的大脑实际上是在正式的、可量化的意义上“同步”的吗？大多数解决这个问题的研究使用高度控制的任务，具有同质的主题池。为了采取更自然的方法，我们与艺术机构合作，众包神经科学数据：在5年的时间里，我们收集了数千名博物馆和节日游客的脑电图（EEG）数据，他们自愿参与10分钟的面对面互动。两组熟悉程度不同的参与者坐在相互波动机内，这是一种艺术性的神经反馈装置，可以将每对参与者的EEG活动的实时相关性转换为光模式。由于这种参与者间的EEG相关性容易受到噪声污染，在随后的离线分析中，我们使用假想相干性和投影功率相关性计算了脑间耦合，这两个同步度量在很大程度上不受瞬时噪声驱动的相关性的影响。当将这些方法应用于两个具有最一致协议的记录数据子集时，我们发现配对的特质同理心，社会亲密度，参与度和社会行为（联合行动和眼神交流）一致地预测了他们的大脑活动同步的程度，最突出的是低α（类似于7-10 Hz）和β（类似于20-22 Hz）振荡。这些发现支持了这样一种解释，即在动态的、自然的社会互动过程中，共享参与和联合行动驱动耦合了神经活动和行为。据我们所知，这项工作首次证明了跨学科、真实世界、众包神经科学方法可以提供一种有前途的方法来收集与现实生活中的面对面互动有关的大型、丰富的数据集。此外，它还展示了公众如何在实验室之外参与和参与科学过程。博物馆、美术馆或任何其他公众积极参与的组织，都可以帮助促进这种类型的公民科学研究，并支持在科学控制的实验条件下收集大型数据集。为了进一步提高公众对实验室外实验方法的兴趣，本研究的数据和结果通过为公众量身定制的网站（wp.nyu.edu/mutualwavemachine）进行传播。
When we feel connected or engaged during social behavior, are our brains in fact "in sync" in a formal, quantifiable sense? Most studies addressing this question use highly controlled tasks with homogenous subject pools. In an effort to take a more naturalistic approach, we collaborated with art institutions to crowdsource neuroscience data: Over the course of 5 years, we collected electroencephalogram (EEG) data from thousands of museum and festival visitors who volunteered to engage in a 10-min face-to-face interaction. Pairs of participants with various levels of familiarity sat inside the Mutual Wave Machine-an artistic neurofeedback installation that translates real-time correlations of each pair's EEG activity into light patterns. Because such inter-participant EEG correlations are prone to noise contamination, in subsequent offline analyses we computed inter-brain coupling using Imaginary Coherence and Projected Power Correlations, two synchrony metrics that are largely immune to instantaneous, noise-driven correlations. When applying these methods to two subsets of recorded data with the most consistent protocols, we found that pairs' trait empathy, social closeness, engagement, and social behavior (joint action and eye contact) consistently predicted the extent to which their brain activity became synchronized, most prominently in low alpha (similar to 7-10 Hz) and beta (similar to 20-22 Hz) oscillations. These findings support an account where shared engagement and joint action drive coupled neural activity and behavior during dynamic, naturalistic social interactions. To our knowledge, this work constitutes a first demonstration that an interdisciplinary, real-world, crowdsourcing neuroscience approach may provide a promising method to collect large, rich datasets pertaining to real-life face-to-face interactions. Additionally, it is a demonstration of how the general public can participate and engage in the scientific process outside of the laboratory. Institutions such as museums, galleries, or any other organization where the public actively engages out of self-motivation, can help facilitate this type of citizen science research, and support the collection of large datasets under scientifically controlled experimental conditions. To further enhance the public interest for the out-of-the-lab experimental approach, the data and results of this study are disseminated through a website tailored to the general public (wp.nyu.edu/mutualwavemachine).