UCLA IMMEX Project, 5601 W. Slauson Avenue #272, Culver City, CA 90230, USA.
Hum Factors. 2012 Aug;54(4):489-502. doi: 10.1177/0018720811427296.
Cognitive neurophysiologic synchronies (NS) are low-level data streams derived from electroencephalography (EEG) measurements that can be collected and analyzed in near real time and in realistic settings. The objective of this study was to relate the expression of NS for engagement to the frequency of conversation between team members during Submarine Piloting and Navigation (SPAN) simulations.
If the expression of different NS patterns is sensitive to changes in the behavior of teams, they may be a useful tool for studying team cognition.
EEG-derived measures of engagement (EEG-E) from SPAN team members were normalized and pattern classified by self-organizing artificial neural networks and hidden Markov models. The temporal expression of these patterns was mapped onto team events and related to the frequency of team members' speech. Standardized models were created with pooled data from multiple teams to facilitate comparisons across teams and levels of expertise and to provide a framework for rapid monitoring of team performance.
The NS expression for engagement shifted across task segments and internal and external task changes.These changes occurred within seconds and were affected more by changes in the task than by the person speaking.Shannon entropy measures of the NS data stream showed decreases associated with periods when the team was stressed and speaker entropy was high.
These studies indicate that expression of neurophysiologic indicators measured by EEG may complement rather than duplicate communication metrics as measures of team cognition.
Neurophysiologic approaches may facilitate the rapid determination of the cognitive status of a team and support the development of novel adaptive approaches to optimize team function.
认知神经生理同步（NS）是从脑电图（EEG）测量中提取的低水平数据流，可以在接近实时的情况下在现实环境中进行收集和分析。本研究的目的是将参与的 NS 表达与潜艇驾驶和导航（SPAN）模拟过程中团队成员之间对话的频率相关联。
如果不同 NS 模式的表达对团队行为的变化敏感，它们可能是研究团队认知的有用工具。
通过自组织人工神经网络和隐马尔可夫模型对 SPAN 团队成员的 EEG 衍生的参与度测量值（EEG-E）进行归一化和模式分类。这些模式的时间表达被映射到团队事件上，并与团队成员讲话的频率相关联。使用来自多个团队的 pooled 数据创建标准化模型，以促进团队之间以及不同专业水平的比较，并为团队绩效的快速监测提供框架。
参与的 NS 表达在任务段之间以及内部和外部任务变化中发生了变化。这些变化在几秒钟内发生，并且更多地受到任务变化的影响，而不是说话者的影响。NS 数据流的香农熵测量值显示出与团队紧张时期和高说话者熵相关的下降。
这些研究表明，通过 EEG 测量的神经生理指标的表达可以补充而不是重复通信指标，作为团队认知的度量。
神经生理方法可以促进团队认知状态的快速确定，并支持开发新的自适应方法来优化团队功能。
