Gonuguntla V, Kim Jae-Hun
Annu Int Conf IEEE Eng Med Biol Soc. 2020 Jul;2020:2853-2856. doi: 10.1109/EMBC44109.2020.9175397.
In more recent times, the network perspective study of the human brain has expanded enormously due to the advancements in the field of network neuroscience. Existing methods to form the functional connectivity from the multichannel EEG leads to a fully connected network. Representation of a fully connected functional network with a significant functional network (SFN) can help to characterize and quantify the complex brain networks. Further, it can also provide novel insight into the brain cognition analysis and is crucial in several brain network-based applications. This paper presents a framework to find the SFN corresponding to any event from its fully connected network. Using the phase-locking value (PLV) in EEG we first identify the difference PLV of an event to the rest. Based on the difference PLV, we then identify the reactive band and event-associated most reactive pairs (MRPs). The SFNs corresponding to their events is then formed with the identified MRPs. The proposed method is employed on 'database for emotion analysis using physiological signals (DEAP)' data set to find the SFNs associated with emotions. Comparable state-of-the-art multiple emotion classification accuracies are obtained with the identified SFNs. Results show that the proposed methods can be used as a general thresholding technique to identify the event-related SFNs which are crucial in brain network-based applications.
近年来，由于网络神经科学领域的进展，对人类大脑的网络视角研究得到了极大扩展。从多通道脑电图形成功能连接的现有方法会导致一个全连接网络。用一个显著功能网络（SFN）来表示全连接功能网络有助于刻画和量化复杂的脑网络。此外，它还能为脑认知分析提供新的见解，并且在一些基于脑网络的应用中至关重要。本文提出了一个从全连接网络中找到与任何事件对应的SFN的框架。我们首先利用脑电图中的锁相值（PLV）确定一个事件与其余部分的差异PLV。基于差异PLV，我们接着确定反应带和与事件相关的最活跃对（MRP）。然后用识别出的MRP形成与其事件对应的SFN。所提出的方法应用于“使用生理信号进行情感分析的数据库（DEAP）”数据集，以找到与情感相关的SFN。通过识别出的SFN获得了可比的当前最先进的多情感分类准确率。结果表明，所提出的方法可作为一种通用的阈值技术，用于识别在基于脑网络的应用中至关重要的与事件相关的SFN。
