Deception is a human behavior and its cognitive process and mechanism involve complex neuronal activities of the brain. In this article, we develop a simple and feasible concealed information test (CIT) method which is based on the audio–visual event-related potentials (ERPs) and its spatial and temporal features. The main purpose of this article is to extend a pattern recognition method with functional network parameters and global feature entropy of the EEG signals from the whole brain. At the same time, a novel quantum neural network (QNN) classifier was developed to distinguish the guilty and innocent conditions. Functional connectivity can provide extra information of interdependence between different brain regions from the spatial dimension, and entropy can reflect the complexity of the whole brain from the temporal dimension. 20 subjects participated in the CIT experiment and 30 channel ERPs were recorded. A high accuracy of 87.67% was got in recognizing the concealed information, which was higher than 85.43% for basic features, demonstrated the effectiveness of this article. Future studies should further clarify the connectivity difference and further improve the accuracy of the QNN classifier for CIT.

中文翻译：

---

基于功能脑连通性和视听ERP信号熵的隐蔽信息测试系统

欺骗是人类的一种行为，其认知过程和机制涉及大脑复杂的神经元活动。在本文中，我们开发了一种基于视听事件相关电位 (ERP) 及其时空特征的简单可行的隐藏信息测试 (CIT) 方法。本文的主要目的是扩展一种具有功能网络参数和全脑脑电信号全局特征熵的模式识别方法。同时，开发了一种新颖的量子神经网络（QNN）分类器来区分有罪和无罪的情况。功能连接可以从空间维度提供不同大脑区域之间相互依赖的额外信息，而熵可以从时间维度反映整个大脑的复杂性。20 名受试者参加了 CIT 实验，记录了 30 个通道 ERP。识别隐藏信息的准确率为87.67%，高于基本特征的85.43%，证明了本文的有效性。未来的研究应进一步阐明连接差异，并进一步提高 CIT 的 QNN 分类器的准确性。