| Deception is viewed as an undesirable behavior in the human society, which can not only prey on the cooperative and unsuspecting, but also dissolve community spirit. Neverthless, deception is ubiquitous in our daily life, occuring quite frequently in interpersonal relationships with various forms. Therefore, there are abundant scientists from a variety of disciplines show interests in its cognitive and neural mechanisms, especially, being eager to find reliable ways by which to detect lies. fMRI techology provides a noninvasive way to investigate the neural basis underlying deception, based on which feasible neural markers could be found to detect lie. However, so far, fMRI studies in this area only focused on the process of conducting deception, and correspondingly, used brain activation patterns associated with deception conducting as markers to detect lies. However, as a complex social phenomenon, deception relates to various psychological components, and only cooperating with these psychological components could deception happen. In this perspective, breaking aways from these psychological processes to study deception independently could not uncover the deception in its essentials. More importantly, theses deception-related psychological processes might also elicit some specific physiological or neural signals, which might become potential new clues to speculate the truth. With this thought, the present study investigated four psychological processes related to deception, including deception trait that is the inner incentive of deception, the memory trace that forms the factual basis of deception, the conducting processes that executes deception, and finally the consequence that is the external feedback to deceptive behaviors, using RSFC and fMRI methods combined with pattern recognition technologies such as discriminant analysis and SVM classifier. By these studies, we not only attempted to reveal the neural mechanism underlying these processes, but also tried to test feasibilities and effectiveness of typical neural singals elicited by each psychological process in detecting lies.Study1investigated the neural basis of deception trait using the RSFC method. Experiment la compared the RSFC pattern of individuals who behaved dishonestly while confronting with opportunities for dishonest gain with that of individuals who behaved honestly in this situation to reveal the neural basis of deception trait. Based on the RSFC patterns associated with deception trait found by experiment la, experiment lb performed the discriminant analysis with cross-validation method to test how correctly could these RSFC patterns predict one individual’s dishonest behavior in the face of temptation. The experiment la revealed that dishonest subjects demonstrated significantly weaker inverse AMG-TPJ and AMG-1OFC couplings, and furthermore, the strength of inverse coupling between AMG and1OFC was significantly and negatively correlated with rejective attitude for deception measured by a feeling thermometer (FT) questionnaire. Considering that AMG is involved in reward seeking while TPJ and1OFC are the key nodes underlying guilt aversion, we speculated that the strength of inverse AMG-TPJ and AMG-10FC couplings reflect the degree to which the impulsion of seeking reward is downregulated by guilt aversion potentially supported by TPJ and1OFC, and thus our findings suggested that the dishonest individuals may reflect an impaired capacity to feel guilt. Experiment2a revealed that, on the basis of the strength of AMG-TPJ and AMG-10FC couplings, we were able to correctly predict32of the39subjects (82.1%) as who would behave dishonestly.Study2investigated the neural basis of memory trace during deceiving using the fMRI method. In the experiment2a, blood-oxygen-level-dependent (BOLD) responses in the brains of17healthy volunteers were recorded while they honestly and deceptively respond to a recognition memory task. The results showed that the bilateral pMTL were activated to larger extent by old words than by new words regardless of whether the participants answered honestly or deceptively about their oldness, demonstrating that the pMTL might objectively and stably store memory trace about stimulus, which would not be affected by the subjective effort to restrain it. Based on acitivities in bilateral pMTL showed by experiment2a, experiment2b used SVM classifier to test whether one participant had memory trace about old words. The results showed that, on the basis of the combined neural signals of bilateral pMTL, we could obtain classification accuracy equal to79.45%while distinguishing old words from new words.Study3investigated the neural mechanism underlying conducting deception using the fMRI method. As this process has been abundantly investigated by previous studies, the present stud)’focused on how the neural signals elicited by deceptive response and their effectiveness in detecting lie would be modulated by whether response would be judged by the lie-detector about its facticity. In the experiment3a, after a mock murder situation, BOLD responses in the brains of31healthy volunteers were recorded while they underwent two versions of GKT paradigm. In the GKT paradigm for the feedback group, a feedback indicating whether the response was judged honest or deceptive was given by the lie detector after participant’s each response, while in the GKT paradigm for the no-feedback group, there was no feedback after their responses. The results showed that the feedback group was associated with increased activity in dorsal anterior cingulate cortex, suggesting an additional mentalising process. Furthermore, deceptive response in the feedback groups was associated with increased activities in insula and striatum. suggesting additional risk evaluation and reward predication processes. In the experiment3b. based on the activities of regions associated with deceptive response revealed by experiment3a. we performed SVM to test how accurate could these neural signals classify deceptive trails from honest trials for two groups, respectively. The results revealed that, for both groups, IPL was the most contributing area for correct classification, based on which we could obtain classification accuracy equal to95.3%for the no-feedback group and91.8%for the feedback group. The classification accuracy was significantly higher for the no-feedback group than for the feedback group. Therefore, these results demonstrate the important influence of social consequence in deception from both mechanism and application perspectives.Study4used fMRI method to investigate the neural mechanism of processing the consequence of deception and test how correctly could the neural signals underlying consequence processing detect lies.In the experiment4. we used fMRI to reveal how the human brain processes the accuracy and the valence of judgments. BOLD responses in the brains of14healthy volunteers were recorded while they encountered four possible judgments by the justice system with two basic components:whether the judgment is right or wrong [accuracy: right vs. wrong (misjudgment)] and wl;ether the judgment is positive or negative [valence:positive vs. negative]. The results showed that the rostral ACC specifically processes the accuracy of judgment, being more active for misjudgment than for right judgment, while the striatum was uniquely responsible for the valence of judgment, being recruited to a larger extent by positive judgment compared to negative judgment. Further, confirming the misjudgment-specificity of rACC, the activity in rACC for positive misjudgment was positively correlated with that for negative misjudgment. These results demonstrate that the brain can distinguish misjudgment from right judgment and regard it as an emotionally arousing stimulus independent of whether the judgment is positive or negative, while positive judgment is considered as hedonic information no matter whether it is right or wrong. This study first reveled the neural mechanism underlying judgment processing, and might inspire future studies to develop a new marker for speculating truth.In the experiment5, we used the same mock murder situation and GKT paradigm as those used by the study3for the feedback group, however, this experiment focused on the fMRI result of feedback stage to investigate how the brain processes the judgment from lie-detector after conducting deceptive response. The results showed that judgment processing involves several brain regions, including STR that was sensitive to the valence of judgment, suggesting its role in reward processing;1OFC that was sensitive to accuracy of judgment, suggesting its role in adaptive learning, especially in processing unexpected outcomes; mPFC and TPJ that were also more active for wrong judgment than for right judgment, suggesting their roles in false belief attribution. Furthermore, all of above regions showed significantly stronger activation to the judgment regard deceptive response compared with judgment regard honest responses, suggesting that the external judgment has important implication for participants after they conducted deceptive response, and thus they showed hypersensitivity to it. Therefore, this study revealed the neural basis of a mental phenomena sounds "A guilty conscience is a self-accuser".In the experiment6, we also used the same mock murder situation and GKT paradigm as those used by experiment5, however, we recruited two groups of participants, the16participants of one group were assigned to be "murderer" and underwent a mock murder situation, while16participants of the other group were assigned to be "innocent" and did not undergo any mock crime. Subsequently, they underwent the same GKT paradigm while the BOLD responses in their brains were recorded. We investigated the brain activation patterns associated with deceptive response and those associated with judgment processing, respectively, and separately computed how correctly could the neural signals associated with each of the two processes to correctly identify participants as who were murderers. The results showed that, among the brain areas associated with deceptive responses, the right VLPF contributed mostly to individual identification with sensitivity of81.3%and specificity of81.3%. Whereas, among the brain areas associated with judgment processing, the right TPJ contributed mostly to individual identification with sensitivity of93.75%and specificity of87.5%. Therefore, this experiment first documented that the neural activity associated with being a successful liar (or not) is a feasible indicator for detecting lies and may be more valid than neural activity associated with producing deception, which has been commonly used by recent studies.Taken together, this study is the first to reveal the neural mechanisms and possible relationships among four psychological components related to deception, including deception trait, memory trace, deception conducting and consequence processing. In light of the present results, we proposed a four components and reciprocal influence model for understanding the specific role of each of these psychological components plays in deception and possible cooperative relationships among them. Furthermore, based on the present results, we found several new neural signals that can directly or indirectly marking deception, which might implicate future study to develop new methods to detect lies and even develop a potential method to measure loyalty. |
PDF Full Text Request