| 其他摘要 | The emotion of fear has important implications for human adaptation and evolution. However, excessive fear can lead to a series of mental illnesses. How to treat fear-related diseases is an urgent problem to be solved clinically. Evidence from a large number of studies shows that compared with neutral, non-emotional information, the brain is less dependent on the state of consciousness when processing emotional information, and can even proceed in a unconscious state. Not only that, the researchers found that fear conditioning can also be successfully established in the subconscious state, and its development process is significantly different from that of conscious fear conditioning. However, the neural mechanism of unconscious fear conditioning is still unclear. In this study, three studies are used to explore the process of unconscious fear acquisition caused by sensory changes and its neural mechanism, and compare it with the process of conscious fear acquisition. The three studies are described as follows:
In Study 1，we designed an unconscious fear conditioning task. The experimental stimulus was two colored gratings with different orientations, presented as low spatial frequency(LSF)and high spatial frequency(HSF)in Study la and Study 1 b, respectively. The orientation of gratings is made invisible using the Critical Flicker Frequency (CFF) paradigm. One of the colored gratings was bound to fear conditioning with negative electrical stimulation (CS+), the other was always presented alone (CS一). Pupil data were recorded when the subjects completed the task. In this study, we recruited 46 healthy subjects (34 were finally included in the analysis). It was found that even when a neutral stimulus that has nothing to do with fear was used as a conditioned stimulus, a significant difference in the conditioned responses of CS+ and CS一could be observed in unconscious fear conditioning, and this difference occurred specifically in the LSF and not in the case of HSF. This provides strong evidence that neutral stimuli can be effectively associated with the US in the unconscious state, while the results support the dependence of unconscious fear processing on LSF.
In Study 2, we used the within-group design method, and a total of 11 epilepsy patients were required to complete the unconscious and conscious fear conditioning successively. The intracranial electroencephalography(iEEG)data and pupil data were simultaneously recorded during the task. Our results found that, at different levels of consciousness, the different responses of the human amygdala to CS+ and CS一all occurred within a very fast latency period (about 60 ms), earlier than the processing of emotional information and the separation of visual features by inferior temporal cortex (ITC). Further analysis found that activity in the amygdala precedes the ITC early in the presentation of subthreshold stimuli, and shifts to ITC earlier than the amygdala in the later stages. These results are consistent with the idea that rapid subthreshold detection of threatening stimuli by the amygdala is modulated by information from subcortical pathways that bypass the visual cortex, thus providing direct empirical support for the 'low-road' idea. At the same time, we also found that there are significant differences in the patterns of event-related potential (ERP) at different levels of consciousness, suggesting that the two may involve different neural mechanisms.
In Study 3, our experimental design was consistent with Study 2 and a total of 10 epilepsy patients' data were included in the analysis二It was found that under different levels of consciousness, both the amygdala and the ventral medial prefrontal cortex (vmPFC) showed significant changes in theta oscillations, and the functional connectivity of the two brain regions was enhanced. This result provides the first direct electrophysiological evidence for the involvement of the amygdala and vmPFC in unconscious fear conditioning. In addition, we found that during unconscious fear learning, the amygdala exhibited enhanced low theta power and decreased high theta power; whereas the vmPFC exhibited enhanced high and low theta oscillations during unconscious and conscious tasks, respectively. This suggests that the low theta and high theta oscillation of the amygdala may play different roles in unconscious fear conditioning, while different frequency bands in theta oscillations of the vmPFC are involved in unconscious and conscious fear processing, respectively. Furthermore, our results found that electrodes located in the ventral posterior vmPFC were more active when CS+ was presented, while electrodes located in the dorsal anterior vmPFC were more active when CS-. This provides key evidence support that different subregion of the vmPFC play different roles in fear learning. Analysis of directionality in two brain regions found that during unconscious fear learning, this synchronicity was initially driven by the amygdala and later gradually shifted to be driven by the vmPFC. This result suggests that during fear acquisition, the amygdala relies on projections to the vmPFC to shape activity in the vmPFC, resulting in enhanced vmPFC output. At the same time, we found that in the conscious fear conditioning, the information transmission between the amygdala and vmPFC was driven by the vmPFC in the early stage of stimulus presentation, but gradually changed to be driven by the amygdala in the later stage of stimulus presentation. This may be due to the heavy use of attentional resources required for conscious tasks. Finally, we found that during conscious fear learning, low theta power in the amygdala and mPFC showed a tendency to increase significantly with learning progress. In the unconscious task, there is no significant change in theta power.
The above findings suggest that unconscious fear conditioning based on neutral stimuli can be established relying on LSF, and during this process, the synchronized activity of the amygdala-mPFC circuit in the theta band promotes fear learning. Although the amygdala-mPFC circuit plays an important role in fear learning at different levels of consciousness, the mechanism of these two is different, which is reflected in many variables involving discrete patterns of oscillatory activity, different learning processes and opposite direction of functional connectivity in amygdala-mPFC circuit. These works will help us better understand the formation of fear memories, and provide some reference for future intervention and treatment of fear diseases.
The above findings suggest that unconscious fear conditioning based on neutral stimuli can be established relying on LSF. In fear learning at different levels of consciousness, the different responses of the human amygdala to CS+ and CS一all occurred at very fast latencies, earlier than processing of emotional information and separation of simple visual features in the ITC. The synchronized activity between the amygdala and mPFC in the theta band promotes the establishment of fear conditioning, but there are different neural mechanisms at different levels of consciousness, which are reflected in the oscillation patterns within and between brain regions, the process of learning and the directionality of information transmission across brain regions. These works will help us better understand the formation of fear memory, and provide some reference for the intervention and treatment of phobia. |
修改评论