Abstract
Self-prioritization is a ubiquitous psychological phenomenon that occurs when prioritized processing is linked to information about oneself compared to information about others. Recent research has identified specific brain regions
that underpin self-prioritization processing. The dynamic characteristics of the region connections that induce such effects, however, remain unknown. To address this issue, we investigated neural couplings when participants carried out a standardized shape-label matching task, which has been used to investigate self-prioritization effects in reaction times and accuracy, while
electroencephalogram (EEG) data were recorded. Behaviorally, the self-prioritization effect was present in faster and more accurate responses to shape-label pairings associated with oneself compared to those associated with a friend or a stranger. Deep learning models were used to test and validate the top-down
regulation and feedforward phases of neural couplings during self-prioritization processing. The analysis based on SqueezeNet revealed enhanced accuracy performance for self-related stimuli during the early top-down regulation phase compared to stimuli associated with others, but this improvement was not observed during the later feedforward processing. Moreover, the analyses
showed better performance in classifying the self-related stimuli during the early top-down phase compared to the later feedforward phase; this pattern, however, was not observed in classifying stranger-related stimuli. These results indicate that deep learning analyses can provide valuable insights about self functions in information processing that would otherwise be difficult to test
using traditional neuroscience methods.
that underpin self-prioritization processing. The dynamic characteristics of the region connections that induce such effects, however, remain unknown. To address this issue, we investigated neural couplings when participants carried out a standardized shape-label matching task, which has been used to investigate self-prioritization effects in reaction times and accuracy, while
electroencephalogram (EEG) data were recorded. Behaviorally, the self-prioritization effect was present in faster and more accurate responses to shape-label pairings associated with oneself compared to those associated with a friend or a stranger. Deep learning models were used to test and validate the top-down
regulation and feedforward phases of neural couplings during self-prioritization processing. The analysis based on SqueezeNet revealed enhanced accuracy performance for self-related stimuli during the early top-down regulation phase compared to stimuli associated with others, but this improvement was not observed during the later feedforward processing. Moreover, the analyses
showed better performance in classifying the self-related stimuli during the early top-down phase compared to the later feedforward phase; this pattern, however, was not observed in classifying stranger-related stimuli. These results indicate that deep learning analyses can provide valuable insights about self functions in information processing that would otherwise be difficult to test
using traditional neuroscience methods.
| Original language | English |
|---|---|
| Number of pages | 12 |
| Publication status | Accepted/In press - 12 Aug 2023 |
| Event | IEEE ICCI*CC 2023: 2023 IEEE 22nd International Conference on Cognitive Informatics and Cognitive Computing - Stanford University, Palo Alto , United States Duration: 19 Aug 2023 → 21 Aug 2023 Conference number: 22nd https://easychair.org/cfp/IEEE_ICCI-CC_2023 |
Conference
| Conference | IEEE ICCI*CC 2023 |
|---|---|
| Country/Territory | United States |
| City | Palo Alto |
| Period | 19/08/23 → 21/08/23 |
| Internet address |
Keywords
- self-prioritization
- EEG
- neural coupling
- deep learning