Functional Profiling of Character Representation in Longitudinal Naturalistic fMRI
Poster No:
637
Submission Type:
Abstract Submission
Authors:
Yibei Chen1, Jefferey Mentch2, Steven Meisler3, Satra Ghosh1
Institutions:
1Massachusetts Institute of Technology, Cambridge, MA, 2Harvard University, Cambridge, MA, 3University of Pennsylvania, Philadelphia, PA
First Author:
Co-Author(s):
Introduction:
Humans possess remarkable abilities to recognize important or personally relevant individuals across diverse contexts, relying on social cognition mechanisms in the brain (Tso et al., 2018). How does the brain distinguish "main characters" from others and identify the brain regions involved in this differentiation? This study investigates these questions by examining the selectivity of brain activity and functional distinctions between six main characters (Chandler, Joey, Monica, Phoebe, Rachel, and Ross) and supporting characters in the sitcom Friends. Using a longitudinal naturalistic fMRI dataset, we explore the brain's character representations, functional composition, and the differentiation mechanisms over time.
Methods:
Dataset: Six participants (aged 31-47, 3 male) watched Friends (seasons 1-6) over 1000 days during fMRI acquisition (Boyle et al., 2023). Episodes were split into two 10-12 minute runs with a 10-second overlap for continuity, yielding 292 runs. Data preprocessing (fMRIPrep 20.2.5, long-term support) included denoising, z-scoring, and confound regression, with analyses performed on CIFTI outputs.
Character Annotation: Faces, rich in social information (Campanella & Belin, 2007), were tracked in each segment to timestamp the six main characters and supporting characters, excluding minor background appearances across all runs.
Functional Localizer: GLMsingle (Prince et al., 2022) was applied to Human Connectome Project tasks (10-15 runs) to identify regions linked to social and cognitive functions, including face perception, mentalizing, language, and relational reasoning. GLMsingle optimizes voxel responses, enhancing test-retest reliability.
Encoding Characters: Two ridge regression models were applied: (1) Character Representation (season 1): Correlations between real and predicted BOLD signals identified brain regions responsive to any character's presence. Top 10% correlations were thresholded and averaged to produce a reliability map. (2) Main-Supporting Contrast (seasons 2-6): One-tailed t-tests on regression coefficients distinguished brain activation differences between main and supporting characters, constrained by the Character Representation map.
Quantifying Functional Localizers: Overlaps between functional localizers and the character representation map were analyzed to identify shared and unique contributions, pinpointing regions associated with specific functions (Figure 1).
Character Annotation: Faces, rich in social information (Campanella & Belin, 2007), were tracked in each segment to timestamp the six main characters and supporting characters, excluding minor background appearances across all runs.
Functional Localizer: GLMsingle (Prince et al., 2022) was applied to Human Connectome Project tasks (10-15 runs) to identify regions linked to social and cognitive functions, including face perception, mentalizing, language, and relational reasoning. GLMsingle optimizes voxel responses, enhancing test-retest reliability.
Encoding Characters: Two ridge regression models were applied: (1) Character Representation (season 1): Correlations between real and predicted BOLD signals identified brain regions responsive to any character's presence. Top 10% correlations were thresholded and averaged to produce a reliability map. (2) Main-Supporting Contrast (seasons 2-6): One-tailed t-tests on regression coefficients distinguished brain activation differences between main and supporting characters, constrained by the Character Representation map.
Quantifying Functional Localizers: Overlaps between functional localizers and the character representation map were analyzed to identify shared and unique contributions, pinpointing regions associated with specific functions (Figure 1).
Results:
Our functional localizer results (Figure 1), thresholded at the top 10% t-values in one direction, align with patterns observed in Barch et al. (2013). Character representation involved cortical regions associated with face perception, mentalizing, language, relational reasoning, and additional subcortical regions. Notably: (1) The left anterior FFA exhibited selective activation for main characters, while posterior FFA regions were more active for supporting characters. (2) Bilateral STG showed significant clusters differentiating main and supporting characters. (3) Visual cortex exhibited greater activation for supporting characters, possibly reflecting broader visual processing demand. Results remained robust after normalizing coefficients by the frequency of character appearances (Figure 2).
Conclusions:
Longitudinal fMRI observations revealed distinct patterns of brain activity differentiating main and supporting characters, with prominent involvement of the FFA and STG. While canonical functional tasks explained part of these patterns, character-specific responses extended beyond regions traditionally associated with social, language, and face processing. This suggests a more intricate character-processing network that may incorporate unique functional dynamics or morphological adaptations in naturalistic settings compared to controlled tasks. Additionally, the role of subcortical regions and their connections with cortical areas warrants further exploration.
Emotion, Motivation and Social Neuroscience:
Social Cognition 1
Language:
Speech Perception
Novel Imaging Acquisition Methods:
BOLD fMRI
Perception, Attention and Motor Behavior:
Perception: Multisensory and Crossmodal 2
Keywords:
Cognition
Cortex
Data analysis
FUNCTIONAL MRI
Infections
Language
Modeling
Social Interactions
Vision
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Provide references using APA citation style.
Campanella et al. (2007). Integrating face and voice in person perception. Trends in cognitive sciences, 11(12), 535-543.
Boyle et al. (2023). The Courtois NeuroMod project: quality assessment of the initial data release (2020). In 2023 Conference on Cognitive Computational Neuroscience. Oxford, UK: Cognitive Computational Neuroscience (pp. 1602-0).
Prince et al. (2022). Improving the accuracy of single-trial fMRI response estimates using GLMsingle. Elife, 11, e77599. https://doi.org/10.7554/eLife.77599
Barch et al. (2013). Function in the human connectome: task-fMRI and individual differences in behavior. Neuroimage, 80, 169-189.