Poster No:
2561
Submission Type:
Abstract Submission
Authors:
Yuxuan Zeng1, Ren Hentz1, David Osher1
Institutions:
1The Ohio State University, Columbus, OH
First Author:
Co-Author(s):
Ren Hentz
The Ohio State University
Columbus, OH
Introduction:
Visual processing of most stimuli involves feature-based visual processing (where stimuli are defined by the presence or absence of specific elements such as shape) whereas other stimulus classes like faces involve configural processing (information about relative locations of individual features e.g. interocular distance). Very little is known about the neural mechanisms that underlie feature and configural processing. We designed stimuli that vary based on either featural properties alone, or configural properties alone, and asked participants to learn to categorize them into groups through trial and error (Zeng et al., 2022). After participants learned to accurately categorize these stimuli in the lab, they performed the same task while fMRI data were collected. By contrasting blocks of featural stimuli against configural stimuli, we identified a mid-level visual region that is selectively recruited while participants extract stimulus features.
Methods:
Participants were first trained to categorize stimuli into two groups, under two conditions: configural and feature-based, each comprising stimuli defined solely by configural or featural information. Participants had to achieve over 80% accuracy in both conditions in order to continue to the fMRI task. Six participants underwent scanning in a Siemens Prisma 3 Tesla MRI with a 64-channel headcoil, acquiring a high-resolution T1-weighted image (1mm isotropic, TR = 2300 ms, TE = 2.9 ms), and task-based fMRI (2mm isotropic, TR = 2000 ms, TE = 29 ms). We collected a 2-run localizer for high-level visual categories, including faces, bodies, object, scenes, and scrambled objects (Pitcher et al., 2011). We also collected 8 runs of fMRI while participants viewed stimuli and categorized them based on the previously learned categorization rules, in a block design that alternated between featural and configural stimuli. Data were processed with Freesurfer 7.2.0.
Results:
We identified a bilateral region in the ventral visual stream that is specifically recruited while participants are extracting stimulus features. This region lies on the ventral surface just beyond areas V4/V8, and extends to the posterior fusiform gyrus. Interestingly, this region is not selective for any high-level visual category. It responds to faces, bodies, and scenes, and strongest to objects and even scrambled objects (object selectivity is typically defined by the contrast of objects > scrambled). In addition, we identified several visual fROIs, and compared configural and featural responses in each fROI. Only the bilateral posterior fusiform sulcus (an object selective region) showed a significant difference between these conditions, preferring featural over configurally processed stimuli.
Conclusions:
Our findings reveal a region in the ventral visual stream associated with feature-based processing that does not overlap with other category-specific areas. We propose this could be an intermediary region that extracts featural information before stimulus identity is determined. Future work will focus on its connectivity with the rest of the ventral visual stream in order to investigate how this region interacts with visual processing at early and late stages of the perceptual process.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 2
Perception, Attention and Motor Behavior:
Perception: Visual 1
Keywords:
FUNCTIONAL MRI
Vision
Other - Visual Features
1|2Indicates the priority used for review
Provide references using author date format
Pitcher, D. (2011), 'Differential selectivity for dynamic versus static information in face-selective cortical regions', Neuroimage, vol. 56, no. 4, pp. 2356-2363
Zeng, Y. (2023), 'A novel framework to study configural and holistic processing', Journal of Vision, vol. 23, no. 9, pp. 5471-5471