Activity in perirhinal and entorhinal cortex predicts perceived visual similarities among category exemplars with highest precision
- Brain and Mind Institute, University of Western Ontario, Canada
- Robarts Research Institute Schulich School of Medicine and Dentistry, University of Western Ontario, Canada
- Cervo Brain Research Center, University of Laval, Canada
- Department of Psychology, Florida State University, United States
- Department of Physiology and Pharmacology, University of Western Ontario, Canada
- School of Biomedical Engineering, University of Western Ontario, Canada
- Department of Medical Biophysics, University of Western Ontario, Canada
- Department of Psychology, University of Western Ontario, Canada
Figures
Figure 1 with 2 supplements
Perceived visual similarity ratings obtained offline with inverse multidimensional scaling (iMDS).
(A) Task required placement of all exemplars from each category in circular arena, with distances reflecting perceived visual similarity. Arrows indicate the six pairwise distances used to compute …
Figure 1—figure supplement 1
Stimuli employed and behavioural data obtained for specific object categories.
(A) Forty images of object stimuli sampled from Migo et al., 2013 and employed in current study. Images of objects are from 10 categories with 4 exemplars in each category. Feature dimensions on …
Figure 1—figure supplement 2
In a follow-up study, a separate group of 30 participants completed 2 sessions of the inverse multidimensional scaling (iMDS) task for the 10 object categories separated by 7 ± 1 days.
(A) Intrasubject correlations for perceived similarity ratings across all exemplars and categories. Correlations were computed between each participants’ perceived similarity representational …
Figure 2
fMRI task: Category-Exemplar 1-Back Task.
(A) Images of objects depicting one of the 4 exemplars from 10 different categories were presented. Participants indicated repetitions on catch trials with two different button presses depending on …
Figure 3 with 1 supplement
Brain-based representational dissimilarity matrices (RDMs) and their relationship to perceived visual similarity.
(A) In each region of interest (ROI), mean multivoxel activation patterns were calculated for every exemplar using the no-response trials in the Category-Exemplar 1-Back Task. Pairwise pattern …
Figure 3—figure supplement 1
Temporal signal-to-noise ratio in perirhinal cortex (PrC; M = 12.25, SD = 2.68) and anterolateral entorhinal cortex (alErC; M = 10.80, SD = 3.05) in each of the 25 participants (average denoted by bolded red dot).
Green arrows indicate two participants with tSNR in alErC below 2 SD of the mean. These participants were excluded from fMRI analyses.
Figure 4 with 1 supplement
Relationship between brain representational dissimilarity matrices (RDMs) and reports at different levels of perceived visual similarity for region of interests (ROIs) showing significant effects in Figure 3D.
(A) Correlation of brain-based RDM and participants’ own behaviour-based RDM at low, medium, and high levels of similarity. Only activation patterns perirhinal cortex (PrC) and anterolateral …
Figure 4—figure supplement 1
Relationship between brain-based representational dissimilarity matrices (RDMs) in even and odd runs, within and between participants at different levels of perceived similarity.
Correlation of brain-based RDMs for even and odd runs at low, medium, and high levels of similarity. Activation patterns in all region of interests (ROIs) show significant correlations at all levels …
Figure 5
Visualization of region of interests (ROIs) and results from whole-volume searchlight analyses.
(A, B) Visual depiction of ROIs. Early visual cortex (EVC = green), lateral occipital complex (LOC = cyan), perirhinal cortex (PrC = pink), parahippocampal cortex (PhC = orange), anterolateral …
Figure 6
Brain-based representational dissimilarity matrices (RDMs) and their relationship to average perceived visual similarity.
Brain-based RDMs were correlated with (A) the average behaviour-based similarity RDMs and (B) the different levels of average similarity RDMs; and with (C) the entire RDM derived from the HMAX model …
Figure 7 with 2 supplements
Brain-based representational dissimilarity matrices (RDMs) and their relationship to observer-specific perceived visual similarity.
(A) Brain-based RDMs were correlated with (1) participants’ own behaviour-based similarity RDMs (black double arrows = within subject r) and (2) other participants’ behaviour-based similarity RDMs …
Figure 7—figure supplement 1
Comparison between inferotemporal (IT) cortex and lateral occipital complex (LOC): brain-based representational dissimilarity matrices (RDMs) and their relationship to perceived visual similarity (A) at all levels of perceived similarity (**p < 0.01, *p < 0.05) and (B) corresponding i-index; and (C) at different levels of perceived similarity and (D) corresponding i-index (*p < 0.05).
These four analyses from the main experiment were included here to compare results in IT cortex with those in LOC. IT cortex is a large swath of cortex that extends across occipital and temporal …
Figure 7—figure supplement 2
Relationship between brain-based representational dissimilarity matrices (RDMs) in even and odd runs, within and between participants.
(A) Pairwise pattern dissimilarity percentiles were used to create observer-specific brain-based RDMs for even and odd runs. (B) Brain-based RDMs from even runs were correlated with (1) observers’ …
Additional files
-
Supplementary file 1
Behavioural performance on Category-Exemplar 1-Back Task.
Proportion of correct responses for each trial type are indicated in green.
- https://cdn.elifesciences.org/articles/66884/elife-66884-supp1-v2.docx
-
Supplementary file 2
Multiple linear regression: brain RDM ~ (own RDM + average RDM).
- https://cdn.elifesciences.org/articles/66884/elife-66884-supp2-v2.docx
-
Transparent reporting form
- https://cdn.elifesciences.org/articles/66884/elife-66884-transrepform1-v2.pdf
