This repository supports the manuscript:
Differential representations for affective and informative components of reward in the striatum and hippocampus
The study uses high-resolution task fMRI and representational similarity analysis to examine affective and informative reward representations in subcortical regions, especially dorsal/ventral striatum and anterior/posterior hippocampus.
The manuscript reports functional BOLD acquisition at 3T with single-shot T2*-weighted EPI, GRAPPA R=2, TR=2000 ms, TE=28 ms, matrix 128 x 128, FOV 204 mm, and 1.75 x 1.75 mm in-plane voxels. The high-resolution acquisition is important for the paper's central goal: resolving representational structure within relatively small subcortical regions.
psychtoolbox/: MATLAB/Psychtoolbox task code.docs/prereg/: preregistration figures.analyses/openneuro_resolution_benchmark/: reproducible OpenNeuro analysis used to contextualize the study's BOLD spatial resolution for reviewer response and future figures.
The OpenNeuro benchmark is intentionally organized as a supporting analysis, not as the identity of this repository. It scans raw public BOLD NIfTI headers from OpenNeuro without full-file downloads under normal operation, aggregates to one observation per dataset x unique spatial-resolution triplet, and generates figures/tables for comparing this study's acquisition with public datasets.
See:
analyses/openneuro_resolution_benchmark/README.md
The next analysis target is to stratify OpenNeuro BOLD resolution by acquisition type, especially likely single-band/conventional EPI versus multiband/SMS datasets. That distinction matters because multiband acceleration can support different spatial/temporal tradeoffs, while the reviewer concern is most directly addressed by comparing against regular single-band task fMRI studies.
The benchmark and reviewer-response figures are being developed here first. Once the figures and text are settled, the relevant materials can be merged into Karen Shen's repository:
https://github.com/KarenShen21/ru-highres