Adaptive recruitment of cortex-wide recurrence for visual object recognition (fMRI)

# Dataset Description

This dataset contains the raw fMRI data accompanying the study
**"Adaptive recruitment of cortex-wide recurrence for visual object recognition"** (Link to preprint: https://www.biorxiv.org/content/10.1101/2025.10.17.682937v2).
Please cite the above paper if you use this data.

---

## Dataset Overview
- 30 participants, each with 2 sessions
- Each session: 8–10 task runs and 1 functional localizer run
- Subjects 1–3 do not have a second localizer
- `anat/`: T1-weighted anatomical scans
- `func/`: whole-brain EPI data with event files specifying onsets, durations, and trial types

---

## Experimental Design
The fMRI experiment used a final stimulus set of 242 images (121 “challenge” and 121 “control” images) derived from comparisons between human behavioral performance and AlexNet.

- **Main task:** Each image was presented for 500 ms followed by a 2500 ms blank. Participants responded to occasional fixation cross color changes (catch trials).
- **Functional localizer:** Block design with fixation, objects, scrambled objects, houses, and faces.

---

## Derivatives
The `derivatives/` folder contains outputs from GLM analyses:
- Subject-level beta estimates in each participant’s individual volume space
- Deformation fields for each participants that allow transforming the subject-level betas estimates to MNI space.
- Object-identity representational dissimilarity matrices (RDMs)
- ROI-based decoding accuracy matrices for object identity
- Searchlight decoding accuracy matrices for object identity