# Hippocampal Reactivation Trades Episodic Detail for Semantic Gist in Human Memory

## Overview

This dataset accompanies Hindy, Coleman, and Selvam (in press, Journal of Cognitive Neuroscience). Twenty-four participants encoded auditory DRM word lists spoken in male or female voices, creating orthogonal semantic (list theme) and episodic (voice source) information. Using high-resolution fMRI, we tracked list-specific reactivation in hippocampal subfields during a 90-second post-encoding delay. CA1 reactivation frequency showed opposing relationships with subsequent memory: greater reactivation predicted increased false recognition of semantic lures (gist extraction) and decreased accuracy for voice discrimination (source memory).

## Task Description

Each of five fMRI runs contained three phases:

1. **Encoding**: Five 6-word DRM lists presented auditorily (male/female synthetic voices, 3 words per voice per list, randomly intermixed). Each list occupied 24 s with 8-s inter-list intervals. Participants indicated word voice via button press.
2. **Delay**: 90-second post-encoding period with self-paced odd/even parity judgments on single digits.
3. **Test**: Recognition confidence (6-point scale) for targets, gist lures, and foils; source voice discrimination (male/female) with 3-point confidence for items endorsed as "old."

## Participants

- N = 24 (16 female, 8 male)
- Age range: 19-31 years
- Recruited from the University of Louisville community
- Compensated $20/hour
- No participants excluded

## MRI Acquisition

All data were collected on a 3T Siemens Skyra with a 20-channel head coil at the University of Louisville.

### Functional (task-drm)
T2*-weighted multi-band EPI (CMRR sequence), TR = 1.6 s, TE = 30 ms, flip angle = 66 deg, voxel size = 1.745 x 1.745 x 1.9 mm, 63 axial slices (interleaved), 110 x 110 matrix, 192 mm FOV, multiband acceleration factor 3, in-plane GRAPPA factor 4. Partial coverage focused on medial temporal lobes. 280 volumes per run, 5 runs per subject. Each run began with 8 s of rest to approach steady-state magnetization.

### Diffusion (dwi)
Single-shell diffusion-weighted EPI (ep2d_diff), TR = 6.8 s, TE = 94 ms, flip angle = 90 deg, voxel size = 2.5 mm isotropic, 49 axial slices (interleaved), 96 x 96 matrix, 240 mm FOV, b = 2000 s/mm2 with 50 diffusion directions and 6 interspersed b=0 volumes per run (56 volumes total). 12 runs per subject (600 diffusion-weighted volumes total).

### T1w structural
MPRAGE, 0.8 mm isotropic.

### T2w structural
Turbo spin-echo, 0.44 x 0.44 x 1.5 mm. Used for hippocampal subfield segmentation with ASHS.

### Fieldmap
GRE magnitude and phase images collected at end of experiment for B0 distortion correction.

## Ethics

University of Louisville IRB Protocol #17.1065

## Contact

Nicholas C. Hindy
Department of Psychology
College of Charleston
hindync@cofc.edu

## License

CC0