================================================================================
BIDS Dataset: Reduced NEXI protocol for the quantification of human gray matter 
microstructure on the Connectome 2.0 scanner
================================================================================

1. OVERVIEW
-----------
This dataset contains in vivo MRI data from healthy human adults acquired to map 
gray matter microstructure using Neurite Exchange Imaging (NEXI). Data were 
collected on an ultra-high gradient 3T Siemens MAGNETOM Connectom.X scanner 
(Connectome 2.0) to evaluate an Explainable AI (XAI)-optimized 14-minute protocol.

2. STUDY PARTICIPANTS & DESIGN
------------------------------
- Number of subjects: 8 healthy adults (mean age: 29.6 ± 3.9 years).
- Sessions & Design: 
  - 7 participants underwent two identical scan sessions (scan-rescan) to evaluate 
    test-retest reproducibility using the standard 2 mm protocol.
  - 1 participant underwent a single high-resolution session (1.6 mm isotropic) 
    without a rescan session.
- Total files: 651 files across 8 subjects and 2 sessions (~18.1 GB).

3. ACQUISITION PROTOCOL (SUMMARY)
---------------------------------
All data were acquired on a MAGNETOM Connectom.X 3T scanner (500 mT/m gradients, 
600 T/m/s slew rate):
- Anatomical Reference: High-resolution T1-weighted MPRAGE (0.9 mm isotropic).
- Diffusion-Weighted Imaging (dMRI): Single-shot spin-echo EPI sequence featuring 
  an XAI-optimized 8-feature subset (b-values up to 12.5 ms/µm², multiple diffusion 
  times).
- Reverse phase-encoding b0 volumes were included for susceptibility distortion 
  correction.

4. RELATED PUBLICATION
----------------------
If you use these data or reference this dataset, please cite the associated publication:
Uhl, Q., Pavan, T., Gerold, J., Chan, K.-S., Jun, Y., Fujita, S., Bhatt, A., 
Ma, Y., Wang, Q., Lee, H.-H., Huang, S. Y., Bilgic, B., & Jelescu, I. (2026).
"Reduced NEXI protocol for the quantification of human gray matter microstructure 
on the Connectome 2.0 scanner." Imaging Neuroscience. 
https://doi.org/10.1162/IMAG.a.1380

5. CODE AVAILABILITY
--------------------
Processing scripts, model fitting routines, and the XGBoost-SHAP-RFE optimization 
pipeline used for this dataset are publicly available at:
https://github.com/Mic-map/graymatter_swissknife
https://github.com/QuentinUhl/XAI-dMRI-protocol-optimization