CTI_Pulseq Tutorial Dataset: Raw Phantom Diffusion MRI Data for Correlation Tensor Imaging (CTI)

This dataset contains raw MRI scanner data (TWIX .dat format) from a Correlation Tensor Imaging (CTI) acquisition, provided as a small tutorial dataset for CTI_Pulseq — an open-source PyQt5 GUI covering the full CTI workflow: pulse sequence generation (Pulseq/pypulseq), image reconstruction, and CTI model fitting. What is CTI? Correlation Tensor Imaging is a diffusion MRI technique using double diffusion encoding (DDE) — two diffusion-sensitizing gradient pairs played within one echo — to decompose diffusional kurtosis into isotropic, anisotropic, and microscopic sources, going beyond conventional DTI/DKI (Henriques, Jespersen & Shemesh, 2021, NeuroImage). Acquisition protocol. A multi-slice, GRAPPA-accelerated, partial-Fourier spin-echo EPI readout, repeated across 4 diffusion-encoding "sets" that differ in how the total b-value is split between the two DDE gradient pairs and their relative orientation: Set b1 b2 Orientation Role 1 1500 s/mm² 0 — single diffusion encoding (SDE) reference 2 750 s/mm² 750 s/mm² parallel symmetric parallel DDE 3 750 s/mm² 750 s/mm² perpendicular symmetric perpendicular DDE 4 375 s/mm² 375 s/mm² parallel parallel DDE, lower b Each set: 45 diffusion directions, FOV 256 mm, 72×72 matrix (~3.6 mm in-plane resolution), 3× 5 mm slices, TR 10 s, TE ≈ 152 ms, partial-Fourier factor 0.75, GRAPPA R=2 (24 ACS lines), δ1=δ2=28 ms, Δ1=Δ2=34 ms. 48 repetitions per set (1 navigator + 2 b0 blip-up/down + 45 directions), ~8 minutes scan time per set. Files. Set_1.dat, Set_2.dat, Set_3.dat, Set_4.dat — one raw TWIX file per CTI set (~405 MB each, ~1.6 GB total), readable in Python via pymapvbvd. How to use. These files are meant to be dropped into CTI_Pulseq's Reconstruction tab (Tab 2) as a hands-on tutorial: generate the matching .seq files in Tab 1 (or use the ones bundled with the tutorial), load each Set_N.dat against its corresponding set, reconstruct, then run the CTI fit in Tab 3 to reproduce parametric maps (FA, MD, MK, K_iso, K_aniso, K_micro, K_total, etc.). See the project's instructions.txt for the full step-by-step walkthrough. Code repository: GitHub