Generalized data stitching for dynamic field monitoring with NMR probes: improving acquisition efficiency and measurement accuracy

Summary This dataset contains the experimental raw data and Pulseq sequence files supporting the manuscript "Generalized data stitching for dynamic field monitoring with NMR probes: improving acquisition efficiency and measurement accuracy." The proposed generalized data stitching framework enables high-fidelity dynamic field monitoring for challenging high-resolution readouts (0.5 mm) with prolonged durations (~90 ms) at 7 T. Key Contributions Enhanced Efficiency: Reduced the number of measurement segments by more than 10-fold compared to the original dephasing-based approach (e.g., 16 vs. 187 segments for spiral; 13 vs. 142 segments for EPI). Time Reduction: Achieved over 90% reduction in total measurement time for field characterization. Improved Accuracy: Eliminated residual artifacts and distortions by accurately capturing zeroth-order field dynamics. Hardware Compatibility: Fully compatible with commercially available NMR field monitoring hardware, such as the Skope Dynamic Field Camera. Dataset Structure 1. Probe_Data Contains raw NMR probe signal data and high-order phase coefficients (up to 3rd order). Includes validation datasets for 2D Spiral and 2D EPI readouts. Compares Standard single-measurement, Dephasing-based stitching, and the Proposed Generalized Stitching methods. 2. MRI_Data Raw in-vivo human brain MR data acquired at 7 T in ISMRMRD (.mrd) format. In_Vivo_Scans: High-resolution (0.5 mm) single-shot 2D Spiral (R=4) and EPI (R=5, 7/8 PF) brain data. Both sequences were acquired at the isocenter as a 2D axial slice. Calibration_Scan: Fully-sampled multi-echo GRE data (1.0 mm) for coil sensitivity (ESPIRiT) and static delta-B0 mapping. 3. Pulse_Sequences The hardware-independent Pulseq (.seq) files used for acquisitions, including those for field monitoring calibration and in-vivo imaging. Includes specific sequence variants tailored for different monitoring and stitching strategies (e.g., standard, dephasing-based, and generalized stitching). Features high-resolution (0.5 mm) single-shot 2D Spiral and EPI sequences developed within the Pulseq framework. Code Availability MATLAB Code (Data Stitching & Segmentation): https://github.com/BennyZhang-Codes/DataStitching Pulseq Sequences: https://github.com/XiaopingWu2020/pulseq-sequences Julia Code (Image Reconstruction): Included in the DataStitching repository. Acknowledgments This work was supported in part by: National Natural Science Foundation of China: 82271985. Brain Science and Brain-like Intelligence Technology – National Science and Technology Major Project: 2022ZD0211900, 2022ZD0211901. Strategic Priority Research Program of the CAS: XDB0930000. USA NIH Grants: R01 NS136490, P41 EB027061.