Respiratory-triggered MRCP acquisition at 3T using a T2-weighted TSE (3D SPACE) sequence for Deep Learning-based reconstruction of MRCP

Description This dataset is the sample data for MRCP_DLRecon (GitHub). Place this dataset in the "Sample_data/" directory along with the "dataset.csv" file. Data The provided 3D MRCP data were acquired at 3T (Skyra, Siemens Healthineers AG, Erlangen) using the 3D-SPACE (3D T2w TSE) sequence for a single healthy volunteer. We provide two 2x and one 6x 3D MRCP data to ensure various training and testing scenarios. Each data in the HDF5 format contains the following structures: Datasets grappa: target data (y * x * slice) kdata_raw: Raw k-space data (x2 or x6) (nCoil * PE * RO * slice) kdata_fs: Fully-sampled k-space data from kdata_raw using GRAPPA (nCoil ×× PE ×× RO ×× Slice) sm_espirit: ESPIRiT-based sensitivity maps (nCoil * y * x * slice) Citation Please cite the following paper if this dataset is helpful for your research :) Kim, J., Nickel, M. and Knoll, F. (2025), Deep Learning-Based Accelerated MR Cholangiopancreatography Without Fully-Sampled Data. NMR in Biomedicine, 38: e70002. https://doi.org/10.1002/nbm.70002