Are the methods for acquiring the training dataset Cartesian, radial, or any other type? How do you get the fully sampled k-space? What is the difference between training and validation datasets from the acquisition perspective?
Thank you.
Created by Md Abrar Akib aia3921 All our data was collected using Cartesian acquisition. The actual k-space data was obtained using low acceleration factors and reconstructed using GRAPPA. We then retrospectively undersampled to release the data for the challenge.
In terms of acquisition patterns, there is no difference among the training, validation, and subsequent test datasets.