Preprocessing data

[FredJones4]

Hi,

Could I get some help understanding how your phantom data is set up? I can see from the code that 5-D raw MRI data is needed, but I am not sure how to take the data setup (load 5-D complex64 .h5 dataset; [frames, slice, coils, height, width]) and create a transform as you describe.

In the dataset.py module, you describe transforms as follows:

transforms: Optional; A sequence of callable objects that
                preprocesses the raw data into appropriate form.
                The transform function should take 'kspace', 'kspace_ref', 'tmap_mask' and `metadata` as inputs.

What is an example of transforms that you use for your datset?

[minipuding]

Thank you for your question.

In fastmrt/data/transforms. py there have a FastmrtDataTransform2D class can help you understand how to build transforms.