search

mmuckley / torchkbnufft

A high-level, easy-to-deploy non-uniform Fast Fourier Transform in PyTorch.
https://torchkbnufft.readthedocs.io/
MIT License
204 stars 44 forks source link

Can the current package handle 2x oversampling data along readout direction？ #4

Closed gongjizhang closed 4 years ago

gongjizhang commented 4 years ago

Can the current package handle 2x oversampling data along readout direction？ For example，if readout point of each spoke is 512（2x oversampling included），image size is 256，grid size is 384。

mmuckley commented 4 years ago

Yeah, it should be able to handle any settings you want. This example is for a 400x400 image with 2x oversampling:

https://github.com/mmuckley/torchkbnufft/blob/master/notebooks/Basic%20Example.ipynb

You can set grid_size to (384, 384) when you initialize your NUFFT object. Just make sure your k-space coordinates are correct in radians/voxel.

gongjizhang commented 4 years ago

Ok thanks for the feedback.