The codes in python just exposes adj_op, op and sense recon.
However, the core CUDA codes can be modified to support some generally just gridding and interpolation of kspace data.
This is of particular use in density estimation for density compensation.
I have written some partial codes in https://github.com/chaithyagr/gpuNUFFT/tree/add_dens_comp, however I dont call it clean yet. I will get a PR after having cleaned it.
Going forward, I would want to add support to density estimation if possible.
The codes in python just exposes adj_op, op and sense recon. However, the core CUDA codes can be modified to support some generally just gridding and interpolation of kspace data. This is of particular use in density estimation for density compensation.
I have written some partial codes in https://github.com/chaithyagr/gpuNUFFT/tree/add_dens_comp, however I dont call it clean yet. I will get a PR after having cleaned it. Going forward, I would want to add support to density estimation if possible.