Added codes

src/mri/cli/dc_adjoint.py

@@ -3,13 +3,13 @@
 from mrinufft.io import read_trajectory
 from mri.operators import NonCartesianFFT
 from mri.operators.fourier.utils import estimate_density_compensation
+from mri.cli.utils import save_data
 from mrinufft.io.nsp import read_arbgrad_rawdat
 from mrinufft.io.utils import add_phase_to_kspace_with_shifts
 from mrinufft.extras.utils import get_smaps
 from pymrt.recipes.coils import compress_svd
 
 import numpy as np
-import pickle as pkl
 import logging
 import os
 
@@ -51,13 +51,23 @@
     implementation="gpuNUFFT",
     name="gpu",
 )
+fourier_store(
+    fourier_op_config,
+    upsampfac=1,
+    implementation="gpuNUFFT",
+    name="gpu_lowmem",
+)
+
 smaps_store = store(group="fourier/smaps")
 smaps_store(smaps_config, name="low_frequency")
 density_store = store(group="fourier/density_comp")
 density_store(density_est_config, implementation="pipe", name="pipe")
+density_store(density_est_config, implementation="pipe", osf=1, name="pipe_lowmem")
 
 
-def recon(obs_file: str, traj_file: str, obs_reader, traj_reader, fourier, coil_compress: str|int = -1):
+
+def recon(obs_file: str, traj_file: str, obs_reader, traj_reader, fourier, coil_compress: str|int = -1,
+          output_filename: str = "dc_adjoint.pkl"):
     """
     Reconstructs an image using the adjoint operator.
 
@@ -78,7 +88,13 @@ def recon(obs_file: str, traj_file: str, obs_reader, traj_reader, fourier, coil_
     coil_compress : str|int, optional default -1
         The number of singular values to keep in the coil compression.
         If -1, coil compression is not applied 
-    
+    output_filename: str, optional default 'dc_adjoint.pkl'
+        The output file name with the right extension.
+        It can be:
+        1) *.pkl / *.mat: Holds the reconstructed results saved in dictionary as `recon`.
+            #TODO: Add scope for debug by saving intermediate results also in output.
+        2) *.nii : NIFTI file holding the reconstructed images.
+        
     Returns
     -------
     None
@@ -117,7 +133,7 @@ def recon(obs_file: str, traj_file: str, obs_reader, traj_reader, fourier, coil_
     recon = fourier_op.adj_op(kspace_data)
     if not fourier_op.uses_sense:
         recon = np.linalg.norm(recon, axis=-1)
-    pkl.dump(recon, open("dc_adjoint.pkl", "wb"))
+    save_data(output_filename, recon, data_header)
 
 store(
     recon,
@@ -131,6 +147,20 @@ def recon(obs_file: str, traj_file: str, obs_reader, traj_reader, fourier, coil_
     ],
     name="dc_adjoint",
 )
+store(
+    recon,
+    obs_reader=raw_config,
+    traj_reader=traj_config,
+    hydra_defaults=[
+        "_self_",
+        {"fourier": "gpu_lowmem"},
+        {"fourier/density_comp": "pipe_lowmem"},
+        {"fourier/smaps": "low_frequency"},
+    ],
+    name="dc_adjoint_lowmem",
+)
+
+
 
 store.add_to_hydra_store()
 zen(recon).hydra_main(

src/mri/operators/fourier/non_cartesian.py

@@ -22,7 +22,7 @@
 class NonCartesianFFT(OperatorBase):
     """This class wraps around different implementation algorithms for NFFT"""
     def __init__(self, samples, shape, implementation='finufft', n_coils=1,
-                 density_comp=None, smaps=None, **kwargs):
+                 density_comp=None, smaps=None, upsampfac: float|int = 2, **kwargs):
         """ A small wrapper around mri-nufft package
         This is mostly maintained just for legacy reasons (all legacy reconstruction
         uses this codes)
@@ -44,6 +44,8 @@ def __init__(self, samples, shape, implementation='finufft', n_coils=1,
         density_comp: np.ndarray (M,) default None
         smaps: np.ndarray, optional, default None
             The sensitivity maps.
+        upsampfac: float|int, optional, default 2
+            The oversampling factor to be used
         kwargs: extra keyword args
             these arguments are passed to underlying operator. Check the docs of 
             mrinufft : https://mind-inria.github.io/mri-nufft/
@@ -60,6 +62,7 @@ def __init__(self, samples, shape, implementation='finufft', n_coils=1,
             density=self.density_comp,
             n_coils=self.n_coils,
             smaps=smaps,
+            upsampfac=upsampfac,
             **self.kwargs
         )
 

src/mri/operators/fourier/utils/processing.py

@@ -264,7 +264,8 @@ def check_if_fourier_op_uses_sense(fourier_op):
         return False
 
 
-def estimate_density_compensation(kspace_loc, volume_shape, implementation='pipe', **kwargs):
+def estimate_density_compensation(kspace_loc, volume_shape, implementation='pipe',
+                                  osf: float|int = 2, **kwargs):
     """ Utils function to obtain the density compensator for a
     given set of kspace locations.
 
@@ -277,13 +278,16 @@ def estimate_density_compensation(kspace_loc, volume_shape, implementation='pipe
     implementation: str default 'pipe'
         the implementation of the non-cartesian operator
         can be 'pipe' which needs gpuNUFFT or 'cell_count'
+    osf: float|int default 2
+        the oversampling factor used in estimation
     kwargs: dict
         extra keyword arguments to be passed to the density
         compensation estimation
     """
     density_comp = get_density(implementation)(
         kspace_loc,
         volume_shape,
+        osf=osf,
         **kwargs
     )
     return np.squeeze(density_comp)