MAINT: Update nitransforms pin

pyproject.toml

@@ -24,7 +24,7 @@ dependencies = [
   "dipy>=1.5.0",
   "joblib",
   "nipype>=1.5.1,<2.0",
-  "nitransforms>=22.0.0,<24",
+  "nitransforms>=25.0.0",
   "nireports",
   "numpy>=1.21.3",
   "nest-asyncio>=1.5.1",

scripts/compute_nufo.py

@@ -164,7 +164,7 @@ def main() -> None:
         _, brain_mask = median_otsu(dwi_data, vol_idx=[0])
 
     dwi_data_masked = dwi_data.copy()
-    dwi_data_masked[~brain_mask, :] = 0
+    dwi_data_masked[~brain_mask.astype(bool), :] = 0
 
     # Create a CSD model
     response, ratio = auto_response_ssst(

scripts/optimize_registration.py

@@ -125,7 +125,7 @@ async def train_coro(
 
             index = i * len(REFERENCES) + j
             moving_path = tmp_folder / f"test-{index:04d}.nii.gz"
-            (~xfm).apply(refnii, reference=refnii).to_filename(moving_path)
+            nt.resampling.apply(~xfm, refnii, reference=refnii).to_filename(moving_path)
 
             _kwargs = {"output_transform_prefix": f"conversion-{index:04d}", **align_kwargs}
 

src/nifreeze/data/base.py

@@ -35,6 +35,7 @@
 import nibabel as nb
 import numpy as np
 from nitransforms.linear import LinearTransformsMapping
+from nitransforms.resampling import apply
 from typing_extensions import Self, TypeVarTuple, Unpack
 
 Ts = TypeVarTuple("Ts")
@@ -265,7 +266,7 @@ def to_nifti(
                 datamoving = nb.Nifti1Image(frame[0], self.affine, self.datahdr)
                 # resample at index
                 resampled[..., i] = np.asanyarray(
-                    xform.apply(datamoving, order=order).dataobj,
+                    apply(xform, datamoving, order=order).dataobj,
                     dtype=self.dataobj.dtype,
                 )
 

src/nifreeze/data/pet.py

@@ -35,6 +35,7 @@
 import numpy as np
 from nibabel.spatialimages import SpatialImage
 from nitransforms.linear import Affine
+from nitransforms.resampling import apply
 from typing_extensions import Self
 
 from nifreeze.data.base import BaseDataset, _cmp, _data_repr
@@ -137,7 +138,7 @@ def set_transform(self, index: int, affine: np.ndarray, order: int = 3) -> None:
 
         # resample and update orientation at index
         self.dataobj[..., index] = np.asanyarray(
-            xform.apply(dmoving, order=order).dataobj,
+            apply(xform, dmoving, order=order).dataobj,
             dtype=self.dataobj.dtype,
         )
 

src/nifreeze/data/utils.py

@@ -28,7 +28,9 @@ def apply_affines(nii, em_affines, output_filename=None):
 
     for ii, bvecnii in enumerate(nb.four_to_three(nii)):
         xfms = nt.linear.Affine(em_affines[ii])
-        transformed_nii[..., ii] = np.asanyarray((~xfms).apply(bvecnii, reference=nii).dataobj)
+        transformed_nii[..., ii] = np.asanyarray(
+            nt.resampling.apply(~xfms, bvecnii, reference=nii).dataobj
+        )
 
     nii_t_img = nii.__class__(transformed_nii, nii.affine, nii.header)
 

src/nifreeze/model/_dipy.py

@@ -139,7 +139,7 @@ def fit(
         self,
         data: np.ndarray,
         gtab: GradientTable | np.ndarray,
-        mask: np.ndarray[bool, Any] | None = None,
+        mask: np.ndarray | None = None,
         random_state: int = 0,
     ) -> GPFit:
         """Fit method of the DTI model class

src/nifreeze/registration/ants.py

@@ -35,6 +35,7 @@
 import numpy as np
 from nipype.interfaces.ants.registration import Registration
 from nitransforms.linear import Affine
+from nitransforms.resampling import apply
 
 PARAMETERS_SINGLE_VALUE = {
     "collapse_output_transforms",
@@ -494,7 +495,7 @@ def _run_registration(
         ),
     )
     # debugging: generate aligned file for testing
-    xform.apply(moving_path, reference=fixed_path).to_filename(
+    apply(xform, moving_path, reference=fixed_path).to_filename(
         dirname / f"dbg_{vol_idx:05d}.nii.gz"
     )
 

test/conftest.py

@@ -113,13 +113,13 @@ def motion_data(tmp_path_factory, datadir):
     )
 
     # Induce motion into dataset (i.e., apply the inverse transforms)
-    moved_nii = (~xfms).apply(b0nii, reference=b0nii)
+    moved_nii = nt.resampling.apply(~xfms, b0nii, reference=b0nii)
 
     # Save the moved dataset for debugging or further processing
     moved_path = tmp_path / "test.nii.gz"
     ground_truth_path = tmp_path / "ground_truth.nii.gz"
     moved_nii.to_filename(moved_path)
-    xfms.apply(moved_nii).to_filename(ground_truth_path)
+    nt.resampling.apply(xfms, moved_nii).to_filename(ground_truth_path)
 
     # Wrap into dataset object
     dwi_motion = DWI(

test/test_integration.py

@@ -49,10 +49,11 @@ def test_proximity_estimator_trivial_model(motion_data, tmp_path):
     )
 
     # Uncomment to see the realigned dataset
-    nt.linear.LinearTransformsMapping(
+    xfm = nt.linear.LinearTransformsMapping(
         dwi_motion.motion_affines,
         reference=b0nii,
-    ).apply(moved_nii).to_filename(tmp_path / "realigned.nii.gz")
+    )
+    nt.resampling.apply(xfm, moved_nii).to_filename(tmp_path / "realigned.nii.gz")
 
     # For each moved b0 volume
     for i, est in enumerate(dwi_motion.motion_affines):

test/test_registration.py

@@ -58,7 +58,7 @@ def test_ANTs_config_b0(datadir, tmp_path, dataset, r_x, r_y, r_z, t_x, t_y, t_z
     T = from_matvec(euler2mat(x=r_x, y=r_y, z=r_z), (t_x, t_y, t_z))
     xfm = nt.linear.Affine(T, reference=b0nii)
 
-    (~xfm).apply(b0nii, reference=b0nii).to_filename(moving)
+    nt.resampling.apply(~xfm, b0nii, reference=b0nii).to_filename(moving)
 
     registration = Registration(
         terminal_output="file",