Update nibabel to latest version, fix int error and add scaled thresh…

…old when masking

setup.py

@@ -41,7 +41,7 @@
         'importlib-metadata ~= 4.0 ; python_version < "3.8"',
         "numpy>=1.21",
         "phantominator~=0.6.4",
-        "nibabel~=3.2.1",
+        "nibabel>=3.2.1",
         "requests",
         "scipy>=1.7",
         "tqdm",

shimmingtoolbox/cli/mask.py

@@ -1,13 +1,13 @@
 #!/usr/bin/env python3
 
 import click
+import logging
 import nibabel as nib
 import numpy as np
 import os
-import logging
 
-import shimmingtoolbox.masking.threshold
 from shimmingtoolbox.masking.shapes import shape_square, shape_cube, shape_sphere
+import shimmingtoolbox.masking.threshold
 from shimmingtoolbox.utils import run_subprocess, create_output_dir, set_all_loggers
 
 CONTEXT_SETTINGS = dict(help_option_names=['-h', '--help'])
@@ -55,7 +55,7 @@ def box(fname_input, output, size, center, verbose):
     if len(data.shape) == 3:
         mask_cb = shape_cube(data, size[0], size[1], size[2], center[0], center[1], center[2])  # creation
         # of the box mask
-        mask_cb = mask_cb.astype(int)
+        mask_cb = mask_cb.astype(np.int32)
         nii_img = nib.Nifti1Image(mask_cb, nii.affine)
         nib.save(nii_img, output)
         click.echo(f"The filename for the output mask is: {os.path.abspath(output)}")
@@ -114,7 +114,7 @@ def rect(fname_input, output, size, center, verbose):
             # on each slice (2D)
             mask_sqr[:, :, z] = mask_slice  # addition of each masked slice to form a 3D array
 
-        mask_sqr = mask_sqr.astype(int)
+        mask_sqr = mask_sqr.astype(np.int32)
         nii_img = nib.Nifti1Image(mask_sqr, nii.affine)
         nib.save(nii_img, output)
         click.echo(f"The filename for the output mask is: {os.path.abspath(output)}")
@@ -151,7 +151,7 @@ def sphere(fname_input, fname_output, radius, center, verbose):
     # API for sphere mask
     mask = shape_sphere(nii.get_fdata(), radius, center[0], center[1], center[2])
 
-    nii_mask = nib.Nifti1Image(mask.astype(int), affine=nii.affine, header=nii.header)
+    nii_mask = nib.Nifti1Image(mask.astype(np.int32), affine=nii.affine, header=nii.header)
     nib.save(nii_mask, fname_output)
 
 
@@ -165,10 +165,14 @@ def sphere(fname_input, fname_output, radius, center, verbose):
 @click.option('-o', '--output', type=click.Path(), default=os.path.join(os.curdir, 'mask.nii.gz'), show_default=True,
               help="Name of output mask. Supported extensions are .nii or .nii.gz. (default: "
                    "(os.curdir, 'mask.nii.gz'))")
-@click.option('--thr', default=30, help="Value to threshold the data: voxels will be set to zero if their "
-                                        "value is equal or less than this threshold. (default: 30)")
+@click.option('--thr', type=click.FLOAT, default=30.0, show_default=True,
+              help="Value to threshold the data: voxels will be set to zero if their value is equal or less than this"
+                   " threshold.")
+@click.option('--scaled-thr', is_flag=True, default=False, show_default=True,
+              help="Indicate whether the --thr option is scaled from 0 to 1 (True) or according to the values within "
+                   "the input (False). Default: False")
 @click.option('-v', '--verbose', type=click.Choice(['info', 'debug']), default='info', help="Be more verbose")
-def threshold(fname_input, output, thr, verbose):
+def threshold(fname_input, output, thr, scaled_thr, verbose):
 
     # Set all loggers
     set_all_loggers(verbose)
@@ -179,8 +183,8 @@ def threshold(fname_input, output, thr, verbose):
     nii = nib.load(fname_input)
     data = nii.get_fdata()  # convert nifti file to numpy array
 
-    mask_thr = shimmingtoolbox.masking.threshold.threshold(data, thr)  # creation of the threshold mask
-    mask_thr = mask_thr.astype(int)
+    mask_thr = shimmingtoolbox.masking.threshold.threshold(data, thr, scaled_thr)  # creation of the threshold mask
+    mask_thr = mask_thr.astype(np.int32)
     nii_img = nib.Nifti1Image(mask_thr, nii.affine)
     nib.save(nii_img, output)
     click.echo(f"The filename for the output mask is: {os.path.abspath(output)}")

test/cli/test_cli_b0shim.py

@@ -56,7 +56,7 @@ def _define_inputs(fmap_dim):
                   center_dim2=int(ny / 2),
                   len_dim1=10, len_dim2=10, len_dim3=nz - 10)
 
-    nii_mask = nib.Nifti1Image(mask.astype(int), nii_anat.affine)
+    nii_mask = nib.Nifti1Image(mask.astype(np.uint8), nii_anat.affine)
 
     return nii_fmap, nii_anat, nii_mask, fm_data, anat_data
 
@@ -1087,7 +1087,7 @@ def test_b0_max_intensity():
                       center_dim1=32,
                       center_dim2=36,
                       len_dim1=10, len_dim2=10, len_dim3=nz)
-        nii_mask = nib.Nifti1Image(mask.astype(int), nii.affine)
+        nii_mask = nib.Nifti1Image(mask.astype(np.uint8), nii.affine)
         nib.save(nii_mask, fname_mask)
 
         runner = CliRunner()

test/cli/test_cli_image.py

@@ -55,8 +55,8 @@ def test_cli_concat_pixdim(self):
 class TestImageLogicalAnd(object):
     def setup_method(self):
         affine_1 = np.eye(4)
-        self.nii1 = nib.Nifti1Image(np.array([[[1, 1], [1, 1]], [[0, 0], [0, 0]]]), affine=affine_1)
-        self.nii2 = nib.Nifti1Image(np.array([[[0, 0], [1, 1]], [[1, 1], [0, 0]]]), affine=affine_1)
+        self.nii1 = nib.Nifti1Image(np.array([[[1, 1], [1, 1]], [[0, 0], [0, 0]]], dtype=np.uint8), affine=affine_1)
+        self.nii2 = nib.Nifti1Image(np.array([[[0, 0], [1, 1]], [[1, 1], [0, 0]]], dtype=np.uint8), affine=affine_1)
 
         mask_3 = np.ones([8, 8])
         mask_3[0, 0] = 0

test/cli/test_cli_mask.py

@@ -88,7 +88,31 @@ def test_cli_mask_threshold():
 
         out = os.path.join(tmp, 'mask.nii.gz')
         thr = 780
-        result = runner.invoke(mask_cli, ['threshold', '--input', inp, '--output', out, '--thr', thr])
+        result = runner.invoke(mask_cli, ['threshold', '--input', inp, '--output', out, '--thr', thr],
+                               catch_exceptions=False)
+
+        # With a threshold value of 780, the expected mask for the positions [58:62, 28:31, 7:9] is :
+        expected = np.array([[[1.0, 1.0], [0.0, 1.0], [0.0, 0.0]],
+                             [[1.0, 1.0], [0.0, 1.0], [0.0, 0.0]],
+                             [[0.0, 1.0], [0.0, 0.0], [0.0, 0.0]],
+                             [[0.0, 1.0], [0.0, 0.0], [0.0, 0.0]]])
+
+        nii = nib.load(out)
+        mask = nii.get_fdata()
+
+        assert result.exit_code == 0
+        assert np.all(mask[58:62, 28:31, 7:9] == expected)
+
+
+def test_cli_mask_threshold_scaled():
+    with tempfile.TemporaryDirectory(prefix='st_' + pathlib.Path(__file__).stem) as tmp:
+        runner = CliRunner()
+
+        out = os.path.join(tmp, 'mask.nii.gz')
+        thr = 0.412
+        result = runner.invoke(mask_cli, ['threshold', '--input', inp, '--output', out, '--thr', thr,
+                                          '--scaled-thr'],
+                               catch_exceptions=False)
 
         # With a threshold value of 780, the expected mask for the positions [58:62, 28:31, 7:9] is :
         expected = np.array([[[1.0, 1.0], [0.0, 1.0], [0.0, 0.0]],

test/cli/test_cli_realtime_shim.py

@@ -1,11 +1,13 @@
 #!usr/bin/env python3
 # coding: utf-8
+
+import json
+import numpy as np
 import os
 import pathlib
+import pytest
 import tempfile
 import nibabel as nib
-import pytest
-import json
 
 from click.testing import CliRunner
 from shimmingtoolbox.cli.realtime_shim import gradient_realtime
@@ -40,7 +42,7 @@ def test_cli_realtime_shim():
                       center_dim2=int(ny / 2),
                       len_dim1=30, len_dim2=30, len_dim3=nz)
 
-        nii_mask_static = nib.Nifti1Image(mask.astype(int), nii_anat.affine)
+        nii_mask_static = nib.Nifti1Image(mask.astype(np.int16), nii_anat.affine)
         fname_mask_static = os.path.join(tmp, 'mask.nii.gz')
         nib.save(nii_mask_static, fname_mask_static)
 
@@ -50,7 +52,7 @@ def test_cli_realtime_shim():
                       center_dim2=int(ny / 2),
                       len_dim1=20, len_dim2=20, len_dim3=nz)
 
-        nii_mask_riro = nib.Nifti1Image(mask.astype(int), nii_anat.affine)
+        nii_mask_riro = nib.Nifti1Image(mask.astype(np.uint8), nii_anat.affine)
         fname_mask_riro = os.path.join(tmp, 'mask.nii.gz')
         nib.save(nii_mask_riro, fname_mask_riro)
 

test/shim/test_sequencer.py

@@ -693,7 +693,7 @@ def save_nii(nii_fieldmap, coil, opt, nii_mask):
     nib.save(nii_coil, fname_coil_res)
 
 
-array = np.array([[1, 2], [3, 4]])
+array = np.array([[1, 2], [3, 4]], dtype=np.uint8)
 array = np.repeat(array, 4, 1)
 array = np.repeat(array[..., np.newaxis], 1, 2)
 array = np.repeat(array[..., np.newaxis], 5, 3)
@@ -836,7 +836,7 @@ def setup_method(self):
                       center_dim1=32,
                       center_dim2=36,
                       len_dim1=10, len_dim2=10, len_dim3=nz)
-        self.nii_mask = nib.Nifti1Image(mask.astype(int), self.nii_input.affine)
+        self.nii_mask = nib.Nifti1Image(mask.astype(np.uint8), self.nii_input.affine)
 
     def test_default_max_intensity(self):
         output = shim_max_intensity(self.nii_input, self.nii_mask)

test/test_load_nifti.py

@@ -18,13 +18,13 @@
 class TestCore(object):
     _data = np.array([[[1, 2, 3], [4, 5, 6], [7, 8, 9]],
                       [[10, 11, 12], [13, 14, 15], [16, 17, 18]],
-                      [[19, 20, 21], [22, 23, 24], [25, 26, 27]]])
+                      [[19, 20, 21], [22, 23, 24], [25, 26, 27]]], dtype=np.uint8)
     _data_volume = np.array([[[[1, 1], [2, 2], [3, 3]], [[4, 4], [5, 5], [6, 6]], [[7, 7], [8, 8], [9, 9]]],
                              [[[10, 10], [11, 11], [12, 12]], [[13, 13], [14, 14], [15, 15]],
                               [[16, 16], [17, 17], [18, 18]]],
                              [[[19, 19], [20, 20], [21, 21]], [[22, 22], [23, 23], [24, 24]],
-                              [[25, 25], [26, 26], [27, 27]]]])
-    _aff = np.array([[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12], [13, 14, 15, 16]])
+                              [[25, 25], [26, 26], [27, 27]]]], dtype=np.uint8)
+    _aff = np.array([[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12], [13, 14, 15, 16]], dtype=np.uint8)
     _json_phase = {"Modality": "MR",
                    "ImageComments": "phase",
                    "MagneticFieldStrength": 3,
@@ -151,8 +151,8 @@ def setup_method(self):
         self.data_path_2.mkdir(exist_ok=True)
         self.data_path_volume.mkdir(exist_ok=True)
 
-        dummy_data = nib.nifti1.Nifti1Image(dataobj=self._data, affine=self._aff)
-        dummy_data_volume = nib.nifti1.Nifti1Image(dataobj=self._data_volume, affine=self._aff)
+        dummy_data = nib.Nifti1Image(self._data, affine=self._aff)
+        dummy_data_volume = nib.Nifti1Image(self._data_volume, affine=self._aff)
 
         nib.save(dummy_data, os.path.join(self.data_path, 'dummy.nii'))
         with open(os.path.join(self.data_path, 'dummy.json'), 'w') as json_file: