Support channeled dicom slices

dicom_numpy/combine_slices.py

@@ -75,27 +75,27 @@ def combine_slices(slice_datasets, rescale=None):
 
 
 def _merge_slice_pixel_arrays(slice_datasets, rescale=None):
-    first_dataset = slice_datasets[0]
-    num_rows = first_dataset.Rows
-    num_columns = first_dataset.Columns
-    num_slices = len(slice_datasets)
-
     sorted_slice_datasets = _sort_by_slice_position(slice_datasets)
 
     if rescale is None:
         rescale = any(_requires_rescaling(d) for d in sorted_slice_datasets)
 
-    if rescale:
-        voxels = np.empty((num_columns, num_rows, num_slices), dtype=np.float32, order='F')
-        for k, dataset in enumerate(sorted_slice_datasets):
+    first_dataset = sorted_slice_datasets[0]
+    slice_dtype = first_dataset.pixel_array.dtype
+    slice_shape = first_dataset.pixel_array.T.shape
+    num_slices = len(sorted_slice_datasets)
+
+    voxels_shape = slice_shape + (num_slices,)
+    voxels_dtype = np.float32 if rescale else slice_dtype
+    voxels = np.empty(voxels_shape, dtype=voxels_dtype, order='F')
+
+    for k, dataset in enumerate(sorted_slice_datasets):
+        pixel_array = dataset.pixel_array.T
+        if rescale:
             slope = float(getattr(dataset, 'RescaleSlope', 1))
             intercept = float(getattr(dataset, 'RescaleIntercept', 0))
-            voxels[:, :, k] = dataset.pixel_array.T.astype(np.float32) * slope + intercept
-    else:
-        dtype = first_dataset.pixel_array.dtype
-        voxels = np.empty((num_columns, num_rows, num_slices), dtype=dtype, order='F')
-        for k, dataset in enumerate(sorted_slice_datasets):
-            voxels[:, :, k] = dataset.pixel_array.T
+            pixel_array = pixel_array.astype(np.float32) * slope + intercept
+        voxels[..., k] = pixel_array
 
     return voxels
 
@@ -136,6 +136,7 @@ def _validate_slices_form_uniform_grid(slice_datasets):
         'SeriesInstanceUID',
         'Rows',
         'Columns',
+        'SamplesPerPixel',
         'PixelSpacing',
         'PixelRepresentation',
         'BitsAllocated',

tests/conftest.py

@@ -11,6 +11,7 @@
 negative_z_cos = (0, 0, -1)
 
 arbitrary_shape = (10, 11)
+arbitrary_rgb_shape = (10, 11, 3)
 
 
 class MockSlice:
@@ -28,15 +29,21 @@ def __init__(self, pixel_array, slice_position, row_cosine=None, column_cosine=N
         if column_cosine is None:
             column_cosine = y_cos
 
-        na, nb = pixel_array.shape
+        shape = pixel_array.shape
+        if len(shape) == 2:
+            na, nb = shape
+            SamplesPerPixel = 1
+        else:
+            na, nb, SamplesPerPixel = shape
 
         self.pixel_array = pixel_array
 
         self.SeriesInstanceUID = 'arbitrary uid'
         self.SOPClassUID = 'arbitrary sopclass uid'
         self.PixelSpacing = [1.0, 1.0]
-        self.Rows = na
-        self.Columns = nb
+        self.Columns = na
+        self.Rows = nb
+        self.SamplesPerPixel = SamplesPerPixel
         self.Modality = 'MR'
 
         # assume that the images are centered on the remaining unused axis
@@ -63,5 +70,16 @@ def axial_slices():
     ]
 
 
+@pytest.fixture
+def axial_rgb_slices():
+    # SamplesPerPixel = 3
+    return [
+        MockSlice(randi(*arbitrary_rgb_shape), 0),
+        MockSlice(randi(*arbitrary_rgb_shape), 1),
+        MockSlice(randi(*arbitrary_rgb_shape), 2),
+        MockSlice(randi(*arbitrary_rgb_shape), 3),
+    ]
+
+
 def randi(*shape):
     return np.random.randint(1000, size=shape, dtype='uint16')

tests/test_combine_slices.py

@@ -30,6 +30,12 @@ def test_single_slice_spacing(self, axial_slices):
         assert np.array_equal(array, dataset.pixel_array.T[:, :, None])
         assert np.isclose(np.linalg.norm(affine[:, 2]), np.abs(slice_spacing))
 
+    def test_rgb_axial_set(self, axial_rgb_slices):
+        combined, _ = combine_slices(axial_rgb_slices)
+
+        manually_combined = np.stack([ds.pixel_array for ds in axial_rgb_slices], axis=0).T
+        assert np.array_equal(combined, manually_combined)
+
 
 class TestMergeSlicePixelArrays:
     def test_casting_if_only_rescale_slope(self):