Merge pull request #400 from mgxd/fix/precomputed-mask-no-preproc

FIX: Do not force masking of anatomicals when using `--derivatives`

nibabies/workflows/anatomical/fit.py

@@ -554,7 +554,7 @@ def init_infant_anat_fit_wf(
             workflow.connect([
                 (t1w_buffer, ds_t1w_mask_wf, [('t1w_mask', 'inputnode.mask_file')]),
             ])  # fmt:skip
-    else:
+    else:  # T1w mask derivative has been found
         LOGGER.info('ANAT Found T1w brain mask')
         if reference_anat == 'T1w':
             desc += (
@@ -563,26 +563,19 @@ def init_infant_anat_fit_wf(
             )
         t1w_buffer.inputs.t1w_mask = t1w_mask
         apply_t1w_mask.inputs.in_mask = t1w_mask
-        workflow.connect(t1w_validate, 'out_file', apply_t1w_mask, 'in_file')
+        workflow.connect(apply_t1w_mask, 'out_file', t1w_buffer, 't1w_brain')
 
         if not t1w_preproc:
             LOGGER.info('ANAT Skipping skull-strip, INU-correction only')
-            t1w_n4_only_wf = init_n4_only_wf(
-                omp_nthreads=omp_nthreads,
-                atropos_use_random_seed=not skull_strip_fixed_seed,
-                bids_suffix='T1w',
-                name='t1w_n4_only_wf',
-            )
+            t1w_n4_wf = init_anat_preproc_wf(name='t1w_n4_wf')
             workflow.connect([
-                (apply_t1w_mask, t1w_n4_only_wf, [('out_file', 'inputnode.in_files')]),
-                (t1w_n4_only_wf, t1w_buffer, [
-                    (('outputnode.bias_corrected', pop_file), 't1w_preproc'),
-                    (('outputnode.out_file', pop_file), 't1w_brain'),
-                ]),
+                (t1w_validate, t1w_n4_wf, [('out_file', 'inputnode.in_anat')]),
+                (t1w_n4_wf, t1w_buffer, [('outputnode.anat_preproc', 't1w_preproc')]),
+                (t1w_n4_wf, apply_t1w_mask, [('outputnode.anat_preproc', 'in_file')]),
             ])  # fmt:skip
         else:
             LOGGER.info('ANAT Skipping T1w masking')
-            workflow.connect(apply_t1w_mask, 'out_file', t1w_buffer, 't1w_brain')
+            workflow.connect(t1w_validate, 'out_file', apply_t1w_mask, 'in_file')
 
     # T2w masking logic:
     #
@@ -709,33 +702,24 @@ def init_infant_anat_fit_wf(
         LOGGER.info('ANAT Found T2w brain mask')
         if reference_anat == 'T2w':
             desc += (
-                'A pre-computed T1w brain mask was provided as input and used throughout the '
+                'A pre-computed T2w brain mask was provided as input and used throughout the '
                 'workflow.'
             )
         t2w_buffer.inputs.t2w_mask = t2w_mask
         apply_t2w_mask.inputs.in_mask = t2w_mask
-        workflow.connect([
-            (t2w_validate, apply_t2w_mask, [('out_file', 'in_file')]),
-        ])  # fmt:skip
+        workflow.connect(apply_t2w_mask, 'out_file', t2w_buffer, 't2w_brain')
 
         if not t2w_preproc:
             LOGGER.info('ANAT Skipping skull-strip, INU-correction only')
-            t2w_n4_only_wf = init_n4_only_wf(
-                omp_nthreads=omp_nthreads,
-                atropos_use_random_seed=not skull_strip_fixed_seed,
-                bids_suffix='T2w',
-                name='t2w_n4_only_wf',
-            )
+            t2w_n4_wf = init_anat_preproc_wf(name='t2w_n4_wf')
             workflow.connect([
-                (apply_t2w_mask, t2w_n4_only_wf, [('out_file', 'inputnode.in_files')]),
-                (t2w_n4_only_wf, t2w_buffer, [
-                    (('outputnode.bias_corrected', pop_file), 't2w_preproc'),
-                    (('outputnode.out_file', pop_file), 't2w_brain'),
-                ]),
+                (t2w_validate, t2w_n4_wf, [('out_file', 'inputnode.in_anat')]),
+                (t2w_n4_wf, t2w_buffer, [('outputnode.anat_preproc', 't2w_preproc')]),
+                (t2w_n4_wf, apply_t2w_mask, [('outputnode.anat_preproc', 'in_file')]),
             ])  # fmt:skip
         else:
             LOGGER.info('ANAT Skipping T2w masking')
-            workflow.connect(apply_t2w_mask, 'out_file', t2w_buffer, 't2w_brain')
+            workflow.connect(t2w_validate, 'out_file', apply_t2w_mask, 'in_file')
 
     # Stage 3: Coregistration
     t1w2t2w_xfm = precomputed.get('t1w2t2w_xfm')
@@ -1643,28 +1627,19 @@ def init_infant_single_anat_fit_wf(
         desc += 'A pre-computed brain mask was provided as input and used throughout the workflow.'
         anat_buffer.inputs.anat_mask = anat_mask
         apply_mask.inputs.in_mask = anat_mask
-        workflow.connect([
-            (anat_validate, apply_mask, [('out_file', 'in_file')]),
-        ])  # fmt:skip
+        workflow.connect(apply_mask, 'out_file', anat_buffer, 'anat_brain')
 
         if not anat_preproc:
             LOGGER.info('ANAT Skipping skull-strip, INU-correction only')
-            n4_only_wf = init_n4_only_wf(
-                omp_nthreads=omp_nthreads,
-                atropos_use_random_seed=not skull_strip_fixed_seed,
-                bids_suffix=reference_anat,
-                name='n4_only_wf',
-            )
+            anat_n4_wf = init_anat_preproc_wf(name='anat_n4_wf')
             workflow.connect([
-                (apply_mask, n4_only_wf, [('out_file', 'inputnode.in_files')]),
-                (n4_only_wf, anat_buffer, [
-                    (('outputnode.bias_corrected', pop_file), 'anat_preproc'),
-                    (('outputnode.out_file', pop_file), 'anat_brain'),
-                ]),
+                (anat_validate, anat_n4_wf, [('out_file', 'inputnode.in_anat')]),
+                (anat_n4_wf, anat_buffer, [('outputnode.anat_preproc', 'anat_preproc')]),
+                (anat_n4_wf, apply_mask, [('outputnode.anat_preproc', 'in_file')]),
             ])  # fmt:skip
         else:
-            LOGGER.info('ANAT Skipping T2w masking')
-            workflow.connect(apply_mask, 'out_file', anat_buffer, 'anat_brain')
+            LOGGER.info(f'ANAT Skipping {reference_anat} masking')
+            workflow.connect(anat_validate, 'out_file', apply_mask, 'in_file')
 
     # Stage 3: Segmentation
     seg_method = 'jlf' if config.execution.segmentation_atlases_dir else 'fast'

nibabies/workflows/anatomical/preproc.py

@@ -12,11 +12,9 @@ def init_anat_preproc_wf(
 
     This workflow accepts T1w/T2w images as inputs (either raw or a merged template) and performs:
     - Intensity clipping
-    - Denoising
     - N4 Bias Field Correction
 
-    The outputs of this workflow will be used to either create the brainmask,
-    or reconstruct the cortical surfaces.
+    The outputs of this workflow will be a structural reference used for subsequent processing.
 
     Inputs
     ------
@@ -45,7 +43,6 @@ def init_anat_preproc_wf(
     # validate image
     validate = pe.Node(ValidateImage(), name='anat_validate', run_without_submitting=True)
     clip = pe.Node(IntensityClip(p_min=10.0, p_max=99.5), name='clip')
-    # denoise = pe.Node(DenoiseImage(dimension=3, noise_model="Rician"), name="denoise")
     n4_correct = pe.Node(
         N4BiasFieldCorrection(
             dimension=3,
@@ -61,14 +58,11 @@ def init_anat_preproc_wf(
     )
     final_clip = pe.Node(IntensityClip(p_min=5.0, p_max=99.5), name='final_clip')
 
-    # fmt:off
     wf.connect([
         (inputnode, validate, [('in_anat', 'in_file')]),
         (validate, clip, [('out_file', 'in_file')]),
         (clip, n4_correct, [('out_file', 'input_image')]),
-        # (denoise, n4_correct, [("output_image", "input_image")]),
         (n4_correct, final_clip, [('output_image', 'in_file')]),
         (final_clip, outputnode, [('out_file', 'anat_preproc')]),
-    ])
-    # fmt:on
+    ])  # fmt:skip
     return wf