ENH: Longitudinal registration - ICON fine tuning

.github/workflows/ci.yml

@@ -407,8 +407,8 @@ jobs:
 # The following tests are excluded from CI and should be run locally:
 #
 # Slow/GPU-intensive tests:
-# - tests/test_register_images_ants.py (slow, computationally intensive)
-# - tests/test_register_images_icon.py (requires CUDA for ICON)
+# - tests/test_register_images_ANTS.py (slow, computationally intensive)
+# - tests/test_register_images_ICON.py (requires CUDA for ICON)
 # - tests/test_transform_tools.py (depends on slow registration tests)
 # - tests/test_segment_chest_total_segmentator.py (requires CUDA for TotalSegmentator)
 #
@@ -421,7 +421,7 @@ jobs:
 #   pytest tests/ -v --run-slow                 # Run all slow tests
 #   pytest tests/ -v --run-gpu --run-slow       # GPU + slow (typical local dev profile)
 #   pytest tests/ -v --run-simpleware --run-gpu --run-slow  # Full Simpleware coverage
-#   pytest tests/test_register_images_ants.py -v --run-slow
+#   pytest tests/test_register_images_ANTS.py -v --run-slow
 #
 # Self-hosted GPU runner enables ALL buckets via --run-all
 # (--run-gpu --run-slow --run-simpleware --run-physicsnemo --run-experiments --run-tutorials).

.gitignore

@@ -32,6 +32,7 @@ network_weights
 *.gz
 *.mat
 *.mhd
+*.raw
 *.zip
 *.nii
 *.tfm

README.md

@@ -134,7 +134,7 @@ print(WorkflowConvertImageToUSD.__name__)
 - **Registration Classes**: Multiple registration methods for different use cases
   - Image-to-Image Registration:
     - `RegisterImagesICON`: Deep learning-based registration using Icon algorithm
-    - `RegisterImagesANTs`: Classical deformable registration using ANTs
+    - `RegisterImagesANTS`: Classical deformable registration using ANTs
     - `RegisterTimeSeriesImages`: Specialized time series registration for 4D CT
   - Model-to-Image/Model Registration:
     - `RegisterModelsPCA`: PCA-based statistical shape model registration
@@ -354,7 +354,7 @@ if "lung" in masks:
 ### Image Registration
 
 ```python
-from physiomotion4d import RegisterImagesICON, RegisterImagesANTs, RegisterTimeSeriesImages
+from physiomotion4d import RegisterImagesICON, RegisterImagesANTS, RegisterTimeSeriesImages
 import itk
 
 # Option 1: Icon deep learning registration (GPU-accelerated)
@@ -364,14 +364,14 @@ registerer.set_fixed_image(itk.imread("reference_frame.mha"))
 results = registerer.register(itk.imread("target_frame.mha"))
 
 # Option 2: ANTs classical registration
-registerer = RegisterImagesANTs()
+registerer = RegisterImagesANTS()
 registerer.set_fixed_image(itk.imread("reference_frame.mha"))
 results = registerer.register(itk.imread("target_frame.mha"))
 
 # Option 3: Time series registration for 4D CT
 time_series_reg = RegisterTimeSeriesImages(
     reference_index=0,
-    registration_method='icon'  # or 'ants'
+    registration_method='ICON'  # or 'ANTS'
 )
 transforms = time_series_reg.register_time_series(
     image_filenames=["time00.mha", "time01.mha", "time02.mha"]
@@ -735,7 +735,7 @@ will change, and flag any coordinate-system or shape implications.
 Use `/impl` for end-to-end implementation: read → summarize → plan → diff → lint.
 
 ```text
-/impl add set_regularization_weight() to RegisterImagesANTs
+/impl add set_regularization_weight() to RegisterImagesANTS
 ```
 
 ```text
@@ -757,7 +757,7 @@ Use `/test-feature` to get a test plan and a complete pytest file using syntheti
 ```
 
 ```text
-/test-feature RegisterImagesANTs with a pair of small synthetic ITK images
+/test-feature RegisterImagesANTS with a pair of small synthetic ITK images
 ```
 
 The agent will state image shapes and axis orders in every test docstring, wire
@@ -772,7 +772,7 @@ Use `/doc-feature` after modifying a public API to refresh docstrings and regene
 the API map.
 
 ```text
-/doc-feature update docstrings for RegisterImagesANTs after adding set_regularization_weight
+/doc-feature update docstrings for RegisterImagesANTS after adding set_regularization_weight
 ```
 
 The agent will update affected docstrings in NumPy style, add shape/axis annotations

docs/api/index.rst

@@ -40,7 +40,7 @@ By Category
 
 **Image Registration**
    * :class:`~physiomotion4d.RegisterImagesBase` - Base registration class
-   * :class:`~physiomotion4d.RegisterImagesANTs` - ANTs registration
+   * :class:`~physiomotion4d.RegisterImagesANTS` - ANTs registration
    * :class:`~physiomotion4d.RegisterImagesICON` - Icon deep learning registration
    * :class:`~physiomotion4d.RegisterTimeSeriesImages` - 4D time series registration
 

docs/api/registration/ants.rst

@@ -5,13 +5,13 @@ ANTs Registration
 .. module:: physiomotion4d.register_images_ants
 .. currentmodule:: physiomotion4d
 
-``RegisterImagesANTs`` provides optimization-based deformable image
+``RegisterImagesANTS`` provides optimization-based deformable image
 registration through ANTs.
 
 Class Reference
 ===============
 
-.. autoclass:: RegisterImagesANTs
+.. autoclass:: RegisterImagesANTS
    :members:
    :undoc-members:
    :show-inheritance:
@@ -23,12 +23,12 @@ Basic Registration
 
    import itk
 
-   from physiomotion4d import RegisterImagesANTs
+   from physiomotion4d import RegisterImagesANTS
 
    fixed = itk.imread("reference.mha")
    moving = itk.imread("moving.mha")
 
-   registrar = RegisterImagesANTs()
+   registrar = RegisterImagesANTS()
    registrar.set_modality("ct")
    registrar.set_transform_type("SyN")
    registrar.set_number_of_iterations([30, 15, 7])

docs/api/registration/index.rst

@@ -19,7 +19,7 @@ PhysioMotion4D image registration classes align moving 3D images to a fixed
 Common Result Shape
 ===================
 
-``RegisterImagesANTs.register()`` and ``RegisterImagesICON.register()`` return:
+``RegisterImagesANTS.register()`` and ``RegisterImagesICON.register()`` return:
 
 * ``forward_transform``
 * ``inverse_transform``
@@ -35,12 +35,12 @@ Basic Example
 
    import itk
 
-   from physiomotion4d import RegisterImagesANTs
+   from physiomotion4d import RegisterImagesANTS
 
    fixed = itk.imread("reference.mha")
    moving = itk.imread("moving.mha")
 
-   registrar = RegisterImagesANTs()
+   registrar = RegisterImagesANTS()
    registrar.set_modality("ct")
    registrar.set_fixed_image(fixed)
 

docs/api/registration/time_series.rst

@@ -6,7 +6,7 @@ Time-Series Registration
 .. currentmodule:: physiomotion4d
 
 ``RegisterTimeSeriesImages`` registers ordered 3D image phases to a reference
-frame using ANTs, Greedy, ICON, or combined ``ants_icon`` / ``greedy_icon``
+frame using ANTs, Greedy, ICON, or combined ``ANTS_ICON`` / ``greedy_ICON``
 methods.
 
 Class Reference
@@ -29,7 +29,7 @@ Basic Usage
 
    images = [itk.imread(f"phase_{idx:02d}.mha") for idx in range(10)]
 
-   registrar = RegisterTimeSeriesImages(registration_method="ants")
+   registrar = RegisterTimeSeriesImages(registration_method="ANTS")
    registrar.set_fixed_image(images[0])
 
    result = registrar.register_time_series(

docs/api/workflows.rst

@@ -157,7 +157,7 @@ High-Resolution 4D CT Reconstruction
        time_series_images=time_series_images,
        fixed_image=time_series_images[0],
        reference_frame=0,
-       registration_method="ants",
+       registration_method="ANTS",
    )
 
    result = workflow.run_workflow(upsample_to_fixed_resolution=True)

docs/architecture.rst

@@ -25,7 +25,7 @@ Data Flow
           v
    RegisterTimeSeriesImages
       |        |
-      |        +--> RegisterImagesANTs / RegisterImagesICON
+      |        +--> RegisterImagesANTS / RegisterImagesICON
       v
    SegmentChestTotalSegmentator / SegmentHeartSimpleware
           |

docs/cli_scripts/4dct_reconstruction.rst

@@ -47,8 +47,8 @@ Registration Options
    physiomotion4d-reconstruct-highres-4d-ct \
        --time-series-images frame_*.mha \
        --fixed-image highres_reference.mha \
-       --registration-method ants \
-       --ants-iterations 30 15 7 3 \
+       --registration-method ANTS \
+       --ANTS-iterations 30 15 7 3 \
        --prior-weight 0.5 \
        --output-dir ./results
 

docs/cli_scripts/fit_statistical_model_to_patient.rst

@@ -119,7 +119,7 @@ Registration Configuration
    Enable mask-to-image refinement registration. Requires
    ``--template-labelmap`` and template label IDs. Disabled by default.
 
-``--use-icon-refinement``
+``--use-ICON-refinement``
    Enable ICON deep learning registration refinement (default: disabled)
 
 Output Options

docs/developer/registration_images.rst

@@ -12,12 +12,12 @@ Basic Pattern
 
    import itk
 
-   from physiomotion4d import RegisterImagesANTs
+   from physiomotion4d import RegisterImagesANTS
 
    fixed = itk.imread("fixed.mha")
    moving = itk.imread("moving.mha")
 
-   registrar = RegisterImagesANTs()
+   registrar = RegisterImagesANTS()
    registrar.set_modality("ct")
    registrar.set_fixed_image(fixed)
 
@@ -38,7 +38,7 @@ Time Series
 
    images = [itk.imread(f"phase_{idx:02d}.mha") for idx in range(10)]
 
-   registrar = RegisterTimeSeriesImages(registration_method="ants")
+   registrar = RegisterTimeSeriesImages(registration_method="ANTS")
    registrar.set_fixed_image(images[0])
    result = registrar.register_time_series(
        moving_images=images,

docs/examples.rst

@@ -66,7 +66,7 @@ phase images:
        time_series_images=time_series_images,
        fixed_image=fixed_image,
        reference_frame=0,
-       registration_method="ants",
+       registration_method="ANTS",
    )
 
    result = workflow.run_workflow(upsample_to_fixed_resolution=True)
@@ -176,10 +176,10 @@ Advanced registration with multiple stages:
 
 .. code-block:: python
 
-   from physiomotion4d import RegisterImagesANTs
+   from physiomotion4d import RegisterImagesANTS
    import itk
 
-   registerer = RegisterImagesANTs()
+   registerer = RegisterImagesANTS()
 
    fixed = itk.imread("reference.mha")
    moving = itk.imread("target.mha")

docs/tutorials.rst

@@ -244,7 +244,7 @@ Script
    ``tutorials/tutorial_08_dirlab_pca_time_series.py``
 
 Workflow
-   ``RegisterTimeSeriesImages`` with ``registration_method='ants_icon'`` and
+   ``RegisterTimeSeriesImages`` with ``registration_method='ANTS_ICON'`` and
    ``TransformTools``
 
 Dataset

experiments/Heart-Create_Statistical_Model/3-registration_based_correspondence.py

@@ -114,7 +114,7 @@
     # This performs progressive multi-stage registration: rigid → affine → SyN deformable
     result = registrar.register(
         transform_type="Deformable",  # Uses ANTs SyN (Symmetric Normalization)
-        use_icon=False,  # Set to True for additional ICON deep learning refinement
+        use_ICON=False,  # Set to True for additional ICON deep learning refinement
     )
 
     forward_transform = result["forward_transform"]
@@ -341,6 +341,6 @@
 #   1. Rigid alignment
 #   2. Affine transformation
 #   3. SyN deformable registration (diffeomorphic)
-# - Setting `use_icon=True` in the `register()` call would add ICON deep learning refinement after SyN
+# - Setting `use_ICON=True` in the `register()` call would add ICON deep learning refinement after SyN
 # - The `roi_dilation_mm` parameter controls the dilation of the ROI mask (default 20mm)
 # - SyN registration provides smooth, invertible deformation fields for anatomical correspondence

experiments/Heart-GatedCT_To_USD/1-register_images.py

@@ -4,7 +4,7 @@
 
 import itk
 
-from physiomotion4d.register_images_ants import RegisterImagesANTs
+from physiomotion4d.register_images_ants import RegisterImagesANTS
 from physiomotion4d.segment_chest_total_segmentator import SegmentChestTotalSegmentator
 from physiomotion4d.test_tools import TestTools
 from physiomotion4d.transform_tools import TransformTools
@@ -76,7 +76,7 @@
     )
 
     # %%
-    reg = RegisterImagesANTs()
+    reg = RegisterImagesANTS()
     reg.set_mask_dilation(5)
     reg.set_number_of_iterations([10, 5, 2])
 

experiments/Heart-GatedCT_To_USD/test_compare_registration_speed.py

@@ -16,7 +16,7 @@
 from itk import TubeTK as ttk
 
 from physiomotion4d.test_tools import TestTools
-from physiomotion4d.register_images_ants import RegisterImagesANTs
+from physiomotion4d.register_images_ants import RegisterImagesANTS
 from physiomotion4d.register_images_greedy import RegisterImagesGreedy
 from physiomotion4d.register_images_icon import RegisterImagesICON
 
@@ -106,7 +106,7 @@
 
 # --- ANTs (deformable SyN) ---
 try:
-    reg_a = RegisterImagesANTs()
+    reg_a = RegisterImagesANTS()
     reg_a.set_modality("ct")
     reg_a.set_transform_type("Deformable")
     reg_a.set_number_of_iterations([10, 5, 2])  # reduced for speed