ENH: Remove unreliable netgen meshing; fix py launcher and CT auto-convert

.github/scripts/build_dashboard.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python3
 """
 Build the nightly health dashboard for PhysioTwin4D.
 

README.md

@@ -25,7 +25,7 @@ for users and contributors. Key sections:
 
 - **Personalized digital twins**: build subject-specific anatomic models and physiological AI surrogates from 3D/4D medical images
 - **Statistical shape models**: capture subject-specific anatomy and establish cross-subject correspondence, aiding AI surrogate generalization and simplifying traditional solver setup
-- **Simplified workflows on industry-leading open-source tools**: ICON and Greedy for registration; MONAI with TotalSegmentator and Simpleware for segmentation; Netgen for meshing (LGPL license); scikit-learn for statistical shape modeling; ITK for image processing; PyVista and OpenUSD/Omniverse for geometry manipulation; CuPy for accelerated computing; and PhysicsNeMo for AI surrogates
+- **Simplified workflows on industry-leading open-source tools**: ICON and Greedy for registration; MONAI with TotalSegmentator and Simpleware for segmentation; scikit-learn for statistical shape modeling; ITK for image processing; PyVista and OpenUSD/Omniverse for geometry manipulation; CuPy for accelerated computing; and PhysicsNeMo for AI surrogates
 - **Extensible class hierarchy**: add new segmentation and registration methods, and extend to new data types, organs, and physiological processes, without reworking the workflow layer
 - **Physiological motion**: cardiac and respiratory motion today, expanding to electrophysiology, blood flow, and organ perfusion
 - **NVIDIA Omniverse as the simulation hub**: the end goal for simulation — a simulation-information hub and gateway to other engines (e.g., Ansys solvers), interactive simulations for treatment planning (e.g., Isaac Sim, Newton), visualization systems (e.g., AR/VR devices), and physical systems (e.g., robots via ROS)
@@ -71,9 +71,8 @@ See the [contributing guide](https://project-monai.github.io/physiotwin4d/contri
 
 This project is licensed under the Apache 2.0 License - see the LICENSE file for details.
 
-Note that some optional dependencies carry their own license terms: meshing
-via Netgen is LGPL-licensed, and NVIDIA Omniverse is distributed under its
-own customer license, which is free for academic and commercial use. To our
-knowledge, none of the dependencies carry commercially prohibitive licenses
-such as GPL, but no guarantees are provided — review the license of each
-dependency for your own use case.
+Note that some optional dependencies carry their own license terms: NVIDIA
+Omniverse is distributed under its own customer license, which is free for
+academic and commercial use. To our knowledge, none of the dependencies
+carry commercially prohibitive licenses such as GPL, but no guarantees are
+provided — review the license of each dependency for your own use case.

data/DirLab-4DCT/Convert4DCTToMHD.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

data/Slicer-Heart-CT/README.md

@@ -18,14 +18,17 @@ assert DataDownloadTools.VerifySlicerHeartCTData("data/Slicer-Heart-CT")
 ```
 
 This fetches a single ~1.2 GB file from
-[github.com/SlicerHeart/SlicerHeart](https://github.com/SlicerHeart/SlicerHeart/releases/download/TestingData/TruncalValve_4DCT.seq.nrrd).
-An existing non-empty file is reused, so re-running the command resumes an
-interrupted download.
+[github.com/SlicerHeart/SlicerHeart](https://github.com/SlicerHeart/SlicerHeart/releases/download/TestingData/TruncalValve_4DCT.seq.nrrd),
+then splits it into per-phase 3D `slice_???.mha` volumes in the same
+directory via `ConvertImage4DTo3D`. An existing non-empty `.seq.nrrd` is
+reused, and the split is skipped once the `slice_???.mha` files are present —
+so re-running the command resumes an interrupted download or conversion.
 
 **Directory structure after download:**
-```
+```text
 data/Slicer-Heart-CT/
 ├── TruncalValve_4DCT.seq.nrrd
+├── slice_000.mha ... slice_020.mha
 └── README.md (this file)
 ```
 
@@ -54,11 +57,14 @@ Philadelphia):
 - Primary dataset for tutorials
 - Primary dataset for `experiments/Heart-GatedCT_To_USD/` and
   `experiments/Heart-VTKSeries_To_USD/`, whose
-  `0-download_and_convert_4d_to_3d.py` scripts download this sequence and
-  split it into per-phase 3D `.mha` slices via `ConvertImage4DTo3D`
+  `0-download_and_convert_4d_to_3d.py` scripts call
+  `DataDownloadTools.DownloadSlicerHeartCTData`, which downloads this
+  sequence and splits it into per-phase 3D `.mha` slices
 - Used in the test suite (`tests/test_download_heart_data.py`)
 - Example data for cardiac motion visualization in NVIDIA Omniverse
 
 ### Files in This Directory
 
 - `TruncalValve_4DCT.seq.nrrd` — the downloaded 4D CT sequence
+- `slice_000.mha` ... `slice_020.mha` — the 21 cardiac phases split out of
+  the sequence above

docs/api/workflows.rst

@@ -27,7 +27,7 @@ Available Workflows
    * - :class:`WorkflowConvertImageToUSD`
      - Convert a 4D cardiac CT sequence into animated USD anatomy.
    * - :class:`WorkflowConvertImageToVTK`
-     - Segment one CT image and export anatomy-group VTK surfaces and meshes.
+     - Segment one CT image and export anatomy-group VTK surfaces.
    * - :class:`WorkflowConvertVTKToUSD`
      - Convert VTK/VTP/VTU meshes or time series into USD.
    * - :class:`WorkflowCreateStatisticalModel`

docs/architecture.rst

@@ -61,8 +61,7 @@ Primary Workflows
    contours and painted animated USD files.
 
 ``WorkflowConvertImageToVTK``
-   Segments a 3D CT image and exports anatomy groups as VTK surfaces and voxel
-   meshes.
+   Segments a 3D CT image and exports anatomy groups as VTK surfaces.
 
 ``WorkflowCreateStatisticalModel``
    Aligns a population of meshes to a reference and builds a PCA statistical

docs/cli_scripts/download_data.rst

@@ -54,15 +54,18 @@ Options
 Output
 ======
 
-For ``Slicer-Heart-CT``, the command downloads or reuses:
+For ``Slicer-Heart-CT``, the command downloads the 4-D sequence and splits it
+into per-phase 3-D volumes:
 
 .. code-block:: text
 
    data/Slicer-Heart-CT/TruncalValve_4DCT.seq.nrrd
+   data/Slicer-Heart-CT/slice_000.mha ... slice_020.mha
 
 The command uses
 :meth:`physiotwin4d.data_download_tools.DataDownloadTools.DownloadSlicerHeartCTData`,
-so repeated runs reuse the existing non-empty file.
+so repeated runs reuse the existing non-empty file and skip the split once
+the ``slice_???.mha`` files are present.
 
 For ``KCL-Heart-Model``, the command downloads, extracts, and reuses:
 

docs/cli_scripts/overview.rst

@@ -54,7 +54,7 @@ Current Scripts
    * - :doc:`heart_gated_ct`
      - Process cardiac gated CT to animated heart models with physiological motion
    * - ``physiotwin4d-convert-image-to-vtk``
-     - Segment one 3D image and export anatomy-group VTK surfaces and meshes
+     - Segment one 3D image and export anatomy-group VTK surfaces
    * - ``physiotwin4d-convert-image-4d-to-3d``
      - Split a 4D medical image into a 3D time series using ITK readers
    * - :doc:`create_statistical_model`

docs/developer/architecture.rst

@@ -40,7 +40,7 @@ Workflow Classes
    transformation, and animated USD generation.
 
 ``WorkflowConvertImageToVTK``
-   Converts one 3D image into labeled VTK surface and voxel-mesh outputs.
+   Converts one 3D image into labeled VTK surface outputs.
 
 ``WorkflowCreateStatisticalModel``
    Builds a PCA statistical shape model from aligned population meshes.

docs/tutorials.rst

@@ -201,7 +201,7 @@ Preview
 
 Inner API usage
    The workflow owns a segmentation method and turns each anatomy group into
-   decimated VTK surfaces and volume meshes:
+   a decimated VTK surface:
 
    .. code-block:: python
 
@@ -213,11 +213,17 @@ Inner API usage
       result = workflow.process(
           input_image=ct_image,
           surface_target_reduction=0.5,
-          mesh_target_reduction=0.7,
       )
 
    Use ``SegmentChestTotalSegmentator`` instead for non-contrast studies.
 
+   By default the script saves one combined ``patient_surfaces.vtp``.  Set
+   ``SAVE_GROUP_SURFACES = True`` and/or ``SAVE_LABEL_SURFACES = True`` near
+   the top of the script to additionally save one VTP per anatomy group
+   (e.g. ``patient_heart.vtp``) and/or one VTP per individual anatomical
+   structure (e.g. ``patient_left_ventricle.vtp``); the latter passes
+   ``extract_label_surfaces=True`` to :meth:`WorkflowConvertImageToVTK.process`.
+
 Run
    .. code-block:: bash
 

experiments/Colormap-VTK_To_USD/colormap_vtk_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Colormap Features for VTK to USD Conversion
 #

experiments/Convert_VTK_To_USD/convert_chop_alterra_valve_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Cardiac Valve 4D Time-Series Conversion to USD
 #

experiments/Convert_VTK_To_USD/convert_chop_heart_vtk_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 from pathlib import Path
 

experiments/Convert_VTK_To_USD/convert_chop_tpv25_valve_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Cardiac Valve 4D Time-Series Conversion to USD
 #

experiments/Convert_VTK_To_USD/convert_vtk_to_usd_using_class.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # VTK to USD Converter Test Notebook
 #

experiments/DisplacementField_To_USD/displacement_field_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Displacement Field to USD for Omniverse Visualization
 #

experiments/Heart-Create_Statistical_Model/1-input_meshes_to_input_surfaces.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Convert VTK to VTP Surface Files
 #

experiments/Heart-Create_Statistical_Model/2-input_surfaces_to_surfaces_aligned.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # ICP Affine Registration: Align Heart Models to Average
 #

experiments/Heart-Create_Statistical_Model/3-registration_based_correspondence.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Registration-Based Correspondence
 #

experiments/Heart-Create_Statistical_Model/4-surfaces_aligned_correspond_to_pca_inputs.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 from pathlib import Path
 

experiments/Heart-Create_Statistical_Model/5-compute_pca_model.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Compute PCA Statistical Model
 #

experiments/Heart-GatedCT_To_USD/0-download_and_convert_4d_to_3d.py

@@ -1,9 +1,7 @@
-#!/usr/bin/env python
 # %%
 import shutil
 from pathlib import Path
 
-from physiotwin4d.convert_image_4d_to_3d import ConvertImage4DTo3D
 from physiotwin4d.data_download_tools import DataDownloadTools
 
 _HERE = Path(__file__).resolve().parent
@@ -12,15 +10,11 @@
 data_dir = _HERE.parent.parent / "data" / "Slicer-Heart-CT"
 output_dir = _HERE / "results"
 
-data_dir.mkdir(parents=True, exist_ok=True)
 output_dir.mkdir(parents=True, exist_ok=True)
 
-input_image_filename = DataDownloadTools.DownloadSlicerHeartCTData(data_dir)
+# Downloads TruncalValve_4DCT.seq.nrrd and splits it into slice_???.mha.
+DataDownloadTools.DownloadSlicerHeartCTData(data_dir)
 
 # %%
-conv = ConvertImage4DTo3D()
-conv.load_image_4d(str(input_image_filename))
-conv.save_3d_images(data_dir, "slice")
-
 # Save the mid-stroke slice as the fixed/reference image
 shutil.copyfile(data_dir / "slice_007.mha", output_dir / "slice_fixed.mha")

experiments/Heart-GatedCT_To_USD/1-register_images.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 from pathlib import Path
 

experiments/Heart-GatedCT_To_USD/2-generate_segmentation.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Heart-GatedCT_To_USD/3-transform_dynamic_and_static_contours.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Heart-GatedCT_To_USD/4-merge_dynamic_and_static_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Heart-Simpleware_Segmentation/simpleware_heart_segmentation.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # Heart Segmentation using Simpleware Medical ASCardio
 #

experiments/Heart-Statistical_Model_To_Patient/heart_model_to_model_icp_itk.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # ICP via ITK Heart Model to Image Registration Experiment
 #

experiments/Heart-Statistical_Model_To_Patient/heart_model_to_model_registration_pca.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # PCA-based Heart Model to Image Registration Experiment
 #

experiments/Heart-Statistical_Model_To_Patient/heart_model_to_patient-CHOPValve.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # ## Setup and Imports
 

experiments/Heart-Statistical_Model_To_Patient/heart_model_to_patient.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # ## Setup and Imports
 

experiments/Heart-VTKSeries_To_USD/0-download_and_convert_4d_to_3d.py

@@ -1,8 +1,6 @@
-#!/usr/bin/env python
 # %%
 import os
 
-from physiotwin4d.convert_image_4d_to_3d import ConvertImage4DTo3D
 from physiotwin4d.data_download_tools import DataDownloadTools
 
 _HERE = os.path.dirname(os.path.abspath(__file__))
@@ -11,16 +9,8 @@
 data_dir = os.path.join(_HERE, "..", "..", "data", "Slicer-Heart-CT")
 output_dir = os.path.join(_HERE, "results")
 
-if not os.path.exists(data_dir):
-    os.makedirs(data_dir)
-
 if not os.path.exists(output_dir):
     os.makedirs(output_dir)
 
-input_image_filename = DataDownloadTools.DownloadSlicerHeartCTData(data_dir)
-
-# %%
-if not os.path.exists(f"{data_dir}/slice_000.mha"):
-    conv = ConvertImage4DTo3D()
-    conv.load_image_4d(str(input_image_filename))
-    conv.save_3d_images(data_dir, "slice")
+# Downloads TruncalValve_4DCT.seq.nrrd and splits it into slice_???.mha.
+DataDownloadTools.DownloadSlicerHeartCTData(data_dir)

experiments/Heart-VTKSeries_To_USD/1-heart_vtkseries_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import glob
 import os

experiments/Lung-GatedCT_To_USD/0-register_dirlab_4dct.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 from typing import Optional

experiments/Lung-GatedCT_To_USD/1-make_dirlab_models.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import itk
 import numpy as np

experiments/Lung-GatedCT_To_USD/2-paint_dirlab_models.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 from data_dirlab_4d_ct import DataDirLab4DCT
 from pxr import Usd

experiments/Lung-GatedCT_To_USD/Experiment_ArrangeOnStage.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Lung-GatedCT_To_USD/Experiment_CombineModels.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Lung-GatedCT_To_USD/Experiment_SegReg.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Lung-GatedCT_To_USD/Experiment_SubSurfaceScatter.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Lung-VesselsAirways/0-GenData.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 from datetime import datetime
 

experiments/Reconstruct4DCT/reconstruct_4d_ct.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %%
 import os
 

experiments/Reconstruct4DCT/reconstruct_4d_ct_class.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 # %% [markdown]
 # # 4D CT Reconstruction Using RegisterTimeSeriesImages Class
 #

pyproject.toml

@@ -78,7 +78,6 @@ dependencies = [
     "pyvista[all]>=0.47.0",
     "usd-core>=23.11",
     "trimesh>=4.0.0",
-    "netgen-mesher>=6.2.2606",
 
     # Utilities
     "ipykernel>=6.0.0",
@@ -229,8 +228,6 @@ module = [
     "itk.*",
     "matplotlib",
     "matplotlib.*",
-    "netgen",
-    "netgen.*",
     "nibabel",
     "nibabel.*",
     "nrrd",

src/physiotwin4d/cli/convert_image_4d_to_3d.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 """Command-line interface for splitting a 3D/4D image into a 3D time series.
 
 Reads a 3D or 4D medical image and writes one ``.mha`` file per temporal frame

src/physiotwin4d/cli/convert_image_to_usd.py

@@ -1,4 +1,3 @@
-#!/usr/bin/env python
 """
 Command-line interface for the Image-to-USD workflow.