geometric-medical-phantoms 1.0.2

Python bindings for the GeometricMedicalPhantoms shared library — synthetic MRI/CT phantoms (Shepp-Logan, Tubes, Torso) with physiological motion signals

geometric-medical-phantoms

Python bindings for the GeometricMedicalPhantoms shared library — a Julia-based toolkit for generating synthetic MRI/CT phantoms.

Supported phantoms

• Shepp-Logan — classic ellipsoid head phantom (CT and MRI variants)
• Tubes — cylindrical quality-control phantom
• Torso — anatomically-motivated torso with respiratory and cardiac motion

Requirements

• Python ≥ 3.9
• NumPy ≥ 1.20
• The pre-built libgeomphantoms shared library (see Releases)

Installation

pip install geometric-medical-phantoms

Then download the pre-built library bundle for your platform from the GitHub Releases page and extract it.

Quick start

from geometric_medical_phantoms import GMPLib

lib = GMPLib("/path/to/build/lib/libgeomphantoms.so")  # .dylib / .dll on other platforms

# Shepp-Logan 3-D phantom
ti = lib.shepp_logan_ct_default()
phantom = lib.create_shepp_logan_phantom_3d(128, 128, 128, ti)
# phantom is a (128, 128, 128) float32 numpy array in Fortran (column-major) order

# Respiratory signal
phys = lib.respiratory_physiology_default()
t, sig = lib.generate_respiratory_signal(60.0, 50.0, 15.0, phys)

# Dynamic torso phantom (4-D cine)
torso = lib.create_torso_phantom_3d(64, 64, 64, resp_signal=sig[:50])
# torso.shape == (64, 64, 64, 50)

Documentation

Full API reference and usage guides are available at https://hakkelt.github.io/GeometricMedicalPhantoms.jl.

License

MIT — see LICENSE.