patminton — GPU-Accelerated 3D Photoacoustic Tomography

CUDA implementations of the forward and adjoint operators of 3D photoacoustic tomography (PAT) as on-the-fly matrix-vector products, for iterative image reconstruction without storing the system matrix, plus CGLS, L-BFGS-B, PGD and Chambolle-Pock TV solvers built on top of them.

For a 201³ grid observed by 11 520 transducer positions with 1000 time samples, the dense system matrix would occupy 201³ × 11 520 × 1000 × 8 B ≈ 750 TB in double precision; patminton applies it and its adjoint on the fly on the GPU instead.

Installation

pip install patminton

Requirements: an NVIDIA GPU (compute capability ≥ 7.5, driver ≥ 580) and Python ≥ 3.9. On Linux x86-64 this installs a prebuilt CUDA 13 wheel and needs no CUDA toolkit; if cuFFT is not already provided by a toolkit or by PyTorch (e.g. with a CUDA 12 build of PyTorch), use pip install patminton[cuda13].

Installing from source instead requires the CUDA toolkit (nvcc ≥ 11 on PATH), which compiles the library for every supported architecture. To target a specific set of GPUs (e.g. V100 + A100 + RTX 30xx, with a CUDA 12 toolkit):

PATMINTON_CUDA_ARCH="70;80;86" pip install --no-binary patminton patminton

Quick example

import torch
from patminton import PAT, translation_rotation_system, least_squares_CG

infos = translation_rotation_system(
    transducer_radius=25e-3, transducer_height=7.5e-3,
    transducer_width=0.250e-3, transducer_pitch=0.298e-3,
    transducer_nbr_elements=64, transducer_wavelength=1500/5e6,
    grid_size=10e-3,
)

pat = PAT(201, 201, 201, 5e-3, 5e-3, 5e-3,
          nT=1024, tStart=12.5e-6, dt=16e-9, c=1500.0,
          mode='cylinder_lut', infos_transducers=infos)

p = torch.zeros((201, 201, 201), dtype=torch.float64, device='cuda')
p[100, 100, 100] = 1.0

s = pat @ p                     # forward:  signals from initial pressure
p_bp = pat.T @ s                # adjoint:  back-propagation

u, *_ = least_squares_CG(pat, s, M_inv=None, max_iter=50, lam=1e-4)

Documentation

patminton.readthedocs.io — installation, quickstart, physical and mathematical model, transducer models, reconstruction algorithms and key parameters.

To preview it locally:

pip install -r docs/requirements.txt
mkdocs serve      # live preview on http://127.0.0.1:8000

Tests

pip install .[test]
pytest              # CPU tests (no GPU needed)
pytest -m gpu       # operator tests on a CUDA GPU

Reference

The on-the-fly matrix-vector product strategy follows:

Lu Ding, Daniel Razansky, Xosé Luís Deán-Ben — "Model-based reconstruction of large three-dimensional optoacoustic datasets", IEEE Transactions on Medical Imaging, 2020.

If you use this package, please cite:

Trung-Thai Do, Paul Escande, Caroline Chaux, Jérôme Gateau, Hwee Kuan Lee — "Implementations of photoacoustic tomography models", 2026.