Toward a Foundation Plug-and-Play Prior for Computed Tomography Reconstruction via a Multimodal Diffusion Model
2026-08-24 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors explore a way to speed up CT scans by using a single neural network model that works for different types of CT imaging problems without needing to be retrained for each one. They use a diffusion model trained on multiple imaging setups and test it on three different kinds of CT scans with different materials and shapes. Their method consistently produces better images than traditional techniques, showing it could be a flexible tool for various CT applications. This could help reduce scanning time and improve image quality across different CT uses.
Computed Tomography (CT)Diffusion ModelNeural Network PriorSparse-View ReconstructionCone-Beam X-ray CTParallel-Beam Neutron CTImaging ModalityBeam GeometryAdditive ManufacturingImage Reconstruction
Authors
Haley Duba-Sullivan, Patxi Fernandez-Zelaia, Obaidullah Rahman, Amirkoushyar Ziabari
Abstract
Computed tomography (CT) throughput is limited by scan time, which grows with both the number of projections acquired and the detector integration time for each. Reconstructing high-quality volumes from sparse-view or low-dose measurements therefore depends on an informative prior, typically a neural network trained for one specific scan setting and retrained whenever the modality, geometry, or material changes. We investigate whether a single diffusion model trained across several imaging domains can instead serve as a prior for many CT problems simultaneously. We evaluate the proposed method using the same frozen model on three datasets that differ in modality, beam geometry, material, and degradation type, spanning flaw analysis in additively manufactured metal parts imaged with cone-beam X-ray CT and concrete microstructure imaged with parallel-beam neutron CT. Our proposed method out-performs analytic reconstructions in all three cases, providing a step toward a reusable foundation prior for heterogeneous CT reconstruction problems.