Saved in:
Bibliographic Details
Main Authors: Webber, George, Hammers, Alexander, King, Andrew P., Reader, Andrew J.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.13441
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918161014587392
author Webber, George
Hammers, Alexander
King, Andrew P.
Reader, Andrew J.
author_facet Webber, George
Hammers, Alexander
King, Andrew P.
Reader, Andrew J.
contents Diffusion models have recently enabled state-of-the-art reconstruction of positron emission tomography (PET) images while requiring only image training data. However, domain shift remains a key concern for clinical adoption: priors trained on images from one anatomy, acquisition protocol or pathology may produce artefacts on out-of-distribution data. We propose integrating steerable conditional diffusion (SCD) with our previously-introduced likelihood-scheduled diffusion (PET-LiSch) framework to improve the alignment of the diffusion model's prior to the target subject. At reconstruction time, for each diffusion step, we use low-rank adaptation (LoRA) to align the diffusion model prior with the target domain on the fly. Experiments on realistic synthetic 2D brain phantoms demonstrate that our approach suppresses hallucinated artefacts under domain shift, i.e. when our diffusion model is trained on perturbed images and tested on normal anatomy, our approach suppresses the hallucinated structure, outperforming both OSEM and diffusion model baselines qualitatively and quantitatively. These results provide a proof-of-concept that steerable priors can mitigate domain shift in diffusion-based PET reconstruction and motivate future evaluation on real data.
format Preprint
id arxiv_https___arxiv_org_abs_2510_13441
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Steerable Conditional Diffusion for Domain Adaptation in PET Image Reconstruction
Webber, George
Hammers, Alexander
King, Andrew P.
Reader, Andrew J.
Medical Physics
Computer Vision and Pattern Recognition
Machine Learning
Diffusion models have recently enabled state-of-the-art reconstruction of positron emission tomography (PET) images while requiring only image training data. However, domain shift remains a key concern for clinical adoption: priors trained on images from one anatomy, acquisition protocol or pathology may produce artefacts on out-of-distribution data. We propose integrating steerable conditional diffusion (SCD) with our previously-introduced likelihood-scheduled diffusion (PET-LiSch) framework to improve the alignment of the diffusion model's prior to the target subject. At reconstruction time, for each diffusion step, we use low-rank adaptation (LoRA) to align the diffusion model prior with the target domain on the fly. Experiments on realistic synthetic 2D brain phantoms demonstrate that our approach suppresses hallucinated artefacts under domain shift, i.e. when our diffusion model is trained on perturbed images and tested on normal anatomy, our approach suppresses the hallucinated structure, outperforming both OSEM and diffusion model baselines qualitatively and quantitatively. These results provide a proof-of-concept that steerable priors can mitigate domain shift in diffusion-based PET reconstruction and motivate future evaluation on real data.
title Steerable Conditional Diffusion for Domain Adaptation in PET Image Reconstruction
topic Medical Physics
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2510.13441