This work introduces an unsupervised, optimization-based reconstruction framework that employs a flow-matching generative model as a learned manifold prior, yielding a differentiable generator of anatomically plausible images.
Abstract
Image reconstruction for positron emission tomography (PET) is an ill-posed Poisson inverse problem that often suffers from severe noise amplification and artifacts. In this work, we introduce an unsupervised, optimization-based reconstruction framework that employs a flow-matching generative model as a learned manifold prior. We train the flow-matching model on high-quality PET images to learn a deterministic ordinary differential equation transport from a Gaussian latent distribution to the empirical PET image distribution, yielding a differentiable generator of anatomically plausible images. We incorporate this generator as an explicit manifold constraint into a regularized Poisson likelihood formulation. We solve the resulting optimization problem using an alternating direction method of multipliers algorithm, in which an expectation-maximization-type surrogate update enforces data consistency and a gradient-based latent-space projection enforces manifold proximity. We evaluate the proposed method on both simulated and real PET datasets, assessing dose-level robustness, lesion-insertion generalization, and cross-scanner transfer. Compared with conventional reconstruction methods and state-of-the-art deep learning baselines, the proposed method provides superior noise suppression, structural preservation, and quantitative accuracy, while maintaining high computational efficiency.
Experimental results across several CT inverse problem settings show that Flow Matching-based approaches consistently outperform diffusion-based methods in terms of PSNR, SSIM, and perceptual quality, while requiring fewer sampling steps.
Generative models have shown strong potential for positron emission tomography (PET) image reconstruction. Although diffusion model-based reconstruction methods have demonstrated promising performance, they often require many reverse sampling steps with data-consistency updates incorporated into the sampling process. F...
F. Hashimoto, Ziqian Huang, Tatsuya Yokota et al.· 0 citations
The proposed approach integrates implicit neural representations with physics-consistent modeling and prior-guided regularization, providing an effective unsupervised framework for TOF-PET reconstruction and highlighting the potential of neural field representations for tomographic imaging.
Yu-Lin Zhang, Yuxuan Long, Hong Wang et al.· Physics in Medicine and Biol...· 0 citations
Positron emission tomography (PET) is a molecular imaging technique that creates an image of radiopharmaceutical distribution using acquired sinogram data. Accurate PET reconstruction requires a balance between data fidelity and regularization. The choice of penalty type and the adjustment of the regularization paramet...
Y. Moussaoui, Saïd Moussaoui, D. Mateus et al.· Physics in Medicine and Biol...· 0 citations
Reducing radiation exposure in Positron Emission Tomography (PET) is important for patient safety; however, ultra-low-dose imaging suffers from severe noise, which may affect diagnostic interpretation without appropriate image enhancement. While current 3D deep generative models, particularly diffusion models, have sho...
Jia-Le Shen, Guo-Lin Wang, Chen-Hao Wang et al.· arXiv.org· 0 citations
Medical image reconstruction began with an exact mathematical result. In 1917, Radon showed how to recover a function from its line integrals, providing the mathematical basis for computed tomography, positron emission tomography, and single-photon emission computed tomography. This review follows the development of re...
Nicholas E. Protonotarios, Sotiris Messinis, G. Kastis et al.· Mathematics· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.