Fusion method improves recovery of missing image structures in medical scans
Consecutive Posterior Fusion for Diffusive Recovery of Unobservable Image Structures
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Recovering details in images that are hard to see or missing is a tough problem, especially in medical scans. The authors propose a new technique that combines multiple steps of a special image restoration process to better fill in these missing parts. Their method works without needing to retrain the model or do extra calculations, making it efficient. Tests show it improves results for CT scans and medical image sharpening compared to previous methods.
What this means in practice
- •For medical imaging teams: Enhance image reconstruction quality in sparse-view and low-dose CT scans by improving recovery of missing image structures without retraining models.
- •For medical image processing developers: Develop super-resolution tools for medical images using fused posterior estimates to better reconstruct fine details.
Authors
Elena Morotti, Davide Evangelista, Elena Loli Piccolomini
Abstract
Solving severely ill-posed imaging inverse problems requires recovering image structures that are unobservable or weakly constrained by the measurements. Diffusion models provide expressive learned priors for inferring such missing information, while posterior sampling incorporates measurement consistency along the reverse process. Standard diffusion posterior samplers, however, rely on instantaneous measurement-aware estimates, without explicitly exploiting information carried by previous posterior corrections. We introduce Consecutive Posterior Fusion Denoising Diffusion Null-Space Models (CPF-DDNM), an inference-time strategy that fuses consecutive measurement-aware estimates to improve the diffusive recovery of unobservable image structures, without requiring retraining or additional denoiser evaluations. We instantiate this principle within DDNM, whose range/null-space decomposition reveals that consecutive fusion preserves the measurement-determined component while acting exclusively on the prior-driven null-space estimate. We thus provide a geometric interpretation of CPF-DDNM and a local error analysis that characterizes the optimal time-dependent fusion coefficient, including the extrapolative regime. Experiments on sparse-view and simulated low-dose computed tomography, as well as medical image super-resolution, show consistent improvements over DDNM and competitive performance against diffusion-based inverse solvers.