analiza aiqbench
This paper proposes a diffusion-based CT super-resolution method that uses an explicit 3D vascular representation to keep generated vessels anatomically plausible, rather than just visually convincing. The reported gains are in segmentation accuracy and vascular connectivity on two datasets, with image quality said to stay close to the backbone. For readers, the useful point is the general pattern of adding structured, verifiable constraints to generative models in high-stakes imaging, though the results still rest on a small set of benchmarks and re-annotated ground truth.
VascuSR: Iterative Anatomy-Aware Diffusion for CT Super-Resolution
arXiv:2610.20509v1 Announce Type: cross Abstract: Generative super-resolution (SR) can recover sharp detail in computed tomography (CT) images from severely downsampled observations, but the missing high-frequency content is not uniquely determined by the low-resolution input. Consequently, visually plausible reconstructions may contain anatomically unsupported vascular detail, such as missing, spurious, or incorrectly connected vessels. To overcome this challenge, we introduce VascuSR, an iterative anatomy-aware diffusion framework that combines image reconstruction with explicit 3D vascular guidance. First, to provide richer anatomical guidance, we use an explicit 3D vascular representation that captures structural information beyond individual slices. This representation also simplifies vascular refinement by allowing a lightweight 3D generative refiner to focus on vascular structure rather than complex CT appearance. Second, to guide slice-wise SR, a lightweight anatomy controller injects the refined structure into the diffusion decoder, while iterative feedback updates this guidance from each new SR volume. On re-annotated MSD Task08 and HVM datasets, VascuSR outperforms interpolation and learning-based SR baselines in segmentation accuracy (Dice and clDice) and 3D vascular structure (connectivity and mean centerline distance), while maintaining image quality comparable to its SR backbone. Compared with learning-based baselines, VascuSR reduces class-averaged connected-component counts by over 20% on both datasets and mean centerline distance by over 10% on MSD.
Abstract:Generative super-resolution (SR) can recover sharp detail in computed tomography (CT) images from severely downsampled observations, but the missing high-frequency content is not uniquely determined by the low-resolution input. Consequently, visually plausible reconstructions may contain anatomically unsupported vascular detail, such as missing, spurious, or incorrectly connected vessels. To overcome this challenge, we introduce VascuSR, an iterative anatomy-aware diffusion framework that combines image reconstruction with explicit 3D vascular guidance. First, to provide richer anatomical guidance, we use an explicit 3D vascular representation that captures structural information beyond individual slices. This representation also simplifies vascular refinement by allowing a lightweight 3D generative refiner to focus on vascular structure rather than complex CT appearance. Second, to guide slice-wise SR, a lightweight anatomy controller injects the refined structure into the diffusion decoder, while iterative feedback updates this guidance from each new SR volume. On re-annotated MSD Task08 and HVM datasets, VascuSR outperforms interpolation and learning-based SR baselines in segmentation accuracy (Dice and clDice) and 3D vascular structure (connectivity and mean centerline distance), while maintaining image quality comparable to its SR backbone. Compared with learning-based baselines, VascuSR reduces class-averaged connected-component counts by over 20% on both datasets and mean centerline distance by over 10% on MSD.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
arXiv:2610.20509 [cs.CV]
(or arXiv:2610.20509v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2610.20509 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 304 cuvinte
Subiecte conexe
TOLA: Text-aware One-Step Latent Adaptation for Diffusion-based Text Image Super-Resolution
arXiv · 25 sept. 2026
Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · 25 sept. 2026
Iterative Refinement Diffusion for Super-Resolved Data Assimilation of Multiscale Physical Systems
arXiv · 19 aug. 2026
Perceptually Regularized Diffusion Model for Image Super-Resolution
arXiv · 3 sept. 2026