Design Choices in Splitting-Based Self-Supervised Sparse-View CT Reconstruction

DOI10.48550/arXiv.2607.10898MaRDI QIDQ7284072arXivFDO

Authors Nadja Gruber, Lukas Neumann, Ander Biguri, Gyeongha Hwang, Markus Haltmeier, Johannes Schwab

Publication date 12 July 2026

Published in arXiv

Tldr We introduce a unified framework to decompose splitting-based self-supervised sparse-view CT reconstruction into partitioning strategy, preprocessing, and inference. Our experiments show that optimal design choices depend on the measurement noise structure.

Full work available at URL https://doi.org/10.48550/arXiv.2607.10898

arXiv classification

cs.CVmath.NA

Cites work

• Nonlinear total variation based noise removal algorithms
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Total generalized variation
• U-Net: Convolutional Networks for Biomedical Image Segmentation

This page was built for publication: Design Choices in Splitting-Based Self-Supervised Sparse-View CT Reconstruction

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7284072)