Eliminating Registration Bias in Synthetic CT Generation: A Physics-Based Simulation Framework Journal: Medical Image Analysis

Lukas Zimmermann, Michael Rauter, Maximilian Schmid, Dietmar Georg, Barbara Knäusl · 2026

Read Eliminating Registration Bias in Synthetic CT Generation: A Physics-Based Simulation Framework Journal: Medical Image Analysis by Lukas Zimmermann, Michael Rauter, Maximilian Schmid, Dietmar Georg, Barbara Knäusl (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-02130.txt.

Description

Lukas Zimmermann and colleagues study how flawed alignment warps scans made by machines that turn cone beam X-ray views into CT-like images. Perfect matching of scans taken apart can never be reached, so flaws seep into trained models and skew scores. They build pairs with a physics based simulation that lines views up by design, judging results by how well shapes line up with the input scan. On two pelvic data sets, models trained this way matched shapes better, and readers picked their outputs in most cases.

Catalog record

CREATORLukas Zimmermann, Michael Rauter, Maximilian Schmid, Dietmar Georg, Barbara Knäusl
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE51812
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602.02130

Browse the full library · About the Tower of Babel