Segmentation-Guided CT Synthesis with Pixel-Wise Conformal Uncertainty Bounds
David Vallmanya Poch, Yorick Estievenart, Elnura Zhalieva, Sukanya Patra, Mohammad Yaqub, Souhaib Ben Taieb · 2025
Read Segmentation-Guided CT Synthesis with Pixel-Wise Conformal Uncertainty Bounds by David Vallmanya Poch, Yorick Estievenart, Elnura Zhalieva, Sukanya Patra, Mohammad Yaqub, Souhaib Ben Taieb (2025) 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-2503-08515.txt.
Description
David Vallmanya Poch and five co-authors present a method that turns low-quality cone beam scans into full CT images for proton therapy planning, in a 2025 paper. Proton dose calculations need sharp CT images, but the cone beam scans taken during treatment are cheaper, gentler, and marred by artefacts. Learning-based conversion exists, yet it can warp anatomy and gives no sense of its own reliability.
Their framework guides the conversion with organ outlines taken from the planning scan, which keeps anatomy consistent. A second part attaches a per-pixel uncertainty range to each generated image, using conformal prediction. Tests on two datasets showed better accuracy on a new soft-tissue-focused score and better-calibrated uncertainty, a step toward clinical trust.
Catalog record
| CREATOR | David Vallmanya Poch, Yorick Estievenart, Elnura Zhalieva, Sukanya Patra, Mohammad Yaqub, Souhaib Ben Taieb |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-08515