Impact of Simulated Radiation Dose Reduction on Deep Learning-Based Renal Segmentation Performance: A Simulation Study Using the KiTS21 (2021 Kidney and Kidney Tumor Segmentation Challenge) Dataset
Jae-Seoung Kim, Sung-Jong Eun · 2026
Read Impact of Simulated Radiation Dose Reduction on Deep Learning-Based Renal Segmentation Performance: A Simulation Study Using the KiTS21 (2021 Kidney and Kidney Tumor Segmentation Challenge) Dataset by Jae-Seoung Kim, Sung-Jong Eun (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/pmc-13248953.txt.
Description
Jae-Seoung Kim and Sung-Jong Eun test how far CT dose can fall before AI kidney outlining suffers in a 2026 simulation study. Lower scan doses spare patients radiation, yet noise may blur the images that AI tools read. The team used a public kidney tumor dataset with expert outlines.
From 299 scans they simulated full, half, quarter, and tenth dose versions. A network trained only on full-dose images was then scored at each level. Accuracy held nearly flat at half dose and slipped a little at quarter dose. At tenth dose the drop grew clear, though scores stayed fair. The authors mark quarter dose as a workable floor for this task.
Catalog record
| CREATOR | Jae-Seoung Kim, Sung-Jong Eun |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |