Assessing the transferability of micro-computed tomography (CT)-based plaque segmentation to clinical CT using super-resolution: Limits and emerging perspectives
Salomé H Kuntz, Katherine Morales, Hugo Gangloff, Yohan Petetin, Anne Lejay, Nabil Chakfé · 2026
Read Assessing the transferability of micro-computed tomography (CT)-based plaque segmentation to clinical CT using super-resolution: Limits and emerging perspectives by Salomé H Kuntz, Katherine Morales, Hugo Gangloff, Yohan Petetin, Anne Lejay, Nabil Chakfé (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/PMC13393156.txt.
Description
Salomé H Kuntz and five colleagues tested whether artificial intelligence trained to outline plaque on ultra-fine laboratory scans could carry that skill to ordinary hospital CT images sharpened by super-resolution software. Leg artery pieces from six patients with peripheral artery disease were imaged both ways and checked against tissue study. Neural networks learned calcified plaque outlines from the fine scans, then were applied to sharpened clinical images without retraining.
Even on fine test images the outlines agreed only modestly with truth, with overlap scores in the low-to-middle range. On sharpened clinical scans the models did pick out calcified structures in a non-random way in some sequences, but results swung widely between slices. The authors conclude that sharpening cannot bridge the basic resolution gap, and they frame the work as a map of current limits rather than a tool ready for clinic use.
Catalog record
| CREATOR | Salomé H Kuntz, Katherine Morales, Hugo Gangloff, Yohan Petetin, Anne Lejay, Nabil Chakfé |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17184 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |