AI-based detection of temporal changes in MR-Linac images acquired during routine prostate radiotherapy
Seungbin Park, Peilin Wang, Joshua Paik, Ryan Pennell, Francesca Khani, Himanshu Nagar, Timothy McClure, Emily S Weg, Mert R Sabuncu, Daniel Margolis, Heejong Kim · 2026
Read AI-based detection of temporal changes in MR-Linac images acquired during routine prostate radiotherapy by Seungbin Park, Peilin Wang, Joshua Paik, Ryan Pennell, Francesca Khani, Himanshu Nagar, Timothy McClure, Emily S Weg, Mert R Sabuncu, Daniel Margolis, Heejong Kim (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-13634774.txt.
Description
Radiotherapy machines that carry their own MRI scanner take pictures at every treatment session, and Seungbin Park and co-authors ask whether software can spot the small changes between sessions. They trained a deep learning model on scans from 761 prostate patients. Its job was to judge which of two images came later, a temporal ordering task that forces the model to learn real change.
The model excelled: scores near the ceiling, beating a radiologist at the ordering task. Accuracy faded for scan pairs closer in time, as expected, and the regions driving its answers were the prostate, bladder, and a pelvic joint. Intriguingly, patients whose cancer later returned showed smaller model-detected changes during treatment. The authors see a path toward imaging markers of treatment response from scans that clinics already collect.
Catalog record
| CREATOR | Seungbin Park, Peilin Wang, Joshua Paik, Ryan Pennell, Francesca Khani, Himanshu Nagar, Timothy McClure, Emily S Weg, Mert R Sabuncu, Daniel Margolis, Heejong Kim |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |