Anatomy-aware, label-informed approach improves image registration for challenging datasets
Rachel A Roston, Nicholas J Tustison, A Murat Maga · 2026
Read Anatomy-aware, label-informed approach improves image registration for challenging datasets by Rachel A Roston, Nicholas J Tustison, A Murat Maga (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/plos-pone-0357448.txt.
Description
Rachel A. Roston, Nicholas J. Tustison, and A. Murat Maga published a 2026 methods paper in PLOS ONE on aligning medical images when anatomy is strikingly abnormal. Registration, warping one image to match a template, underpins modern morphometrics, but standard tools align by brightness alone and fail when organs are rearranged. Their label-informed approach lets experts supply segmentations declaring which regions correspond, guiding the warp by biology rather than pixels, inside a widely used open ecosystem.
The demonstration was dramatic: mouse embryos lacking a key developmental gene develop severe spinal curvature scrambling their organs, and ordinary registration failed to match them to a normal template. With label guidance, the knockout embryos registered properly, restoring correspondence organ by organ, raising the power of measurements drawn from the alignment. The method is offered as a general fix for developmental and comparative studies, where the interesting subjects are those breaking ordinary registration.
Catalog record
| CREATOR | Rachel A Roston, Nicholas J Tustison, A Murat Maga |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60256 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |