![[feed]](/style/images/feed-icon-14x14.png){kind=icon}
Tools
Tools
Preview |
PDF (Preprint incl. Supplementary Material)
- Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
17MB |
| Item Type: | Preprint |
|---|---|
| Title: | FEATMAP: targeted correction of acquisition signatures harmonizes medical foundation model embeddings and enables robust task generalization |
| Creators: |
Donle, Leonhard  ORCID: https://orcid.org/0000-0002-7931-4621, Phillips, Michael, Gaber, Farieda ORCID: https://orcid.org/0009-0004-5512-8527, Ramesh, Siddhi, Sacco, Matteo, Hautaniemi, Sampsa, Virtanen, Anni, Bressem, Keno, Adams, Lisa, Goon, Kelsey, Nevins, Elena, Robinett, Ryan A., Kochanny, Sara, Hassan, Sasha, Dolezal, James, Pearson, Alexander T. and Lengyel, Ernst
|
| Abstract: | Medical foundation models compress biomedical data into embeddings that support diverse downstream clinical tasks. However, successful model deployment is hampered by performance degradation on external data. It is recognized that embeddings capture acquisition signatures, such as hardware and technical differences, in addition to biology. Effective harmonization must remove the acquisition signature while preserving biological signals, a trade-off that current methods fail to balance adequately. Input-level normalization fails to eliminate acquisition signatures from embeddings, whereas embedding-level methods adjust features in an untargeted manner. We present FEATMAP, a harmonization approach that models acquisition signatures as geometric distortions between manifolds of similarly arranged embeddings. Using paired data that isolate the effect of acquisition signatures, FEATMAP fits a single global affine transformation per foundation model to correct acquisition signatures directly in the embedding space. This targeted, reusable correction aims to preserve biological and demographic variation while harmonizing across acquisition signatures. Across scanner and foundation-model harmonization in digital pathology and field-strength harmonization in brain MRI, FEATMAPimprovescross-condition embedding similarity, reduces performance gaps without retraining, and suggests potential for the alignment of disparate embedding spaces. |
| Source: | bioRxiv |
| Publisher: | Cold Spring Harbor Laboratory Press |
| Article Number: | 2026.07.02.736184 |
| Date: | 8 July 2026 |
| Official Publication: | https://doi.org/10.64898/2026.07.02.736184 |
Repository Staff Only: item control page
