Article

FEATMAP: Targeted Correction of Acquisition Signatures Harmonizes Medical Foundation Model Embeddings and Enables Robust Task Generalization

2026-07-08

Abstract excerpt

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 preserv...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: d0fed128-9630-58b0-9085-8d809734dbdc
DOI: 10.64898/2026.07.02.736184

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

FEATMAP: Targeted Correction of Acquisition Signatures Harmonizes Medical Foundation Model Embeddings and Enables Robust Task GeneralizationDOI 10.64898/2026.07.02.736184Read

Normal is All You Need: A Symmetry-Informed Inverse Learning Foundation Model for Neuroimaging DiagnosticsSemantic proximity Read

Metadata Supervised Imaging Representations for Modelling and Controlling Acquisition VariabilitySemantic proximity Read

Cracks in the Foundation: How Data-Hungry and Sensitive to Domain Shift are Vision Foundation Models for Computational Pathology?Semantic proximity Read

Foundation model embeddings for quantitative tumor imaging biomarkersSemantic proximity Read

MICAFlow: Fast and robust MRI preprocessing bridging research neuroimaging and clinical practiceSemantic proximity Read

Comparison of different methods for average anatomical templates creation: do we really gain anything from a diffeomorphic framework?Semantic proximity Read

Modeling patient tissues at molecular resolution with EvaSemantic proximity Read

Domain Shift-Robust Tumor Segmentation Across HospitalsSemantic proximity Read

Anatomy-guided, modality-agnostic segmentation of neuroimaging abnormalitiesSemantic proximity Read

DUNE: a versatile neuroimaging encoder captures brain complexity across three major diseases: cancer, dementia and schizophreniaSemantic proximity Read

A Survey of Generalization and Adaptation in Medical Imaging Foundation ModelsSemantic proximity Read

A foundation model for generalized brain MRI analysisSemantic proximity Read