Metraipolis › archive › dc_6e041ee53cd4

MedHEB: Benchmarking Medical Embeddings Across Heterogeneous Clinical Evidence — arXiv:2611.02519

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.02519
License (as archived)
CC0-1.0
Archived
2026-11-03T18:55:00.872Z
SHA-256 of the archived text
823faf51dd197c32b3e244abef7f12f775093fb6fc29461fdefce662f2e675f2
Stable URL of this page
https://metraipolis.com/documento/dc_6e041ee53cd4

How to cite

MedHEB: Benchmarking Medical Embeddings Across Heterogeneous Clinical Evidence — arXiv:2611.02519 — arXiv (en), https://arxiv.org/abs/2611.02519 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6e041ee53cd4 (Metraipolis archive dc_6e041ee53cd4, sha256 823faf51dd197c32b3e244abef7f12f775093fb6fc29461fdefce662f2e675f2)
BibTeX
@misc{dc6e041ee53cd4,
  title = {MedHEB: Benchmarking Medical Embeddings Across Heterogeneous Clinical Evidence — arXiv:2611.02519},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02519},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6e041ee53cd4; sha256 823faf51dd197c32b3e244abef7f12f775093fb6fc29461fdefce662f2e675f2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02519. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Clinical evidence is inherently relational: medical images are interpreted with prior studies, cross-modal examinations, reports, and localized findings, and retrieving relevant evidence is therefore central to clinical interpretation. Recent foundation models increasingly encode these heterogeneous sources into shared embedding spaces, yet existing embedding evaluation often focuses on isolated matches, leaving unclear whether current models retrieve relevant evidence across changes in modality, appearance, and granularity. We introduce MedHEB, a unified benchmark with 107 retrieval tasks from 54 datasets, including both in-distribution and held-out-source evaluation, across three axes: evidence form (2D images, 3D volumes, text), pairing type (image--image, image--text, text--text, and modality-bridging visual matching), and relevance granularity (global, question-conditioned, region-level). Under a single ranking protocol, MedHEB varies clinically relevant evidence definitions to test whether embedding proximity remains meaningful across heterogeneous evidence, including modality-bridging visual retrieval and region-level text-to-region retrieval. Across representative models, performance is uneven across retrieval families: models strong on conventional image--text or text--text retrieval are less reliable on modality-bridging and region-level tasks. Current embedding spaces capture some clinical relationships well but do not yet provide balanced retrieval across heterogeneous evidence. We further provide MedEmb, an MLLM-based reference embedding model trained on the MedHEB ID split, as a reproducible baseline for unified 2D--3D--text medical retrieval. We release MedHEB with preprocessing scripts, evaluation code, access manifests, and MedEmb to support research on clinically meaningful medical embeddings.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.02519 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T18:16:53.000Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).