{
  "oags": "0.2",
  "status": "alpha",
  "entry": {
    "id": "IB-arxiv-2606.29949-histopathology-retrieval-25x",
    "type": "evidence",
    "claim": "On a multi-cancer cohort (N=1,720), contrastive alignment of frozen histopathology and RNA-seq foundation models yielded a 25-fold improvement in open-vocabulary molecular retrieval over baseline methods.",
    "evidence_grade": "E3",
    "state": "active",
    "primary_source_ref": [
      {
        "kind": "arxiv",
        "id": "2606.29949",
        "version": "v1",
        "url": "https://arxiv.org/abs/2606.29949",
        "locator": "abstract"
      }
    ],
    "quantities": {
      "n_cohort": 1720,
      "retrieval_improvement_fold": 25
    },
    "declared_loss": null,
    "source_title": "Data-Efficient Multimodal Alignment for Histopathology-based Molecular Prediction",
    "authors": [
      "Dominik Winter",
      "Dominik Vonficht",
      "Loïc Le Bescond",
      "Christian Gebbe",
      "Marco Rosati",
      "Richard J. Chen",
      "Markus Schick",
      "Ross Stewart",
      "Nicolas Brieu"
    ],
    "result_origin": "measured",
    "support_quote": "Using contrastive learning on a multi-cancer cohort (N=1,720), we achieve a 25-fold improvement in retrieval over baseline methods",
    "source_version": "v1",
    "abstract_check": {
      "version": "v1",
      "date": "2026-10-08",
      "quote_verbatim": true
    },
    "domain_v2": "med/imaging"
  },
  "expand": [],
  "incoming": [],
  "links": [],
  "declared_losses": [
    "Public evidence-graph projection: internal analytic/theory edges and private provenance are withheld by design.",
    "Alpha: some primary-source refs await precise citation verification."
  ],
  "generated": "2026-10-08"
}