{
  "abstract": "Histology images contain rich morphological information and can provide insights into pathological processes. However, deriving hypotheses relating morphological phenotypes to clinical attributes is bottlenecked by a manual image interpretation step. Here, we demonstrate that this process can be automated through interpretable deep learning. We present SCOPE, a method to interpret slide-level classifiers by combining pathology-specific vision--language models with sparse concept attribution onto a generalist histomorphological concept bank. To measure whether such explanations recover known morphology, we introduce MorphoRecoveryBench, a benchmark of seven tasks with pathologist-curated reference descriptions. On this benchmark, dense concept attribution is indistinguishable from a random baseline, whereas sparse attribution recovers substantial known morphology; decomposing the pooled slide embedding reaches similar explanation correctness at a fraction of the computational cost. Post-hoc interpretation of whole-slide classifiers can thus generate morphological hypotheses at scale, for expert validation.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.02985",
  "author_total": 9,
  "authors": [
    {
      "name": "Tristan Lazard"
    },
    {
      "name": "Kenza Bouzid"
    },
    {
      "name": "Julius Hense"
    },
    {
      "name": "Shruthi Bannur"
    },
    {
      "name": "Daniel Coelho de Castro"
    },
    {
      "name": "Daniel Shao"
    },
    {
      "name": "Rajesh Jena"
    },
    {
      "name": "Drew Williamson"
    },
    {
      "name": "Stephanie Hyland"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/2m0t6t5tj7e69.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.02985",
  "fulltext_url": null,
  "id": "2m0t6t5tj7e69",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.02985",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "eess.IV"
    ],
    [
      "arxiv",
      "q-bio.QM"
    ],
    [
      "openalex",
      "1702|Explainable Artificial Intelligence (XAI)"
    ]
  ],
  "openalex": "W7208777689",
  "pdf_url": "https://arxiv.org/pdf/2609.02985",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-02",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/2m0t6t5tj7e69.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:18:37Z",
      "native_id": "2609.02985",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.02985"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "arxiv:q-bio.QM",
      "subfield": "bioinformatics"
    }
  ],
  "title": "Sparse concept attribution for histomorphological hypothesis generation from whole-slide classifiers",
  "title_text": "Sparse concept attribution for histomorphological hypothesis generation from whole-slide classifiers",
  "type": "preprint",
  "updated_at": "2026-10-30T12:18:37Z",
  "url": "https://research.draconitic.com/paper/2m0t6t5tj7e69",
  "venue": "arXiv",
  "versions": []
}