{"record": {"aliases": ["PathReportEval: A Systematic Benchmark for Pathology Report Generation"], "artifacts": [{"id": "arxiv:2607.18448", "kind": "paper", "url": "https://arxiv.org/abs/2607.18448"}], "categories": ["Biology & Drug Discovery", "Language & Knowledge", "cs.CL"], "description": {"en": "A standardized benchmark for generating pathology reports from whole-slide images. It covers TCGA, HistAI and REG 2025 with three foundation encoders and scores reports with the Clinical Report Quality Score (CRQS)."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.18448", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.18448", "title": "PathReportEval"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.18448", "modality": null, "name": "PathReportEval", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "false", "origin_record_id": "bm_pathreporteval-a-systematic-benchmark-for-_445812b1", "origin_source": "arxiv", "recognition_confidence": "0.55", "record_sha256": "56371cf6ab89cfd84227eb5025931c0eca3d72a1a46959aeadda9e35ec649d8d", "relation": "evaluates_only", "review_model": "claude-code-default", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-24T20:45:26.375696Z", "source_url": "https://arxiv.org/abs/2607.18448"}, "publisher": null, "released": "2026-07-20", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.18448", "source_url": "https://arxiv.org/abs/2607.18448"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-18448", "source": "claire_radar", "source_benchmark_id": "2607.18448", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["PathReportEval: A Systematic Benchmark for Pathology Report Generation"], "annotation": "Unknown", "applicationDomains": ["Biology & Drug Discovery"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "This is a niche medical-imaging benchmark with a new clinical metric. Interest is likely to stay within computational pathology, and no public code link is supplied yet.", "score": 45}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T20:45:26.375696Z", "inputHash": "e97e985f32fbc3e19ea328f1e0c7f9214dbd9771b349b68cf611ba515d52d919", "model": "claude-code-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "PathReportEval: A Systematic Benchmark for Pathology Report Generation", "canonicalNameSource": "paper_title", "decisionReason": "The benchmark name and CRQS scoring procedure are supported, but no actionable code, data, project, leaderboard, or demo link is supplied. The abstract's use of the word public is insufficient to verify reuse.", "evaluationEvidence": {"reuse": "the benchmark, public plug-and-play framework, and CRQS establish a reproducible foundation", "scoring": "measures four complementary dimensions: clinical fact coverage, key information recall, hallucination rate, and clinical discordance", "task": "Pathology report generation from whole-slide images (WSIs)"}, "model": "claude-code-default", "reviewedAt": "2026-09-24T20:45:26.375696Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A standardized benchmark for generating pathology reports from whole-slide images. It covers TCGA, HistAI and REG 2025 with three foundation encoders and scores reports with the Clinical Report Quality Score (CRQS).", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "benchmark study on existing or public datasets without a named artifact release"], "snippet": "We present a standardized benchmark and evaluation framework for pathology report generation."}, "familyId": "bmf_332ab570f521", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_pathreporteval-a-systematic-benchmark-for-_445812b1", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Pharma & Biotech"], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2607.18448", "pdf": "https://arxiv.org/pdf/2607.18448", "project": null, "report": "https://arxiv.org/abs/2607.18448"}, "metrics": [], "motivation": "Pathology report generation from whole-slide images (WSIs) is a rapidly growing multimodal learning problem, yet progress is difficult to measure because existing studies use heterogeneous datasets, model settings, visual encoders, and evaluation protocols.", "name": "PathReportEval", "oneLine": "A standardized benchmark for generating pathology reports from whole-slide images. It covers TCGA, HistAI and REG 2025 with three foundation encoders and scores reports with the Clinical Report Quality Score (CRQS).", "paperTitle": "PathReportEval: A Systematic Benchmark for Pathology Report Generation", "primaryDomain": "Biology & Drug Discovery", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.18448", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Paper only", "recognitionConfidence": 0.55, "relation": "evaluates_only", "releasedAt": "2026-07-20", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Suryakant Singh", "Sejuti Majumder", "Beatrice Knudsen", "Joel Saltz", "Prateek Prasanna"], "categories": ["cs.CL", "cs.CV"], "id": "2607.18448", "title": "PathReportEval: A Systematic Benchmark for Pathology Report Generation", "type": "arxiv", "url": "https://arxiv.org/abs/2607.18448"}, "sourceUpdatedAt": "2026-07-20T18:59:47Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Lexical metrics such as BLEU and ROUGE can miss clinically important errors. PathReportEval standardizes the evaluation pipeline and adds a clinically grounded score for factual correctness, which makes pathology report generation models easier to compare."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
