{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:V5VOWTROSYMFNXPKWZRO5GTKY5","short_pith_number":"pith:V5VOWTRO","schema_version":"1.0","canonical_sha256":"af6aeb4e2e961856ddeab662ee9a6ac74bd09f5c86fd7b31f45da3c007334153","source":{"kind":"arxiv","id":"2204.09817","version":4},"attestation_state":"computed","paper":{"title":"Making the Most of Text Semantics to Improve Biomedical Vision--Language Processing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Aditya Nori, Anton Schwaighofer, Benedikt Boecking, Daniel C. Castro, Hoifung Poon, Javier Alvarez-Valle, Maria Wetscherek, Naoto Usuyama, Ozan Oktay, Shruthi Bannur, Stephanie Hyland, Tristan Naumann","submitted_at":"2022-04-21T00:04:35Z","abstract_excerpt":"Multi-modal data abounds in biomedicine, such as radiology images and reports. Interpreting this data at scale is essential for improving clinical care and accelerating clinical research. Biomedical text with its complex semantics poses additional challenges in vision--language modelling compared to the general domain, and previous work has used insufficiently adapted models that lack domain-specific language understanding. In this paper, we show that principled textual semantic modelling can substantially improve contrastive learning in self-supervised vision--language processing. We release "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2204.09817","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-04-21T00:04:35Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"881e892a9ea725d3bc299cae7d5d6f4d3b4a311c3b13943d293fe0a786e98f64","abstract_canon_sha256":"e76d144c8d9d7e17249d2e60672e46cf301fa82a170d544938488493fe084842"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:23:08.827191Z","signature_b64":"HQKboJwPpry9O5UPWoVYhPUVDSvhyjZxOZoIbhg+Ldd7KpEInymtUCAOMo/d9JpHyt7s4kw4QM5KtJ4tQ9MqDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"af6aeb4e2e961856ddeab662ee9a6ac74bd09f5c86fd7b31f45da3c007334153","last_reissued_at":"2026-07-05T05:23:08.826675Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:23:08.826675Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Making the Most of Text Semantics to Improve Biomedical Vision--Language Processing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Aditya Nori, Anton Schwaighofer, Benedikt Boecking, Daniel C. Castro, Hoifung Poon, Javier Alvarez-Valle, Maria Wetscherek, Naoto Usuyama, Ozan Oktay, Shruthi Bannur, Stephanie Hyland, Tristan Naumann","submitted_at":"2022-04-21T00:04:35Z","abstract_excerpt":"Multi-modal data abounds in biomedicine, such as radiology images and reports. Interpreting this data at scale is essential for improving clinical care and accelerating clinical research. Biomedical text with its complex semantics poses additional challenges in vision--language modelling compared to the general domain, and previous work has used insufficiently adapted models that lack domain-specific language understanding. In this paper, we show that principled textual semantic modelling can substantially improve contrastive learning in self-supervised vision--language processing. We release "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2204.09817","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2204.09817/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2204.09817","created_at":"2026-07-05T05:23:08.826736+00:00"},{"alias_kind":"arxiv_version","alias_value":"2204.09817v4","created_at":"2026-07-05T05:23:08.826736+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2204.09817","created_at":"2026-07-05T05:23:08.826736+00:00"},{"alias_kind":"pith_short_12","alias_value":"V5VOWTROSYMF","created_at":"2026-07-05T05:23:08.826736+00:00"},{"alias_kind":"pith_short_16","alias_value":"V5VOWTROSYMFNXPK","created_at":"2026-07-05T05:23:08.826736+00:00"},{"alias_kind":"pith_short_8","alias_value":"V5VOWTRO","created_at":"2026-07-05T05:23:08.826736+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00967","citing_title":"MedSyn2: Flexible Control of 3D CT Generation via Text and Semantically-Defined Segmentation Prompts","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15736","citing_title":"BiomedAP: A Vision-Informed Dual-Anchor Framework with Gated Cross-Modal Fusion for Robust Medical Vision-Language Adaptation","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5","json":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5.json","graph_json":"https://pith.science/api/pith-number/V5VOWTROSYMFNXPKWZRO5GTKY5/graph.json","events_json":"https://pith.science/api/pith-number/V5VOWTROSYMFNXPKWZRO5GTKY5/events.json","paper":"https://pith.science/paper/V5VOWTRO"},"agent_actions":{"view_html":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5","download_json":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5.json","view_paper":"https://pith.science/paper/V5VOWTRO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2204.09817&json=true","fetch_graph":"https://pith.science/api/pith-number/V5VOWTROSYMFNXPKWZRO5GTKY5/graph.json","fetch_events":"https://pith.science/api/pith-number/V5VOWTROSYMFNXPKWZRO5GTKY5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5/action/storage_attestation","attest_author":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5/action/author_attestation","sign_citation":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5/action/citation_signature","submit_replication":"https://pith.science/pith/V5VOWTROSYMFNXPKWZRO5GTKY5/action/replication_record"}},"created_at":"2026-07-05T05:23:08.826736+00:00","updated_at":"2026-07-05T05:23:08.826736+00:00"}