{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:SSLOKSFKR5M34VC37UN5IKFGV7","short_pith_number":"pith:SSLOKSFK","schema_version":"1.0","canonical_sha256":"9496e548aa8f59be545bfd1bd428a6afce3b0e7caef882e624480383d1bf7c08","source":{"kind":"arxiv","id":"2306.17555","version":2},"attestation_state":"computed","paper":{"title":"Why does my medical AI look at pictures of birds? Exploring the efficacy of transfer learning across domain boundaries","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Constantin Seibold, Enrico Nasca, Felix Nensa, Frederic Jonske, Jan Egger, Janis Evers, Jens Kleesiek, Johannes Haubold, Michael Kamp, Moon Kim, Ren\\'e Hosch","submitted_at":"2023-06-30T11:15:26Z","abstract_excerpt":"It is an open secret that ImageNet is treated as the panacea of pretraining. Particularly in medical machine learning, models not trained from scratch are often finetuned based on ImageNet-pretrained models. We posit that pretraining on data from the domain of the downstream task should almost always be preferred instead. We leverage RadNet-12M, a dataset containing more than 12 million computed tomography (CT) image slices, to explore the efficacy of self-supervised pretraining on medical and natural images. Our experiments cover intra- and cross-domain transfer scenarios, varying data scales"},"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":"2306.17555","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2023-06-30T11:15:26Z","cross_cats_sorted":[],"title_canon_sha256":"b61ab946376f70ddd7ca86e58312255438ed270c245cc81ad208c4d4a7a59a63","abstract_canon_sha256":"ff206e7ab4c1bf21d055a242beaed491bd85c3b2d85480479bac576852a24e92"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:14:01.609713Z","signature_b64":"ZTzp50jMt/2FwGXEztnyb1LRfj/YYBYwWx/546oDHeV7x15t3iyQDB8SIOJQIDob3C1QCAeXrksQRBtQY9pXBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9496e548aa8f59be545bfd1bd428a6afce3b0e7caef882e624480383d1bf7c08","last_reissued_at":"2026-07-05T10:14:01.609276Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:14:01.609276Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Why does my medical AI look at pictures of birds? Exploring the efficacy of transfer learning across domain boundaries","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Constantin Seibold, Enrico Nasca, Felix Nensa, Frederic Jonske, Jan Egger, Janis Evers, Jens Kleesiek, Johannes Haubold, Michael Kamp, Moon Kim, Ren\\'e Hosch","submitted_at":"2023-06-30T11:15:26Z","abstract_excerpt":"It is an open secret that ImageNet is treated as the panacea of pretraining. Particularly in medical machine learning, models not trained from scratch are often finetuned based on ImageNet-pretrained models. We posit that pretraining on data from the domain of the downstream task should almost always be preferred instead. We leverage RadNet-12M, a dataset containing more than 12 million computed tomography (CT) image slices, to explore the efficacy of self-supervised pretraining on medical and natural images. Our experiments cover intra- and cross-domain transfer scenarios, varying data scales"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.17555","kind":"arxiv","version":2},"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/2306.17555/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":"2306.17555","created_at":"2026-07-05T10:14:01.609345+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.17555v2","created_at":"2026-07-05T10:14:01.609345+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.17555","created_at":"2026-07-05T10:14:01.609345+00:00"},{"alias_kind":"pith_short_12","alias_value":"SSLOKSFKR5M3","created_at":"2026-07-05T10:14:01.609345+00:00"},{"alias_kind":"pith_short_16","alias_value":"SSLOKSFKR5M34VC3","created_at":"2026-07-05T10:14:01.609345+00:00"},{"alias_kind":"pith_short_8","alias_value":"SSLOKSFK","created_at":"2026-07-05T10:14:01.609345+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7","json":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7.json","graph_json":"https://pith.science/api/pith-number/SSLOKSFKR5M34VC37UN5IKFGV7/graph.json","events_json":"https://pith.science/api/pith-number/SSLOKSFKR5M34VC37UN5IKFGV7/events.json","paper":"https://pith.science/paper/SSLOKSFK"},"agent_actions":{"view_html":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7","download_json":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7.json","view_paper":"https://pith.science/paper/SSLOKSFK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.17555&json=true","fetch_graph":"https://pith.science/api/pith-number/SSLOKSFKR5M34VC37UN5IKFGV7/graph.json","fetch_events":"https://pith.science/api/pith-number/SSLOKSFKR5M34VC37UN5IKFGV7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7/action/storage_attestation","attest_author":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7/action/author_attestation","sign_citation":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7/action/citation_signature","submit_replication":"https://pith.science/pith/SSLOKSFKR5M34VC37UN5IKFGV7/action/replication_record"}},"created_at":"2026-07-05T10:14:01.609345+00:00","updated_at":"2026-07-05T10:14:01.609345+00:00"}