{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:RCJI5PC6L6N5D4LBHELS54255W","short_pith_number":"pith:RCJI5PC6","schema_version":"1.0","canonical_sha256":"88928ebc5e5f9bd1f16139172ef35ded9e699b05cecbdf0fe25ac6020ba61701","source":{"kind":"arxiv","id":"2403.09920","version":3},"attestation_state":"computed","paper":{"title":"Predicting Generalization of AI Colonoscopy Models to Unseen Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.CY"],"primary_cat":"eess.IV","authors_text":"Atsushi Hamabe, Carson McNeil, Daisuke Tsurumaru, Ehud Rivlin, Eiji Oki, Haruei Ogino, Hiroki Kayama, Hiro-o Yamano, Hiroshi Nakase, Ichiro Takemasa, Joel Shor, Joseph R Ledsam, Kaho Kobayashi, Koji Ando, Masaaki Miyo, Mitsuhiko Ota, Roman Goldenberg, Yotam Intrator","submitted_at":"2024-03-14T23:41:00Z","abstract_excerpt":"$\\textbf{Background}$: Generalizability of AI colonoscopy algorithms is important for wider adoption in clinical practice. However, current techniques for evaluating performance on unseen data require expensive and time-intensive labels.\n  $\\textbf{Methods}$: We use a \"Masked Siamese Network\" (MSN) to identify novel phenomena in unseen data and predict polyp detector performance. MSN is trained to predict masked out regions of polyp images, without any labels. We test MSN's ability to be trained on data only from Israel and detect unseen techniques, narrow-band imaging (NBI) and chromendoscoy "},"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":"2403.09920","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.IV","submitted_at":"2024-03-14T23:41:00Z","cross_cats_sorted":["cs.AI","cs.CV","cs.CY"],"title_canon_sha256":"cdc435b95b30446e67e32ead8d891c2b9fd0b2ba90165472c99282a6eedd506a","abstract_canon_sha256":"e08942fdec20100426805cca790a26e8661059572e6751db309ad0437c8babd9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:59:22.653483Z","signature_b64":"OhHNVNh7/5Iy/XCu7rYsY9F1QHKc19lpOC22CjmYZVXB5aR1pL5fono/o9Aq2PBQj2DY5nLPvcVXPpIfxv0UBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"88928ebc5e5f9bd1f16139172ef35ded9e699b05cecbdf0fe25ac6020ba61701","last_reissued_at":"2026-07-05T07:59:22.653013Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:59:22.653013Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Predicting Generalization of AI Colonoscopy Models to Unseen Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.CY"],"primary_cat":"eess.IV","authors_text":"Atsushi Hamabe, Carson McNeil, Daisuke Tsurumaru, Ehud Rivlin, Eiji Oki, Haruei Ogino, Hiroki Kayama, Hiro-o Yamano, Hiroshi Nakase, Ichiro Takemasa, Joel Shor, Joseph R Ledsam, Kaho Kobayashi, Koji Ando, Masaaki Miyo, Mitsuhiko Ota, Roman Goldenberg, Yotam Intrator","submitted_at":"2024-03-14T23:41:00Z","abstract_excerpt":"$\\textbf{Background}$: Generalizability of AI colonoscopy algorithms is important for wider adoption in clinical practice. However, current techniques for evaluating performance on unseen data require expensive and time-intensive labels.\n  $\\textbf{Methods}$: We use a \"Masked Siamese Network\" (MSN) to identify novel phenomena in unseen data and predict polyp detector performance. MSN is trained to predict masked out regions of polyp images, without any labels. We test MSN's ability to be trained on data only from Israel and detect unseen techniques, narrow-band imaging (NBI) and chromendoscoy "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.09920","kind":"arxiv","version":3},"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/2403.09920/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":"2403.09920","created_at":"2026-07-05T07:59:22.653073+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.09920v3","created_at":"2026-07-05T07:59:22.653073+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.09920","created_at":"2026-07-05T07:59:22.653073+00:00"},{"alias_kind":"pith_short_12","alias_value":"RCJI5PC6L6N5","created_at":"2026-07-05T07:59:22.653073+00:00"},{"alias_kind":"pith_short_16","alias_value":"RCJI5PC6L6N5D4LB","created_at":"2026-07-05T07:59:22.653073+00:00"},{"alias_kind":"pith_short_8","alias_value":"RCJI5PC6","created_at":"2026-07-05T07:59:22.653073+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/RCJI5PC6L6N5D4LBHELS54255W","json":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W.json","graph_json":"https://pith.science/api/pith-number/RCJI5PC6L6N5D4LBHELS54255W/graph.json","events_json":"https://pith.science/api/pith-number/RCJI5PC6L6N5D4LBHELS54255W/events.json","paper":"https://pith.science/paper/RCJI5PC6"},"agent_actions":{"view_html":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W","download_json":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W.json","view_paper":"https://pith.science/paper/RCJI5PC6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.09920&json=true","fetch_graph":"https://pith.science/api/pith-number/RCJI5PC6L6N5D4LBHELS54255W/graph.json","fetch_events":"https://pith.science/api/pith-number/RCJI5PC6L6N5D4LBHELS54255W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W/action/storage_attestation","attest_author":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W/action/author_attestation","sign_citation":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W/action/citation_signature","submit_replication":"https://pith.science/pith/RCJI5PC6L6N5D4LBHELS54255W/action/replication_record"}},"created_at":"2026-07-05T07:59:22.653073+00:00","updated_at":"2026-07-05T07:59:22.653073+00:00"}