{"id":"W4399544575","doi":"10.1038/s41467-024-48666-7","title":"Mapping the landscape of histomorphological cancer phenotypes using self-supervised learning on unannotated pathology slides","year":2024,"lang":"en","type":"article","venue":"Nature Communications","topic":"AI in cancer detection","field":"Computer Science","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Engineering and Physical Sciences Research Council; National Cancer Institute; School of Medicine, New York University; Vetenskapsrådet; National Institutes of Health; Biotechnology and Biological Sciences Research Council; University of Glasgow; Research Councils UK; York University","keywords":"Phenotype; Digital pathology; Computational biology; Pathology; Artificial intelligence; Biology; Immunophenotyping; Computer science; Phenotypic screening; Medicine; Gene; Genetics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000388498,0.0001087958,0.0001354419,0.00015142,0.0003741403,0.0000699751,0.001543333,0.0001910396,0.00001446964],"category_scores_gemma":[0.00009587956,0.00007754084,0.00006401713,0.0009922347,0.0001144228,0.0001390602,0.0004411096,0.001262719,0.000009803353],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001442758,"about_ca_system_score_gemma":0.0001205419,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004155678,"about_ca_topic_score_gemma":0.00005207193,"domain_scores_codex":[0.9988703,0.0003619298,0.0002042261,0.0002460291,0.0001646655,0.0001528876],"domain_scores_gemma":[0.9979743,0.0006456241,0.00008994648,0.001130663,0.0001353838,0.00002413346],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007966458,0.0006779527,0.01437757,0.0003356416,0.0007690006,0.00009012979,0.0459706,0.0346909,0.1279994,0.3450885,0.01236611,0.4175545],"study_design_scores_gemma":[0.0001886033,0.0001172689,0.005474342,0.0001883622,0.00003816882,0.00006484234,0.0002837084,0.924949,0.001463397,0.001298541,0.06570829,0.0002254271],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6479881,0.1495836,0.1371787,0.04627941,0.004503948,0.001171925,0.00003913477,0.003042183,0.01021299],"genre_scores_gemma":[0.9731123,0.0008747348,0.0255196,0.0003522335,0.00005096058,0.00003588155,0.000005932054,0.00001137538,0.00003700373],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8902581,"threshold_uncertainty_score":0.5485954,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04570803936177387,"score_gpt":0.3089186293740342,"score_spread":0.2632105900122604,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}