{"id":"W6976877773","doi":"10.6084/m9.figshare.19759380.v1","title":"Additional file 2 of Automated medical chart review for breast cancer outcomes research: a novel natural language processing extraction system","year":2022,"lang":"en","type":"article","venue":"Figshare","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Prevention of Organ Failure; University of British Columbia","funders":"","keywords":"Breast cancer; Chart; Data extraction; Natural language; Gold standard (test)","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00159674,0.001012996,0.000758618,0.003278639,0.000535015,0.001457844,0.001436189,0.0008755269,0.6593007],"category_scores_gemma":[0.01875455,0.0004414732,0.0008404264,0.003564793,0.0002382989,0.001312074,0.00125301,0.0007234714,0.1137051],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009827464,"about_ca_system_score_gemma":0.001942426,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003678074,"about_ca_topic_score_gemma":0.005926862,"domain_scores_codex":[0.9991112,0.0001337827,0.0002247888,0.0002436623,0.0002080036,0.00007865346],"domain_scores_gemma":[0.9826717,0.01222746,0.001228776,0.0009833799,0.002499627,0.0003892107],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003630374,0.00009297599,0.002727196,0.002342417,0.00005830915,0.0001381781,0.00007568309,0.0004647164,0.0005999813,0.0007379249,0.9737857,0.01861394],"study_design_scores_gemma":[0.00211584,0.0002099977,0.02472213,0.001353404,0.0002107716,0.0007118265,0.0003476576,0.00457367,0.005519207,0.008421266,0.9516581,0.0001560624],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0002811582,0.00001472551,0.0009532653,0.0001088396,0.00002261984,0.00008766399,0.9964204,0.001310295,0.0008011109],"genre_scores_gemma":[0.006328249,0.0000789042,0.009770464,0.0003094668,0.00009348639,0.00115888,0.9761199,0.001218633,0.004921907],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.6593007,"threshold_uncertainty_score":0.4859662,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06545233424165164,"score_gpt":0.4051565814282834,"score_spread":0.3397042471866318,"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."}}