{"id":"W6977259263","doi":"10.6084/m9.figshare.19759386","title":"Additional file 4 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":[0.001671441,0.001065869,0.0007905045,0.003453514,0.0005482053,0.001469406,0.001532664,0.0009276581,0.68614],"category_scores_gemma":[0.01946769,0.0004552328,0.0009046023,0.003643862,0.0002460174,0.001359842,0.00127607,0.0007297028,0.116945],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001078075,"about_ca_system_score_gemma":0.002018982,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004001167,"about_ca_topic_score_gemma":0.006318447,"domain_scores_codex":[0.9990619,0.0001455882,0.0002459843,0.0002416815,0.0002177495,0.00008726944],"domain_scores_gemma":[0.9815381,0.01295621,0.00133826,0.001044091,0.002718882,0.0004045137],"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":[0.0003986916,0.00009766706,0.002883006,0.002525373,0.00006484831,0.0001390011,0.00007495566,0.0004943216,0.0005803662,0.00073143,0.9734237,0.01858666],"study_design_scores_gemma":[0.002386896,0.0002200927,0.0254121,0.001498773,0.000227756,0.0006717266,0.0003501553,0.00458695,0.00558989,0.008969552,0.9499213,0.0001647981],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.0002842426,0.0000156888,0.000916143,0.000115992,0.00002399245,0.00009330971,0.9963185,0.001391111,0.0008410373],"genre_scores_gemma":[0.006613003,0.00008620294,0.009743799,0.0003210041,0.00009925096,0.001193479,0.9753749,0.001258335,0.005309931],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.68614,"threshold_uncertainty_score":0.4476832,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06701088399541694,"score_gpt":0.4057983267847717,"score_spread":0.3387874427893548,"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."}}