{"id":"W4403064636","doi":"10.1186/s12911-024-02677-y","title":"Validation of large language models for detecting pathologic complete response in breast cancer using population-based pathology reports","year":2024,"lang":"en","type":"article","venue":"BMC Medical Informatics and Decision Making","topic":"AI in cancer detection","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary; Alberta Health Services","funders":"Canadian Institutes of Health Research","keywords":"Health informatics; Breast cancer; Medicine; Pathology; Population; Cancer; Computer science; Natural language processing; Internal medicine; Public health","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01025136,0.001183495,0.0005144964,0.001223346,0.000279694,0.00102518,0.001063319,0.0008288836,0.001116719],"category_scores_gemma":[0.02882116,0.0003222469,0.001266369,0.0005092621,0.0005632,0.001031155,0.001449853,0.001360399,0.0008228175],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00105387,"about_ca_system_score_gemma":0.001202551,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005347109,"about_ca_topic_score_gemma":0.004669883,"domain_scores_codex":[0.9966725,0.002008938,0.0002634618,0.0006795594,0.0002425775,0.0001329661],"domain_scores_gemma":[0.9832513,0.01377835,0.00073364,0.0009889826,0.001005768,0.0002418753],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"observational","study_design_scores_codex":[0.002863778,0.00135934,0.3010825,0.0005867372,0.001505879,0.0007029923,0.001059666,0.3892806,0.01172104,0.001554224,0.006879851,0.2814035],"study_design_scores_gemma":[0.00009283202,0.0004929423,0.02419507,0.0000660136,0.0001604445,0.0002570356,0.0002602027,0.966758,0.005048976,0.001703502,0.0009208838,0.00004418944],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9309388,0.000686136,0.06170249,0.0007061493,0.0001311385,0.0002289861,0.002586374,0.001863249,0.001156626],"genre_scores_gemma":[0.9758715,0.0001120767,0.01917593,0.0001464727,0.00003573047,0.0001060376,0.004136642,0.00005122906,0.000364486],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01025136,"threshold_uncertainty_score":0.05421501,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05551034995085824,"score_gpt":0.3693355669972379,"score_spread":0.3138252170463796,"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."}}