{"id":"W3180779372","doi":"10.1038/s43856-021-00008-0","title":"A BERT model generates diagnostically relevant semantic embeddings from pathology synopses with active learning","year":2021,"lang":"en","type":"article","venue":"Communications Medicine","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"Juravinski Hospital; University of Waterloo; McMaster University","funders":"","keywords":"Computer science; Artificial intelligence; Deep learning; Natural language processing; Information extraction; Context (archaeology); Semantics (computer science); Training set; Information retrieval","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":[],"consensus_categories":[],"category_scores_codex":[0.0009626536,0.001051927,0.0003797654,0.0008357192,0.0002445499,0.0007987326,0.001359761,0.001127882,0.002128233],"category_scores_gemma":[0.003609661,0.0003432754,0.00101887,0.0004508523,0.000708553,0.001357842,0.001007794,0.001584177,0.0008360788],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001131829,"about_ca_system_score_gemma":0.001240869,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002950128,"about_ca_topic_score_gemma":0.003760303,"domain_scores_codex":[0.9996386,0.00007908469,0.00002485799,0.0001449569,0.00006655986,0.00004581888],"domain_scores_gemma":[0.9988281,0.000678115,0.00009734956,0.0001044804,0.00023335,0.00005854249],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0006669816,0.0005758875,0.009202477,0.0002792875,0.0001845684,0.000469961,0.0003720439,0.5170787,0.01795992,0.01574107,0.01194043,0.4255287],"study_design_scores_gemma":[0.00001199537,0.00003554305,0.0001586319,0.000008449592,0.00001311725,0.00003714514,0.00001361324,0.9918588,0.002442352,0.004974655,0.0004407699,0.000004871889],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09280151,0.0002898674,0.8969671,0.001365267,0.0001482014,0.0002129062,0.001038506,0.003916754,0.003259942],"genre_scores_gemma":[0.8303912,0.0001811426,0.1615212,0.0006162,0.00007125915,0.0002646627,0.002264597,0.0001406148,0.004549003],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002950128,"threshold_uncertainty_score":0.00821203,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04592017625887133,"score_gpt":0.3451619568347017,"score_spread":0.2992417805758303,"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."}}