{"id":"W4320065330","doi":"10.2196/43014","title":"Extracting Medical Information From Free-Text and Unstructured Patient-Generated Health Data Using Natural Language Processing Methods: Feasibility Study With Real-world Data","year":2023,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":59,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Center for Advancing Translational Sciences; National Institutes of Health","keywords":"Computer science; Natural language processing; Pipeline (software); Artificial intelligence; Named-entity recognition; Unstructured data; Sentence; SNOMED CT; Parsing; Set (abstract data type); Unified Medical Language System; Information retrieval; Big data; Terminology; Data mining; Linguistics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01171034,0.0006291649,0.0004940317,0.001374444,0.000845836,0.001474745,0.001378667,0.001217924,0.001990042],"category_scores_gemma":[0.03570452,0.0003228631,0.000713094,0.0009830494,0.001218297,0.002660828,0.001533158,0.0007026089,0.000800339],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006084887,"about_ca_system_score_gemma":0.001916532,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002336836,"about_ca_topic_score_gemma":0.003286368,"domain_scores_codex":[0.9910098,0.005681351,0.0008020571,0.001256817,0.0009737477,0.0002762812],"domain_scores_gemma":[0.9436969,0.04805673,0.001322197,0.002073992,0.003980057,0.0008701236],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.008130769,0.03278922,0.2931685,0.008955165,0.0007036383,0.00986673,0.04542106,0.02548062,0.0624864,0.003701556,0.01020618,0.4990903],"study_design_scores_gemma":[0.004801752,0.03466174,0.3946495,0.00124199,0.001042052,0.01132387,0.05241669,0.35524,0.08726319,0.01037255,0.0461516,0.0008351242],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9307455,0.0001791682,0.0587579,0.0006216168,0.00004447357,0.005681312,0.002473265,0.0003839942,0.001112753],"genre_scores_gemma":[0.7846575,0.000199646,0.2051447,0.0003521797,0.00007063853,0.005057845,0.003983335,0.00005980538,0.0004743116],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01171034,"threshold_uncertainty_score":0.06193089,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2386185537144399,"score_gpt":0.5579463831016811,"score_spread":0.3193278293872412,"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."}}