{"id":"W3113296611","doi":"10.2196/22397","title":"Using General-purpose Sentiment Lexicons for Suicide Risk Assessment in Electronic Health Records: Corpus-Based Analysis","year":2020,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Mental Health via Writing","field":"Psychology","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Academy of Medical Sciences; Medical Research Council; National Institute for Health and Care Research; UK Research and Innovation","keywords":"Health records; Natural language processing; Computer science; Electronic health record; Artificial intelligence; Medical emergency; Medicine; Health care","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":[0.004184866,0.0006475531,0.000432515,0.007755162,0.0009177366,0.001778079,0.0004478925,0.0004438952,0.001328034],"category_scores_gemma":[0.02209321,0.0002608021,0.0007961069,0.005275368,0.0006808683,0.001903865,0.0013694,0.0006049414,0.0005929789],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001157388,"about_ca_system_score_gemma":0.001195679,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004644821,"about_ca_topic_score_gemma":0.007510698,"domain_scores_codex":[0.9968954,0.001333757,0.0005970241,0.0004916467,0.0005885877,0.00009353932],"domain_scores_gemma":[0.9803924,0.01295265,0.001776429,0.0009575093,0.003701732,0.0002193221],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001006669,0.0006947558,0.2490224,0.008054464,0.0007612505,0.002201067,0.01702089,0.01102352,0.07824343,0.01006745,0.04912838,0.5727758],"study_design_scores_gemma":[0.000328937,0.0006257636,0.4423582,0.001957062,0.001054675,0.003315828,0.02037891,0.3481953,0.03940928,0.01670794,0.1253279,0.0003400268],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8653719,0.002781201,0.09228323,0.001049396,0.0003136978,0.001677464,0.02544055,0.001459283,0.009623162],"genre_scores_gemma":[0.7628417,0.001116336,0.1890832,0.0002319688,0.0001257227,0.001742004,0.04343731,0.0002546846,0.001167016],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.007755162,"threshold_uncertainty_score":0.02213198,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07971806615987065,"score_gpt":0.4636556540803959,"score_spread":0.3839375879205252,"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."}}