{"id":"W3010499704","doi":"10.1101/2020.02.26.967208","title":"Automated detection of glaucoma with interpretable machine learning using clinical data and multi-modal retinal images","year":2020,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Retinal Imaging and Analysis","field":"Medicine","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Eye Institute; Alberta Innovates - Technology Futures; Lowy Medical Research Institute; Alfred P. Sloan Foundation; University of Washington; Research to Prevent Blindness; Gordon and Betty Moore Foundation; National Institutes of Health; National Science Foundation","keywords":"Glaucoma; Optical coherence tomography; Artificial intelligence; Computer science; Data set; Optic disc; Modal; Fundus (uterus); Intraocular pressure; Medical diagnosis; Segmentation; Population; Machine learning; Pattern recognition (psychology); Ophthalmology; Medicine; Radiology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"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":[0.001592201,0.0006246959,0.0004561926,0.0008776372,0.0001863069,0.00100334,0.0005029313,0.0008673258,0.0009219851],"category_scores_gemma":[0.005719509,0.0003008153,0.0006892899,0.0003283479,0.0003822578,0.000441772,0.000484706,0.0007498293,0.0003240983],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001096264,"about_ca_system_score_gemma":0.0006664491,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009333841,"about_ca_topic_score_gemma":0.007247639,"domain_scores_codex":[0.9995264,0.000198361,0.00002809877,0.0001419113,0.0000627814,0.00004245544],"domain_scores_gemma":[0.9971994,0.001914451,0.0003166947,0.0002112964,0.0002809131,0.00007721972],"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":[0.0002360239,0.0002938648,0.0608211,0.00005650005,0.0001616433,0.0002392903,0.0000899568,0.8861826,0.003434769,0.0006474252,0.001205319,0.04663148],"study_design_scores_gemma":[0.000003518501,0.00002039579,0.00320884,0.000004318129,0.000005945848,0.00002129314,0.000005786914,0.9957979,0.000365981,0.0005013555,0.00005966817,0.000004886408],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7775762,0.0005490744,0.2157829,0.00148322,0.00008264776,0.0001180587,0.001325303,0.001498228,0.001584295],"genre_scores_gemma":[0.9857053,0.00004308626,0.01333678,0.00009537349,0.00002296904,0.00003257815,0.0003608941,0.00001440714,0.0003887499],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.009333841,"threshold_uncertainty_score":0.01855904,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03880419940726593,"score_gpt":0.311283024103766,"score_spread":0.2724788246965001,"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."}}