{"id":"W2968097744","doi":"10.1109/jbhi.2019.2934477","title":"Direct Cup-to-Disc Ratio Estimation for Glaucoma Screening via Semi-Supervised Learning","year":2019,"lang":"en","type":"article","venue":"IEEE Journal of Biomedical and Health Informatics","topic":"Retinal Imaging and Analysis","field":"Medicine","cited_by":86,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"National Key Research and Development Program of China Stem Cell and Translational Research; National Natural Science Foundation of China","keywords":"Glaucoma; Optic disc; Optic cup (embryology); Artificial intelligence; Computer science; Segmentation; Convolutional neural network; Optic disk; Pattern recognition (psychology); Robustness (evolution); Feature extraction; Deep learning; Supervised learning; Feature (linguistics); Artificial neural network; Ophthalmology; Medicine","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001660751,0.0001164195,0.0005355818,0.0003246296,0.0001400445,0.0000445918,0.00006982211,0.00006603281,0.00002330378],"category_scores_gemma":[0.0001794551,0.00008095361,0.0001235344,0.0002493314,0.00004289228,0.0002186022,0.0000151183,0.0003143888,0.00001383536],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005368535,"about_ca_system_score_gemma":0.0001930408,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001690506,"about_ca_topic_score_gemma":6.690901e-7,"domain_scores_codex":[0.9980295,0.00003359437,0.001148616,0.00006267503,0.0004692009,0.0002563841],"domain_scores_gemma":[0.9984761,0.0001269245,0.0005604623,0.00008431161,0.0002078902,0.0005443062],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007293068,0.0002413993,0.04166677,0.005285914,0.0004025365,0.00001538322,0.01019699,0.001267571,0.002702581,0.00003115971,0.01549126,0.9219691],"study_design_scores_gemma":[0.003064951,0.002960327,0.007189284,0.00162411,0.0001534457,0.0003878916,0.001290359,0.9570652,0.0003948357,0.00004257565,0.02566719,0.0001597785],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3877909,0.0001703821,0.5996112,0.0117648,0.0002111504,0.0002746696,0.000003284173,0.00001942588,0.0001542658],"genre_scores_gemma":[0.8763421,0.000236869,0.1183937,0.004208637,0.0004236955,0.000003631204,0.00003163,0.00001557716,0.0003441437],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9557977,"threshold_uncertainty_score":0.3301192,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02525178807350699,"score_gpt":0.3356276650895234,"score_spread":0.3103758770160164,"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."}}