{"id":"W2909847169","doi":"10.21037/jmai.2019.01.01","title":"Promises and limitations of deep learning for medical image segmentation","year":2019,"lang":"en","type":"article","venue":"Journal of Medical Artificial Intelligence","topic":"Radiomics and Machine Learning in Medical Imaging","field":"Medicine","cited_by":45,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Artificial intelligence; Deep learning; Computer vision; Segmentation; Image (mathematics); Computer science","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.002409445,0.00009868558,0.0003879659,0.0001515356,0.00005359038,0.00002441674,0.0001699394,0.0001363847,0.0005284449],"category_scores_gemma":[0.01854407,0.00007399959,0.000127174,0.0001445549,0.0002600372,0.0001166643,0.00004044278,0.0007045718,0.000009816397],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002623638,"about_ca_system_score_gemma":0.0002618643,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00000878825,"about_ca_topic_score_gemma":0.000004836406,"domain_scores_codex":[0.9973783,0.00008200524,0.0008925025,0.0001383656,0.001332419,0.0001764131],"domain_scores_gemma":[0.9971366,0.001628051,0.0003853249,0.00007666478,0.0003500252,0.0004233174],"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":[0.0003009683,0.0002353085,0.002691327,0.0002519012,0.00008520801,0.00005150995,0.0009134556,0.0001259567,0.01203759,0.003493119,0.0001157906,0.9796979],"study_design_scores_gemma":[0.0020072,0.004531862,0.003989395,0.002434116,0.0004412397,0.001721331,0.00726337,0.9093662,0.04524851,0.01281531,0.009807581,0.0003738839],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5838106,0.000516384,0.3961141,0.01837559,0.0004751456,0.0003008133,4.918653e-7,0.00001230145,0.0003946566],"genre_scores_gemma":[0.9815713,0.001270451,0.01620914,0.0004710897,0.0003967542,0.000004480082,0.000004313863,0.00001603632,0.00005643801],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.979324,"threshold_uncertainty_score":0.9897231,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04344906224562343,"score_gpt":0.3625607305346126,"score_spread":0.3191116682889892,"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."}}