{"id":"W4405068114","doi":"10.2196/59045","title":"Intersection of Performance, Interpretability, and Fairness in Neural Prototype Tree for Chest X-Ray Pathology Detection: Algorithm Development and Validation Study","year":2024,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"COVID-19 diagnosis using AI","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Health Sciences Centre; Sunnybrook Health Science Centre; St. Michael's Hospital; University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; University of Toronto","keywords":"Interpretability; Artificial intelligence; Classifier (UML); Receiver operating characteristic; Machine learning; Decision tree; Computer science; Gradient boosting; Deep learning; Artificial neural network; Pattern recognition (psychology); Random forest","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02025886,0.001170679,0.001147121,0.001750901,0.0005892625,0.001717153,0.001604928,0.001750053,0.001142994],"category_scores_gemma":[0.055984,0.0003275837,0.0009379386,0.0008178677,0.0008866742,0.001504101,0.001445557,0.001815116,0.0003068065],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002338963,"about_ca_system_score_gemma":0.00225406,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006144405,"about_ca_topic_score_gemma":0.003872846,"domain_scores_codex":[0.993777,0.003095096,0.0005113493,0.0008888331,0.001397056,0.000330635],"domain_scores_gemma":[0.9561877,0.03125966,0.001859675,0.002373268,0.007742283,0.0005773823],"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":[0.002416625,0.001223079,0.1018428,0.0004785291,0.000633254,0.000290654,0.000439973,0.4596112,0.006373241,0.00327591,0.005542652,0.4178721],"study_design_scores_gemma":[0.00005614117,0.0003983114,0.005732215,0.00004023865,0.00007488869,0.00009353386,0.00005472593,0.9886997,0.003392072,0.001004893,0.0004347595,0.00001842597],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7924328,0.003217279,0.1975355,0.0007379008,0.0002128811,0.0006366316,0.0005660111,0.001435783,0.003225113],"genre_scores_gemma":[0.9246553,0.0003147046,0.07285663,0.0001274396,0.00004994471,0.0002780689,0.0009397718,0.00008815803,0.0006899153],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02025886,"threshold_uncertainty_score":0.1071404,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06756871581205094,"score_gpt":0.4246695908152568,"score_spread":0.3571008750032059,"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."}}