{"id":"W4284978235","doi":"10.1016/j.cmpb.2022.107012","title":"Constrained multiple instance learning for ulcerative colitis prediction using histological images","year":2022,"lang":"en","type":"article","venue":"Computer Methods and Programs in Biomedicine","topic":"AI in cancer detection","field":"Computer Science","cited_by":34,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"European Regional Development Fund; Generalitat Valenciana; Ministerio de Economía y Competitividad; European Commission","keywords":"Ulcerative colitis; Artificial intelligence; Inflammatory bowel disease; Colitis; Rectum; Computer science; Focus (optics); Relevance (law); Medicine; Machine learning; Pattern recognition (psychology); Disease; Pathology; Gastroenterology","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":[0.001881068,0.0001531972,0.0002921586,0.0002025756,0.0004079879,0.00006562546,0.0002629534,0.00005795139,0.000005746771],"category_scores_gemma":[0.00007301631,0.0001435769,0.00004355003,0.0006606096,0.0002529794,0.000179097,0.0003219901,0.0003225132,8.282348e-8],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002407471,"about_ca_system_score_gemma":0.00005753681,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000448811,"about_ca_topic_score_gemma":0.000005298901,"domain_scores_codex":[0.997959,0.0006852014,0.000340569,0.0005344665,0.0001873623,0.0002934563],"domain_scores_gemma":[0.9990281,0.0004850806,0.0001587079,0.0001760168,0.00007575188,0.00007638035],"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":[0.00007886549,0.00006030271,0.002532384,0.00003801562,0.00001678762,0.00001193587,0.001106798,0.0005880207,0.001292778,0.0007587083,0.00006645342,0.993449],"study_design_scores_gemma":[0.001836342,0.002069641,0.004877545,0.00005071006,0.00001133568,0.0001837956,0.0002306702,0.9533797,0.0002948986,0.00276585,0.03411449,0.0001849991],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01433418,0.0005999131,0.9823517,0.0004515419,0.001341167,0.0007067044,0.000006583235,0.0001769759,0.00003119727],"genre_scores_gemma":[0.2300653,0.00002205975,0.7693059,0.0001784692,0.0001727678,0.000217412,0.00001421661,0.000008324485,0.00001549032],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.993264,"threshold_uncertainty_score":0.5854897,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08131422063713314,"score_gpt":0.361383803857829,"score_spread":0.2800695832206959,"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."}}