{"id":"W4405923428","doi":"10.1038/s41598-024-81724-0","title":"Attention-guided convolutional network for bias-mitigated and interpretable oral lesion classification","year":2024,"lang":"en","type":"article","venue":"Scientific Reports","topic":"AI in cancer detection","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University Health Centre; McGill University","funders":"Canadian Institutes of Health Research","keywords":"Interpretability; Artificial intelligence; Computer science; Ground truth; Deep learning; Segmentation; Robustness (evolution); Pattern recognition (psychology); Lesion; Categorization; Convolutional neural network; Machine learning; Medicine; Pathology","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"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":[0.001424293,0.001077358,0.0006480798,0.0007208703,0.0003119147,0.0007133139,0.001397733,0.0009615207,0.00140061],"category_scores_gemma":[0.003325659,0.000343204,0.0006770567,0.0005629655,0.0004638934,0.0008849815,0.001044621,0.001518365,0.0004936549],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001647645,"about_ca_system_score_gemma":0.00146947,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01571411,"about_ca_topic_score_gemma":0.02188911,"domain_scores_codex":[0.9995773,0.00007968472,0.00002299089,0.0001391243,0.00009671349,0.00008404562],"domain_scores_gemma":[0.9989777,0.0004492007,0.000119494,0.0001489462,0.0002590152,0.00004578959],"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":[0.0007456077,0.000306794,0.01235943,0.0001584911,0.0002655447,0.0003501823,0.0001721004,0.3532957,0.03562404,0.002922808,0.009444206,0.5843551],"study_design_scores_gemma":[0.00001175107,0.0000515018,0.00141445,0.000009685479,0.00003824626,0.00005228725,0.000009110508,0.9896277,0.006454391,0.001741255,0.0005813469,0.000008303499],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3860722,0.002283734,0.5953531,0.001441508,0.0002869664,0.0001748852,0.001516229,0.007636494,0.005234907],"genre_scores_gemma":[0.9336807,0.0003047255,0.05981459,0.000315306,0.00007075124,0.00007055052,0.001448561,0.0001133315,0.004181434],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01571411,"threshold_uncertainty_score":0.03124529,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07339478128583878,"score_gpt":0.3112205826537794,"score_spread":0.2378258013679406,"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."}}