{"id":"W4417467340","doi":"10.1159/000550153","title":"When Jack of All Trades Is a Master of None: Comparing the Performance of GPT-4 Omni against Specialised Neural Networks in Identifying Malignant Dermatological Lesions from Smartphone Images and Structured Clinical Data","year":2025,"lang":"en","type":"article","venue":"Dermatology","topic":"Cutaneous Melanoma Detection and Management","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary; Queen's University; University of British Columbia; Public Health Ontario; University of Toronto","funders":"","keywords":"Skin lesion; Artificial neural network; Patient data; Dermatological diseases; Triage","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":[0.0001794131,0.000137785,0.0007567596,0.0001384032,0.00003786368,0.0000102859,0.0002753107,0.0001383313,0.00004580315],"category_scores_gemma":[0.0000572479,0.00009839686,0.00007044835,0.0001415361,0.0003552736,0.00005778124,0.0003926758,0.0002641075,6.488978e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009560684,"about_ca_system_score_gemma":0.00002516734,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009435449,"about_ca_topic_score_gemma":0.0001525272,"domain_scores_codex":[0.9982773,0.0001936135,0.0009194527,0.0002910748,0.0001259826,0.0001925851],"domain_scores_gemma":[0.998831,0.000296925,0.0002661939,0.000525289,0.00003615565,0.00004437526],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0007684402,0.0001934356,0.982885,0.0004035847,0.000292932,0.0002089581,0.000696554,0.0000893516,0.004188662,0.00009733641,0.004117999,0.006057731],"study_design_scores_gemma":[0.003318331,0.00006885684,0.8285773,0.0003194105,0.000280792,0.0004359662,0.0007985306,0.1567852,0.008215612,0.000081435,0.0009877561,0.0001308207],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9961757,0.000712089,0.0005938095,0.001633758,0.0002973539,0.0003390615,0.00002528669,0.00001294416,0.0002099716],"genre_scores_gemma":[0.99794,0.0004924938,0.0005441334,0.0008999247,0.00002852422,0.000004671881,0.00006021538,0.00000744388,0.00002257213],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1566958,"threshold_uncertainty_score":0.4012507,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08904350149265933,"score_gpt":0.3314629814763169,"score_spread":0.2424194799836576,"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."}}