{"id":"W4390837349","doi":"10.1101/2024.01.12.24301222","title":"Assessing Performance of Multimodal ChatGPT-4 on an image based Radiology Board-style Examination: An exploratory study","year":2024,"lang":"en","type":"preprint","venue":"medRxiv","topic":"Artificial Intelligence in Healthcare and Education","field":"Medicine","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Subspecialty; Radiology; Likert scale; Medicine; Medical physics; Confidence interval; Institutional review board; Multiple choice; Modalities; Significant difference; Psychology; Pathology; Surgery","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008276668,0.0007439435,0.0006065233,0.001402467,0.000413942,0.000815101,0.0006392347,0.0007919552,0.001801999],"category_scores_gemma":[0.03588032,0.0004224624,0.0006784623,0.0005426327,0.0004911008,0.001238022,0.001747125,0.0004932473,0.001210532],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005485059,"about_ca_system_score_gemma":0.0004329973,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008522068,"about_ca_topic_score_gemma":0.001324654,"domain_scores_codex":[0.9945931,0.002739188,0.0005875078,0.0006944619,0.0009825035,0.0004032479],"domain_scores_gemma":[0.9696296,0.01398636,0.007369896,0.002333967,0.004477263,0.002202861],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.004276376,0.002445075,0.9583638,0.0001257772,0.0001539039,0.0005595341,0.005259055,0.0005052835,0.004334905,0.00003480356,0.0003712175,0.02357026],"study_design_scores_gemma":[0.00008790018,0.01372096,0.9794891,0.00002895888,0.00004907545,0.0007185409,0.001746312,0.001437057,0.002194645,0.00004137296,0.0004499667,0.00003604819],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.999332,0.00001938923,0.0001965077,0.000009171202,0.0000032578,0.00007396942,0.00005134082,0.000009662711,0.000304773],"genre_scores_gemma":[0.9985803,0.00002813841,0.0007325928,0.00002739864,0.00001546986,0.0001482163,0.0001654548,0.000007924602,0.0002944986],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008276668,"threshold_uncertainty_score":0.04377174,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1557306516319711,"score_gpt":0.4376000450642064,"score_spread":0.2818693934322353,"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."}}