{"id":"W4391654261","doi":"10.2196/32690","title":"Vision-Language Model for Generating Textual Descriptions From Clinical Images: Model Development and Validation Study","year":2024,"lang":"en","type":"article","venue":"JMIR Formative Research","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":37,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"China Postdoctoral Science Foundation; National Natural Science Foundation of China","keywords":"Computer science; Natural language processing; Artificial intelligence; Standardization; Deep learning; Metric (unit); Closed captioning; Quality (philosophy); Information retrieval; Machine learning; Image (mathematics)","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002750237,0.0001424011,0.0001674769,0.0002825971,0.0007230772,0.0008363012,0.000548852,0.00006988031,0.000003707661],"category_scores_gemma":[0.0001748805,0.000121838,0.00005087475,0.0004231843,0.00008604677,0.001008933,0.000608271,0.0005931453,0.0000709361],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001220946,"about_ca_system_score_gemma":0.0003183788,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006782267,"about_ca_topic_score_gemma":0.00001664365,"domain_scores_codex":[0.9976703,0.0003203292,0.0005028126,0.0005741948,0.0005811969,0.0003512],"domain_scores_gemma":[0.998152,0.0009531198,0.00005451189,0.0004291924,0.0002687802,0.0001423507],"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":[0.00003477842,0.001425796,0.0008119273,0.0001363294,0.0001588658,0.000006744279,0.2358743,0.08863649,0.005326156,0.02244325,0.006265263,0.6388801],"study_design_scores_gemma":[0.0003421786,0.0001445878,0.002677695,0.00004016866,0.000004040339,8.308344e-7,0.001541893,0.9927561,0.0002647081,0.00201869,0.00007329626,0.0001358617],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4243904,0.0000529406,0.574213,0.0002970564,0.00002862509,0.0007330449,0.0000257558,0.0001187957,0.0001403486],"genre_scores_gemma":[0.7458573,0.000003985958,0.2528514,0.00003214442,0.00005281886,0.0009561817,0.00004451102,0.00001397339,0.0001876384],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9041196,"threshold_uncertainty_score":0.8064472,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1517744827311691,"score_gpt":0.505880174163475,"score_spread":0.3541056914323059,"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."}}