{"id":"W4414189846","doi":"10.1016/j.imu.2025.101691","title":"FedDeepInsight—A privacy-first federated learning architecture for medical data","year":2025,"lang":"en","type":"article","venue":"Informatics in Medicine Unlocked","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute of Infection and Immunity","funders":"Nederlandse Brandwonden Stichting; Nederlandse Organisatie voor Wetenschappelijk Onderzoek; Health~Holland; National Weather Association","keywords":"Differential privacy; Federated learning; Deep learning; Dimensionality reduction; Architecture; Information privacy; Stability (learning theory); Data transformation; Data modeling","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004947288,0.0005079798,0.0008247249,0.0007583093,0.0009251597,0.002267638,0.002650048,0.001220309,0.00176901],"category_scores_gemma":[0.007596513,0.0003243923,0.0008770048,0.001080779,0.001190592,0.004328948,0.003582567,0.002080836,0.0006351761],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001553271,"about_ca_system_score_gemma":0.002797066,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002869188,"about_ca_topic_score_gemma":0.003097805,"domain_scores_codex":[0.9981508,0.0005274061,0.0001612682,0.0004874952,0.0004660999,0.000206954],"domain_scores_gemma":[0.997221,0.0005670976,0.0001926332,0.001372836,0.0004948486,0.0001516139],"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":[0.001258909,0.0006173814,0.007822207,0.0002120974,0.0003048877,0.0003584689,0.0004730428,0.3671089,0.009919913,0.08313872,0.01314097,0.5156446],"study_design_scores_gemma":[0.00003317917,0.0001286103,0.000484944,0.00002793167,0.00002600748,0.0001497874,0.00005782533,0.9330576,0.01003396,0.0501749,0.005800435,0.00002476708],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01917366,0.0001537871,0.9747127,0.0004921471,0.00006281621,0.0001058567,0.000172635,0.004143558,0.0009826412],"genre_scores_gemma":[0.5963932,0.0002505307,0.3978997,0.000665021,0.00004624727,0.0002657053,0.001065635,0.000182987,0.003231035],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004947288,"threshold_uncertainty_score":0.02616405,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03629533177841489,"score_gpt":0.3234449371476492,"score_spread":0.2871496053692343,"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."}}