{"id":"W3165750456","doi":"10.1038/s42256-021-00337-8","title":"End-to-end privacy preserving deep learning on multi-institutional medical imaging","year":2021,"lang":"en","type":"article","venue":"Nature Machine Intelligence","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":470,"is_retracted":false,"has_abstract":false,"ca_institutions":"Institute on Governance","funders":"Deutschen Konsortium für Translationale Krebsforschung; Technische Universität München; Deutsche Forschungsgemeinschaft; Imperial College London; UK Research and Innovation","keywords":"Computer science; USable; Inference; Convolutional neural network; Encryption; Deep learning; Artificial intelligence; Machine learning; Information privacy; End-to-end principle; Computer security; Data mining; World Wide Web","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":["metaresearch","metaepi_narrow","open_science","research_integrity"],"consensus_categories":["open_science"],"category_scores_codex":[0.0009987683,0.0004028429,0.0003318272,0.0003406635,0.0004497687,0.0003524912,0.02759207,0.0004248859,0.0003653807],"category_scores_gemma":[0.1229651,0.0003563722,0.0001324621,0.001466298,0.0001883942,0.0007297582,0.07477295,0.003827776,0.0002582227],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000243035,"about_ca_system_score_gemma":0.0003477205,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008853612,"about_ca_topic_score_gemma":0.0001228311,"domain_scores_codex":[0.9956165,0.0001931963,0.0004947986,0.001362061,0.001617992,0.0007154858],"domain_scores_gemma":[0.9929295,0.0007315982,0.0001375705,0.005589839,0.0002971437,0.0003143571],"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":[0.00002334378,0.0003734915,0.004674393,0.00006150921,0.00007157546,0.002232591,0.0003578792,0.005285458,0.00147398,0.06271638,0.008054907,0.9146745],"study_design_scores_gemma":[0.0002516472,0.00005298749,0.002009378,0.0003502325,0.000009030034,0.0004782987,0.00004797841,0.8594131,0.05903198,0.04263984,0.0351075,0.0006079609],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002017092,0.003051688,0.9294896,0.0612027,0.00119904,0.0001798949,0.00001102472,0.001088207,0.001760765],"genre_scores_gemma":[0.603992,0.0001919211,0.3930202,0.002495129,0.000138561,0.00002121676,0.0000398863,0.00002253521,0.00007844169],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9140666,"threshold_uncertainty_score":0.9998888,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01857445252678108,"score_gpt":0.3149303193573083,"score_spread":0.2963558668305272,"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."}}