{"id":"W4392202241","doi":"10.1016/j.bpj.2024.02.019","title":"VSG-GAN: A high-fidelity image synthesis method with semantic manipulation in retinal fundus image","year":2024,"lang":"en","type":"article","venue":"Biophysical Journal","topic":"Retinal Imaging and Analysis","field":"Medicine","cited_by":9,"is_retracted":false,"has_abstract":false,"ca_institutions":"York University","funders":"Horizon 2020; European Research Council; Hong Kong Baptist University; Horizon 2020 Framework Programme; National Natural Science Foundation of China; European Commission","keywords":"Fundus (uterus); Image (mathematics); Retinal; Fidelity; Computer science; Fundus camera; Artificial intelligence; High fidelity; Ophthalmology; Computer vision; Medicine; Physics; Ophthalmoscopy; Telecommunications","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":[0.0006944413,0.000224386,0.0005031976,0.0003037745,0.00009720371,0.0002774516,0.0001074841,0.00006714916,0.0001791416],"category_scores_gemma":[0.0001745588,0.0001524666,0.000259569,0.0006856367,0.000109319,0.0002888577,0.00002513813,0.000787372,0.0001345263],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001585938,"about_ca_system_score_gemma":0.000108116,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002012746,"about_ca_topic_score_gemma":0.000005217822,"domain_scores_codex":[0.9980841,0.0002167638,0.0004362618,0.0003745937,0.0005396589,0.0003486219],"domain_scores_gemma":[0.9990659,0.0002231408,0.0001031186,0.0002267163,0.0001585596,0.0002225322],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"observational","study_design_scores_codex":[0.0005164772,0.0003511069,0.002763778,0.0003586611,0.000304694,0.005602726,0.0001671248,0.00004257803,0.9700978,0.0005596085,0.0007313333,0.01850405],"study_design_scores_gemma":[0.003159824,0.001491032,0.3753917,0.008030727,0.005128636,0.01170846,0.0009856649,0.2825675,0.305065,0.00375968,0.001226436,0.001485381],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9080191,0.000152672,0.08113519,0.009096615,0.0001306699,0.0001443974,0.000007084308,0.0001130514,0.001201259],"genre_scores_gemma":[0.9672989,0.00004189185,0.03144223,0.0001120307,0.0007833166,0.000006941275,0.000006724415,0.0000366429,0.0002713665],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6650328,"threshold_uncertainty_score":0.6217409,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01820131349746352,"score_gpt":0.318829243944009,"score_spread":0.3006279304465455,"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."}}