{"id":"W4417117542","doi":"10.21203/rs.3.rs-8194963/v1","title":"Accelerating Machine Learning in Healthcare: Addressing the Labelling Bottleneck","year":2025,"lang":"","type":"preprint","venue":"Research Square","topic":"ECG Monitoring and Analysis","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"SickKids Foundation; Hospital for Sick Children","funders":"Canadian Institutes of Health Research; University of Toronto","keywords":"Bottleneck; Scalability; Annotation; Transfer of learning; Modalities; Deep learning","routes":{"ca_aff":true,"ca_fund":true,"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":[0.006367934,0.001227268,0.002226572,0.001510296,0.000883476,0.002984032,0.002245811,0.003487169,0.007749315],"category_scores_gemma":[0.04046335,0.0008943493,0.0007112816,0.00241601,0.001029381,0.004403103,0.003506616,0.003742323,0.004351771],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009877129,"about_ca_system_score_gemma":0.003540846,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005778309,"about_ca_topic_score_gemma":0.005402517,"domain_scores_codex":[0.9969743,0.001477555,0.0001353078,0.0005032345,0.0006509706,0.00025864],"domain_scores_gemma":[0.9790164,0.01401073,0.0006278839,0.002898473,0.002960742,0.000485746],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00123498,0.0005349845,0.004660649,0.0007044938,0.0001666749,0.0001635353,0.0002904117,0.1733511,0.01045157,0.0324945,0.04539245,0.7305548],"study_design_scores_gemma":[0.00006009791,0.00008117224,0.000643737,0.00003436046,0.00003236463,0.00007145474,0.00004188326,0.9452441,0.003896399,0.04540505,0.004477002,0.00001245225],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03232825,0.002401345,0.9510648,0.005327325,0.000606088,0.0001154129,0.0005362276,0.00399373,0.003626813],"genre_scores_gemma":[0.3421468,0.001917942,0.6391562,0.001446285,0.001155862,0.0002224636,0.001939976,0.001069067,0.01094535],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007749315,"threshold_uncertainty_score":0.03367728,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2645008355560837,"score_gpt":0.4908586045679754,"score_spread":0.2263577690118918,"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."}}