{"id":"W4386355230","doi":"10.1101/2023.08.31.23294925","title":"Efficient Edge-AI Models for Robust ECG Abnormality Detection on Resource-Constrained Hardware","year":2023,"lang":"en","type":"preprint","venue":"medRxiv","topic":"ECG Monitoring and Analysis","field":"Medicine","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Center for Long-Term Cybersecurity, University of California Berkeley; University of Sydney; Crohn's and Colitis Foundation of Canada","keywords":"Abnormality; Computer science; Generalization; Identification (biology); Enhanced Data Rates for GSM Evolution; Resource (disambiguation); Artificial intelligence; Machine learning; Pattern recognition (psychology); Data mining; Medicine; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.001150762,0.0004107432,0.0007574646,0.0004311464,0.0002233231,0.00006390732,0.0002280045,0.0004729815,0.00001907682],"category_scores_gemma":[0.0004477017,0.0003680511,0.0006981567,0.0002995218,0.00008400229,0.0000160545,0.0001970635,0.001004417,0.00005766541],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002470893,"about_ca_system_score_gemma":0.0001414234,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001346271,"about_ca_topic_score_gemma":0.00002742777,"domain_scores_codex":[0.9973723,0.000108639,0.0005719443,0.0009318445,0.0005412297,0.0004740889],"domain_scores_gemma":[0.9980091,0.0002137357,0.0002327165,0.001001476,0.0002947031,0.0002482268],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004276604,0.0002893007,0.005913273,0.001317335,0.0005614092,0.000081654,0.0003717525,0.9718856,0.0005025565,0.00002675421,0.0008280356,0.01779463],"study_design_scores_gemma":[0.001013007,0.0002173716,0.004857302,0.001067161,0.0007761719,0.00001237875,0.0001852694,0.9849215,0.005154001,0.0004133765,0.0009666291,0.0004158796],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.747622,0.0001245726,0.244483,0.0022572,0.001940372,0.001250516,0.0002013308,0.0008788733,0.001242132],"genre_scores_gemma":[0.9914874,0.0000224076,0.00206314,0.0001654837,0.001426615,0.000314658,0.0002101098,0.0001033554,0.004206771],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2438655,"threshold_uncertainty_score":0.9998772,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0715977255041392,"score_gpt":0.3045120493612354,"score_spread":0.2329143238570962,"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."}}