{"id":"W4379919680","doi":"10.48550/arxiv.2306.03407","title":"LESS: Label-efficient Multi-scale Learning for Cytological Whole Slide Image Screening","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"AI in cancer detection","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Compute Canada; Nvidia","keywords":"Computer science; Artificial intelligence; Pattern recognition (psychology); Task (project management); Inefficiency; Machine learning; Image (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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00193449,0.00151967,0.001826926,0.001237616,0.0005190545,0.001237486,0.004150619,0.002136093,0.003190895],"category_scores_gemma":[0.003811746,0.0007003893,0.001316326,0.001056681,0.0008193046,0.001940211,0.002420008,0.001714208,0.001475151],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001056094,"about_ca_system_score_gemma":0.001157058,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005166643,"about_ca_topic_score_gemma":0.006737388,"domain_scores_codex":[0.9988527,0.0002821245,0.00004562617,0.000413353,0.0002765052,0.0001297221],"domain_scores_gemma":[0.9985805,0.0004947634,0.0001556225,0.0003611054,0.0002751531,0.0001329705],"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":[0.0006374082,0.0006424348,0.006985537,0.0003689482,0.0003378805,0.0002744014,0.0001571993,0.1747977,0.03462842,0.004399545,0.01968331,0.7570873],"study_design_scores_gemma":[0.00003014577,0.00005740321,0.0004966732,0.000005933322,0.00001613975,0.00005811537,0.00001672823,0.9924594,0.003140611,0.002879825,0.0008282092,0.00001090613],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05598797,0.001239492,0.9253018,0.0005751124,0.0001359244,0.0002323782,0.0005906146,0.0141725,0.001764228],"genre_scores_gemma":[0.494968,0.0004666579,0.4907955,0.001498186,0.0002325755,0.0003531034,0.004140181,0.0009497711,0.006596082],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005166643,"threshold_uncertainty_score":0.01067466,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1876221967195119,"score_gpt":0.2426375994714252,"score_spread":0.05501540275191333,"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."}}