{"prediction": {"id": "eaed069e-6e13-561f-8bcb-d3b2b9816ed7", "work_id": "f7bb253b-2359-567f-a2ef-83e114ac89c2", "study_id": "71694fad-fa5d-5e7c-80f3-787c4eeaed32", "kind": "Model evaluation", "pages": [9], "version": "Version of record", "source_work_id": "f7bb253b-2359-567f-a2ef-83e114ac89c2", "title": "TwinRL-Onco: A World Model-Empowered Digital Twin Framework with Hierarchical Reinforcement Learning for Venetoclax Resistance Trajectory Prediction and Adaptive Therapy Optimization in Chronic Lymphocytic Leukemia.", "doi": "10.1109/jbhi.2026.3696685", "pdf_sha256": "86eeffde96e51d1a4e7d5d61c8ebd8ef1956f0f3f10a6b697ae40f00fcbce736", "physical_page_count": 12}, "fields": {"measured_variable": {"value": "lymphocyte_count", "source_basis": "verified_profile", "pages": [9]}, "measurement_role": {"value": "prediction_error", "source_basis": "original_pdf_review", "pages": [9]}, "metric": {"value": "root_mean_square_error", "source_basis": "original_pdf_review", "pages": [9]}, "population": {"value": "simulated_system", "source_basis": "verified_profile", "pages": [1, 9, 10]}, "reference_type": {"value": "synthetic_reference", "source_basis": "verified_profile", "pages": [9]}, "scenario": {"value": "within_model_multistep_forecast", "source_basis": "verified_profile", "pages": [9]}}, "endpoints": [{"term": "agreement", "label": "Correlation or agreement"}, {"term": "cell_state", "label": "Cell state and composition"}, {"term": "prediction_error", "label": "Prediction or fitting error"}], "validations": [{"id": "a1f8311c-3504-5bf0-b04c-aa4cc56efe5d", "claim_id": "eaed069e-6e13-561f-8bcb-d3b2b9816ed7", "coverage": "direct_comparison", "support": "supports", "source": "same_source", "overlap": "overlap", "comparability": "matched", "statistics": "not_reported", "anchors": [{"work_id": "f7bb253b-2359-567f-a2ef-83e114ac89c2", "version": "Version of record", "pages": [10]}]}], "model_resources": {"total": 0, "page": 1, "limit": 50, "rows": [], "review_status": "indexed_source"}}