{"paper":{"title":"From Static Risk to Dynamic Trajectories: Toward World-Model-Inspired Clinical Prediction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A unified framework links patient forecasts, counterfactual treatment paths, and policy checks by jointly modeling disease, treatment choices, and observation biases.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bin Cui, Erik Cambria, Min Hun Lee, Pujun Feng, Seyed Ehsan Saffari, Siew-Kei Lam, Tao Tan, Tong Yang, Xiaoyu Guo, Xiaoyu Zhang, Xibin Sun, Yangtao Zhou, Yue Sun","submitted_at":"2026-05-16T10:45:26Z","abstract_excerpt":"Clinical decision-making is a feedback system where risk estimates influence treatment, which in turn changes disease trajectories, and both shape clinicians' measurement practices. Static prediction often fails clinically: models trained on observational care logs conflate disease biology with clinician behavior, particularly under treatment confounder feedback and irregular or informative observation. This Review focuses on intervention-aware disease trajectory modeling in clinical AI--methods estimating patient-specific longitudinal disease evolution and assessing trajectory changes under a"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We present the first unified framework bridging forecasting, counterfactual trajectories, and policy evaluation across discrete/continuous time, explicitly addressing treatment assignment, time-varying confounding, and observation bias.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the six linked components (three decision tasks and three data-generating mechanisms) are sufficient to determine identifiability and to comprehensively map existing method families without material loss of structure or coverage from prior literature.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A review proposing a unified framework for intervention-aware disease trajectory modeling in clinical AI, organized around three decision tasks and three data-generating mechanisms.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A unified framework links patient forecasts, counterfactual treatment paths, and policy checks by jointly modeling disease, treatment choices, and observation biases.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b30e528fd971965b9f657d64b5a1e2af421df3b0efd4a8dc37bc16d200030b79"},"source":{"id":"2605.16927","kind":"arxiv","version":1},"verdict":{"id":"1353bd7c-5733-47cd-9b20-9e84aac0e1d4","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T20:34:32.107867Z","strongest_claim":"We present the first unified framework bridging forecasting, counterfactual trajectories, and policy evaluation across discrete/continuous time, explicitly addressing treatment assignment, time-varying confounding, and observation bias.","one_line_summary":"A review proposing a unified framework for intervention-aware disease trajectory modeling in clinical AI, organized around three decision tasks and three data-generating mechanisms.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the six linked components (three decision tasks and three data-generating mechanisms) are sufficient to determine identifiability and to comprehensively map existing method families without material loss of structure or coverage from prior literature.","pith_extraction_headline":"A unified framework links patient forecasts, counterfactual treatment paths, and policy checks by jointly modeling disease, treatment choices, and observation biases."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16927/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T21:01:19.158117Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T20:40:52.139328Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"cited_work_retraction","ran_at":"2026-05-19T20:22:26.545207Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T18:41:56.257130Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.338199Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9c0fd07674535714248a9405f8a15cad28a2b54009cbd17e3fe7d26e25b53f05"},"references":{"count":107,"sample":[{"doi":"","year":1978,"title":"An empirical transition matrix for non-homogeneous markov chains based on censored observations.Scandinavian Journal of Statistics, 5(3):141–150, 1978","work_id":"72d6615f-697e-49fe-88d9-4629c808484c","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Analyzing patient trajectories with artificial intelligence.Journal of medical internet research, 23(12):e29812, 2021","work_id":"9a167847-45fa-4eca-b199-82720d3f8c68","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1993,"title":"Gill, and Niels Keiding.Statistical Models Based on Counting Processes","work_id":"b633b7a3-0586-42c1-a71e-69d6fe672a5d","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1996,"title":"Joshua D. Angrist, Guido W. Imbens, and Donald B. Rubin. Identification of causal effects using instrumental variables.Journal of the American Statistical Association, 91(434):444–455, 1996","work_id":"f8dbc394-89f9-4899-b1a7-272330153140","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Peter C. Austin, Frank E. Harrell, and Davide van Klaveren. Graphical calibration curves and the integrated calibration index (ICI) for survival models.Statistics in Medicine, 39(21):2714–2742, 2020","work_id":"90446b6c-af46-41a0-9134-ae9fc833ed14","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":107,"snapshot_sha256":"a5e06950e24e33cde2bc51d1a41be5a1a02b4e4e66daf4b0f2d0594c426bc768","internal_anchors":3},"formal_canon":{"evidence_count":2,"snapshot_sha256":"f3e4c4ed8bf45bd243b4ed13299520aa3e32ac56249845c129a553d864ce3ae5"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}