aiqbench analysis
This research introduces a novel architectural approach to maintaining ICU prediction accuracy by decoupling stable physiological signals from evolving treatment protocols, effectively isolating model drift to the latter stream alone. By integrating temporal retrieval with PubMed evidence and automated audit logs for adaptation triggers, the framework ensures that clinical AI evolves alongside changing hospital practices without distorting fundamental patient biology or losing historical consistency.
Biological Amnesia in ICU Time-Series Prediction: A Drift-Adaptive Two-Stream Architecture with Temporal Retrieval
arXiv:2607.19020v1 Announce Type: cross Abstract: Background: Clinical decision support systems degrade silently as treatment protocols evolve, yet standard adaptation methods treat models as monolithic blocks, unable to distinguish stable patient physiology from shifting institutional practice. Methods: We propose an adaptive clinical intelligence architecture for ICU intervention prediction that structurally decouples physiological from treatment representations, confining parameter updates to the treatment stream upon a dual distributional and accuracy trigger. Automated audit logs record which treatment features drove each adaptation event and how their importance shifted. At inference, an attribution-driven Temporal RAG module grounds each prediction in patient-specific, era-matched PubMed evidence anchored to the patient's dominant physiological features. Experiments used 84,792 MIMIC-IV stays (2008-2022) under strict chronological split. Results: Drift localised entirely to the treatment stream, validating the structural prior. Selective adaptation improved vasopressor and septic shock discrimination and calibration over the static source model. A fully retrained baseline yielded marginally higher aggregate discrimination but missed 26 septic shock cases the framework correctly identified, with none in the reverse direction; retrieval consistency with the pre-adaptation source model was preserved by the framework but degraded substantially in the retrained baseline. Conclusions: Structurally constraining adaptation to drifting components while preserving stable physiological representations enables clinical AI to evolve with practice without distorting learned patient biology. This architecture offers a template for governable, interpretable deployment of adaptive models in high-stakes clinical environments.
Abstract:Background: Clinical decision support systems degrade silently as treatment protocols evolve, yet standard adaptation methods treat models as monolithic blocks, unable to distinguish stable patient physiology from shifting institutional practice. Methods: We propose an adaptive clinical intelligence architecture for ICU intervention prediction that structurally decouples physiological from treatment representations, confining parameter updates to the treatment stream upon a dual distributional and accuracy trigger. Automated audit logs record which treatment features drove each adaptation event and how their importance shifted. At inference, an attribution-driven Temporal RAG module grounds each prediction in patient-specific, era-matched PubMed evidence anchored to the patient's dominant physiological features. Experiments used 84,792 MIMIC-IV stays (2008-2022) under strict chronological split. Results: Drift localised entirely to the treatment stream, validating the structural prior. Selective adaptation improved vasopressor and septic shock discrimination and calibration over the static source model. A fully retrained baseline yielded marginally higher aggregate discrimination but missed 26 septic shock cases the framework correctly identified, with none in the reverse direction; retrieval consistency with the pre-adaptation source model was preserved by the framework but degraded substantially in the retrained baseline. Conclusions: Structurally constraining adaptation to drifting components while preserving stable physiological representations enables clinical AI to evolve with practice without distorting learned patient biology. This architecture offers a template for governable, interpretable deployment of adaptive models in high-stakes clinical environments.
10 pages, 3 figures, 8 tables. Code and aggregate audit logs available at: [this https URL]. Under Review
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Quantitative Methods (q-bio.QM)
arXiv:2607.19020 [cs.LG]
(or arXiv:2607.19020v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2607.19020 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 343 words
Related coverage
NVIDIA Cosmos-H-Dreams: Bringing Real-Time Generative Simulation to Surgical Robotics
Hugging Face · Jul 27, 2026
Amazon is investing in the Lean Focused Research Organization
Amazon Science · Jul 26, 2026
Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · Sep 25, 2026
GNSS-FM: A Self-Supervised Foundation Model for Daily GNSS Displacement Time Series
arXiv · Jun 9, 2026