A human-centered framework for Miss ICU: A multimodal, evidence-integrated artificial intelligence agent for clinical decision-making in intensive care
Keywords:
Artificial Intelligence, Critical Care, Clinical Decision Support Systems, Machine Learning, Explainable AI, Digital healthAbstract
Background and aim: The modern Intensive Care Unit (ICU) is a data-rich environment where an overwhelming volume of information can cause cognitive overload for clinicians. Artificial Intelligence (AI) is promising for predictive tasks, but adoption is limited by an implementation gap driven by poor data interoperability, model opacity, and inadequate integration into clinical workflows. This paper presents the "Miss ICU" framework, a novel AI agent designed as an intelligent partner for the critical care team. Its goal is to create an integrated, interpretable, and evidence-based system that augments, rather than replaces, clinical intelligence to improve patient outcomes.
Methods: The framework uses a three-tiered architecture. A Data Harmonization Layer applies standards such as Health Level Seven/Fast Healthcare Interoperability Resources (HL7/FHIR) to build a unified patient data model. An Analytics Engine integrates machine learning for patient-state modeling, Natural Language Processing (NLP) for real-time evidence synthesis from medical literature, and Explainable AI (XAI) for transparency. A human-centered Presentation Layer provides an intuitive dashboard to display patient trajectory and risk.
Results: The envisioned system offers clinicians a holistic view of patient status, including a stability curve to depict clinical trajectory. It aims to generate predictions (e.g., sepsis) with human-readable explanations of contributing factors and follows a multi-phase validation pathway ending in a pragmatic randomized controlled trial.
Conclusions: Miss ICU offers a blueprint for next-generation clinical decision support, emphasizing interoperability, interpretability, and integration of patient data with medical evidence to reduce cognitive load, enable proactive care, and improve safety and quality in critical care.
References
1. Gal DB, Han B, Longhurst C, Scheinker D, Shin AY. Quantifying Electronic Health Record Data: A Potential Risk for Cognitive Overload. Hosp Pediatr. 2021 Feb;11(2):175-8. doi: 10.1542/hpeds.2020-002402
2. Park J, Zhong X, Dong Y, Barwise A, Pickering BW. Investigating the cognitive capacity constraints of an ICU care team using a systems engineering approach. BMC Anesthesiol. 2022 Jan 4;22(1):10. doi: 10.1186/s12871-021-01548-7.
3. Arnold M, Goldschmitt M, Rigotti T. Dealing with information overload: a comprehensive review. Front Psychol. 2023 Jun 21;14:1122200. doi: 10.3389/fpsyg.2023.1122200
4. Asgari E, Kaur J, Nuredini G, et al. Impact of Electronic Health Record Use on Cognitive Load and Burnout Among Clinicians: Narrative Review. JMIR Med Inform. 2024 Apr 12;12:e55499. doi: 10.2196/55499
5. Biesheuvel LA, Dongelmans DA, Elbers PWG. Artificial intelligence to advance acute and intensive care medicine. Curr Opin Crit Care. 2024 Jun 1;30(3):246-50. doi: 10.1097/MCC.0000000000001150.
6. Lv X, Liu D, Chen X, et al. Machine learning for the prediction of mortality in patients with sepsis-associated acute kidney injury: a systematic review and meta-analysis. BMC Infect Dis. 2024 Dec 21;24(1):1454. doi: 10.1186/s12879-024-10380-6
7. Yue S, Li S, Huang X, et al. Machine learning for the prediction of acute kidney injury in patients with sepsis. J Transl Med. 2022 May 13;20(1):215. doi: 10.1186/s12967-022-03364-0
8. Zhuang C, Hu R, Li K, et al. Machine learning prediction models for mortality risk in sepsis-associated acute kidney injury: evaluating early versus late CRRT initiation. Front Med (Lausanne). 2025 Jan 22;11:1483710. doi: 10.3389/fmed.2024.1483710
9. Gao T, Nong Z, Luo Y, et al. Machine learning-based prediction of in-hospital mortality for critically ill patients with sepsis-associated acute kidney injury. Ren Fail. 2024 Dec;46(1):2316267. doi: 10.1080/0886022X.2024.2316267
10. Pinsky MR, Bedoya A, Bihorac A, et al. Use of artificial intelligence in critical care: opportunities and obstacles. Crit Care. 2024 Apr 8;28(1):113. doi: 10.1186/s13054-024-04860-z
11. Gonzalez FA, Santonocito C, Lamas T, et al. Is artificial intelligence prepared for the 24-h shifts in the ICU? Anaesth Crit Care Pain Med. 2024 Dec;43(6):101431. doi: 10.1016/j.accpm.2024.101431
12. Tungushpayev M, Suleimenova D, Sarria-Santamerra A, Aimyshev T, Gaipov A, Viderman D. The value of machine and deep learning in management of critically ill patients: An umbrella review. Int J Med Inform. 2025 Aug 8;204:106081. doi: 10.1016/j.ijmedinf.2025.106081
13. Shapoo N, Shapoo N, Rehman A, Boma N. Smart Healthcare: The Role of Digital Health in Modern Medicine. Health Care Sci. 2025 Jun 1;4(3):179-87. doi: 10.1002/hcs2.70019
14. Suresh V, Singh KK, Vaish E, et al. Artificial Intelligence in the Intensive Care Unit: Current Evidence on an Inevitable Future Tool. Cureus. 2024 May 7;16(5):e59797. doi: 10.7759/cureus.59797
15. Cao Y, Wang Y, Liu H, Wu L. Artificial intelligence revolutionizing anesthesia management: advances and prospects in intelligent anesthesia technology. Front Med (Lausanne). 2025 Aug 6;12:1571725. doi: 10.3389/fmed.2025.1571725
16. Iapascurta V, Fiodorov I, Belii A, Bostan V. Multi-Agent Approach for Sepsis Management. Healthc Inform Res. 2025 Apr;31(2):209-14. doi: 10.4258/hir.2025.31.2.209
17. Chen CJ, Fang YS, Liu CF. An Integrated AI Cloud Sharing Framework for Predictive AI and Generative AI in Healthcare. Stud Health Technol Inform. 2025 May 15;327:233-4. doi: 10.3233/SHTI250314
18. Bai Y, Gu B, Tang C. Enhancing Real-Time Patient Monitoring in Intensive Care Units with Deep Learning and the Internet of Things. Big Data. 2025 Jan 17. doi: 10.1089/big.2024.0113
19. Chen T, Li Q, Zhao D, et al. Utilisation of AI-driven chatbots for perioperative health information seeking: a descriptive qualitative study of orthopaedic patients and family members. BMJ Open. 2025 Sep 4;15(9):e099824. doi: 10.1136/bmjopen-2025-099824
20. González-Nóvoa JA, Campanioni S, Busto L, et al. Improving Intensive Care Unit Early Readmission Prediction Using Optimized and Explainable Machine Learning. Int J Environ Res Public Health. 2023 Feb 16;20(4):3455. doi: 10.3390/ijerph20043455
21. Mollerus F, Lynch C, Bruining H. Data interoperability for a systems approach to developmental conditions. Neurosci Biobehav Rev. 2025 Sep;176:106245. doi: 10.1016/j.neubiorev.2025.106245
22. Mataraso S, D'Souza S, Seong D, Berson E, Espinosa C, Aghaeepour N. Benchmarking of pre-training strategies for electronic health record foundation models. JAMIA Open. 2025 Aug 13;8(4):ooaf090. doi: 10.1093/jamiaopen/ooaf090
23. Contreras M, Silva B, Shickel B, et al. Real-time prediction of intensive care unit patient acuity and therapy requirements using state-space modelling. Nat Commun. 2025 Aug 8;16(1):7315. doi: 10.1038/s41467-025-62121-1
24. Huang J, Galal G, Etemadi M, Vaidyanathan M. Evaluation and Mitigation of Racial Bias in Clinical Machine Learning Models: Scoping Review. JMIR Med Inform. 2022 May 31;10(5):e36388. doi: 10.2196/36388
25. Cecconi M, Greco M, Shickel B, et al. Implementing Artificial Intelligence in Critical Care Medicine: a consensus of 22. Crit Care. 2025 Jul 8;29(1):290. doi: 10.1186/s13054-025-05532-2
26. Amith MT, Cui L, Zhi D, et al. Toward a standard formal semantic representation of the model card report. BMC Bioinformatics. 2022 Jul 14;23(Suppl 6):281. doi: 10.1186/s12859-022-04797-6
27. Gilbert S, Adler R, Holoyad T, Weicken E. Could transparent model cards with layered accessible information drive trust and safety in health AI? NPJ Digit Med. 2025 Feb 25;8(1):124. doi: 10.1038/s41746-025-01482-9
28. Wilkinson MD, Dumontier M, Aalbersberg IJ, et al. The FAIR Guiding Principles for scientific data management and stewardship. Sci Data. 2016 Mar 15;3:160018. doi: 10.1038/sdata.2016.18
29. Lyu G. Data-driven decision making in patient management: a systematic review. BMC Med Inform Decis Mak. 2025 Jul 1;25(1):239. doi: 10.1186/s12911-025-03072-x
30. Bignami EG, Russo M, Bellini V. Reclaiming Patient-Centered Care: How Intelligent Time is Redefining Healthcare Priorities. J Med Syst. 2025 Feb 21;49(1):30. doi: 10.1007/s10916-025-02163-6
31. Barrett M, Boyne J, Brandts J, et al. Artificial intelligence supported patient self-care in chronic heart failure: a paradigm shift from reactive to predictive, preventive and personalised care. EPMA J. 2019 Nov 22;10(4):445-64. doi: 10.1007/s13167-019-00188-9
32. Workum JD, Meyfroidt G, Bakker J, et al. AI in critical care: A roadmap to the future. J Crit Care. 2025 Sep 23;91:155262. doi: 10.1016/j.jcrc.2025.155262
33. Karako K, Tang W. Applications of and issues with machine learning in medicine: Bridging the gap with explainable AI. Biosci Trends. 2025 Jan 14;18(6):497-504. doi: 10.5582/bst.2024.01342
34. Müller L, Gangadharaiah R, Klein SC, et al. An open access medical knowledge base for community driven diagnostic decision support system development. BMC Med Inform Decis Mak. 2019 Apr 27;19(1):93. doi: 10.1186/s12911-019-0804-1
35. Wu JY, Wang Y, Ching CTS, Wang HD, Liao LD. IoT-based wearable health monitoring device and its validation for potential critical and emergency applications. Front Public Health. 2023 Jun 16;11:1188304. doi: 10.3389/fpubh.2023.1188304
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Elena Giovanna Bignami, Michele Russo, Erika Cerdelli, Roberto Lanza, Valentina Bellini
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Transfer of Copyright and Permission to Reproduce Parts of Published Papers.
Authors retain the copyright for their published work. No formal permission will be required to reproduce parts (tables or illustrations) of published papers, provided the source is quoted appropriately and reproduction has no commercial intent. Reproductions with commercial intent will require written permission and payment of royalties.
How to Cite
