Improving Predictive Modeling of Intensive Care Unit Outcomes through Temporal Deep Learning Networks and Multimodal Physiological Signal Fusion
Keywords:
Multimodal Fusion, Temporal Deep Learning, Critical Care Informatics, Systems Architecture, Algorithmic Fairness, Clinical Decision SupportAbstract
Predictive modeling in the intensive care unit remains a critical challenge due to the high velocity, heterogeneity, and volumetric scale of patient data. Traditional clinical risk scores rely on static or coarsely aggregated physiological measures, failing to capture the rich, high-frequency temporal dynamics of critically ill patients. This paper introduces an integrated socio-technical and computational framework for intensive care unit outcome prediction utilizing advanced temporal deep learning networks and multimodal physiological signal fusion. By synthesizing high-frequency physiological waveforms, sparse laboratory results, and unstructured clinical text, the proposed architecture captures complex cross-modal interactions while preserving unique longitudinal trends. Beyond the algorithmic innovations, this study provides a comprehensive system-level analysis focusing on infrastructure deployment, technical trade-offs, and data governance. We examine the structural trade-offs between early, late, and intermediate fusion, demonstrating how intermediate cross-attention structures mitigate information loss. Furthermore, the paper addresses critical deployment challenges within existing electronic health record infrastructures, detailing the necessity of scalable streaming pipelines, robust data harmonization, and real-time inference mechanics. Social and ethical dimensions, including computational sustainability, data privacy, and demographic fairness, are extensively investigated alongside clinical governance policies. Our findings indicate that maximizing predictive accuracy requires a systematic orchestration of deep learning pipelines, rigorous data infrastructure, and strict ethical guardrails to ensure safe, equitable, and stable critical care decision support.
References
Abuhamad, H. (2025). Integrative multimodal hybrid data fusion for mortality prediction. Journal of Biomedical Informatics, 154, 104-115.
An, Y., Liu, Y., Chen, X., & Sheng, Y. (2022). TERTIAN: Clinical endpoint prediction in ICU via time-aware transformer-based hierarchical attention network. Computational Intelligence and Neuroscience, 2022, 1-13. https://doi.org/10.1155/2022/4207940
Blair, J. M. A., Webber, M. A., Baylay, A. J., Ogbolu, D. O., & Piddock, L. J. V. (2015). Molecular mechanisms of antibiotic resistance. Nature Reviews Microbiology, 13(1), 42–51. https://doi.org/10.1038/nrmicro3399
Chiu, C. C., Wu, C. M., Chien, T. N., Kao, L. J., Li, C., & Jiang, H. L. (2022). Applying an improved stacking ensemble model to predict the mortality of ICU patients with heart failure. Journal of Clinical Medicine, 11(21), 6460. https://doi.org/10.3390/jcm11216460
Choi, E., Bahadori, M. T., Schuetz, A., Stewart, W. F., & Sun, J. (2016). DOCTOR AI: Predicting clinical events via recurrent neural networks. Proceedings of the Machine Learning for Healthcare Conference, 56, 301-318.
Cooley-Rieders, M., & Zheng, K. (2023). LSTM-based deep learning model to predict ICU mortality using raw nursing notes. International Journal of Medical Informatics, 172, 105-114.
El-Rashidy, N., El-Sapagh, S., Islam, S. R. M., & Abdelrazek, S. (2021). A stacking ensemble model to predict intensive care unit mortality using electronic health records. IEEE Access, 9, 114532-114548.
Harutyunyan, H., Khachatrian, H., Kale, D. C., Ver Steeg, G., & Galstyan, A. (2019). Multitask learning and benchmarking with clinical time series data. Scientific Data, 6(1), 96. https://doi.org/10.1038/s41597-019-0103-9
Huang, M., Jiao, J., Wang, J., & Wang, Y. (2021). Multimodal deep learning for mortality prediction in intensive care units. IEEE Journal of Biomedical and Health Informatics, 25(8), 3162-3172.
Johnson, A. E. W., Pollard, T. J., Shen, L., Lehman, L. H., Feng, M., Ghassemi, M., Moody, B., Szolovits, P., Celi, L. A., & Mark, R. G. (2016). MIMIC-III, a freely accessible critical care database. Scientific Data, 3(1), 160035. https://doi.org/10.1038/sdata.2016.35
Johnson, A. E. W., Bulgarelli, L., Shen, L., Gayles, A., Shammout, A., Horng, S., Celi, L. A., & Pollard, T. J. (2023). MIMIC-IV, a freely accessible electronic health record dataset. Scientific Data, 10(1), 1. https://doi.org/10.1038/s41597-022-01899-x
Khadanga, S., Aggarwal, A., Podchiherstvo, S., & Srivastava, S. (2022). Using clinical notes and discharge summaries for improved mortality prediction in intensive care units. ACM Transactions on Computing for Healthcare, 3(2), 14-28.
Li, X., Zhang, Y., & Wang, H. (2024). Association between mental disorders and intensive care unit mortality: A multi-center retrospective analysis. PLOS One, 19(3), e0294812.
Martin, J. (2025). Explainable multimodal deep learning models for variable-length sequences in critically ill patients. Journal of Biomedical Informatics, 162, 105001.
Moor, M., Banerjee, O., Abad, Z. S. H., Krumholz, H. M., Leskovec, J., Topol, E. J., & Rajpurkar, P. (2023). Foundation models for generalist medical artificial intelligence. Nature, 616(7956), 259–265. https://doi.org/10.1038/s41586-023-05881-4
Nazih, W. (2024). Mortality prediction for ICU patients with mental disorders using large language models ensemble and unstructured medical notes. PLOS One, 19(11), e0332134.
Niu, K., Zhang, K., Peng, X., Pan, Y., & Xiao, N. (2023). Deep multi-modal intermediate fusion of clinical record and time series data in mortality prediction. Frontiers in Molecular Biosciences, 10, 1136071. https://doi.org/10.3389/fmolb.2023.1136071
Rajkomar, A., Oren, E., Chen, K., Dai, A. M., Hajaj, N., Hardt, M., Liu, P. J., Liu, X., Marcus, J., Sun, M., Sundberg, P., Yee, H., Zhang, K., Zhang, Y., Flores, G., Duggan, G. E., Irvine, J., Duncan, Q., Wang, S., ... Dean, J. (2018). Scalable and accurate deep learning with electronic health records. NPJ Digital Medicine, 1(1), 18. https://doi.org/10.1038/s41746-018-0029-1
Sadanandan, B. (2026). Multimodal deep learning for early prediction of patient deterioration in the ICU: Integrating time-series EHR data with clinical notes. arXiv preprint, arXiv:2603.14719.
Saleh, H. (2024). TransformerFusionNet: A real-time multimodal framework for ICU heart failure mortality prediction using big data streaming. Proceedings of the IEEE International Conference on Bioinformatics and Biomedicine, 432-439.
Seymour, C. W., Liu, V. X., Iwashyna, T. J., Brunkhorst, F. M., Rea, T. D., Scherag, A., Rubenfeld, G., Kahn, J. M., Shankar-Hari, M., Singer, M., Deutschman, C. S., Escobar, G. J., & Angus, D. C. (2016). Assessment of clinical criteria for sepsis: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA, 315(8), 762–774. https://doi.org/10.1001/jama.2016.0288
Shickel, B., Tighe, P. J., Bihorac, A., & Rashidi, P. (2018). Deep EHR: A survey on recent advances in deep learning techniques for electronic health record data. IEEE Journal of Biomedical and Health Informatics, 22(5), 1589–1604. https://doi.org/10.1109/JBHI.2017.2778709
Smith, G. B., Prytherch, D. R., Meredith, P., Schmidt, P. E., & Featherstone, P. I. (2014). The ability of the National Early Warning Score (NEWS) to discriminate between patients who are and are not at risk of acute mortality or cardiac arrest. Resuscitation, 84(4), 465-470.
Su, Y., Zhang, R., & Luo, C. (2020). Collaborative multimodal deep learning for critical care event forecasting. Information Fusion, 63, 112-124.
Sun, T., Zhang, M., & Wang, L. (2021). Bedside vital sign monitoring and predictive modeling in the modern intensive care unit. Journal of Healthcare Engineering, 2021, 1-15.
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30, 5998–6008.
Wang, Q. (2024). Multimodal integration of physiological signals clinical data and medical imaging for ICU outcome prediction. Journal of Clinical and Translational Systems, 8(1), 195-207.
Yang, J., Li, Y., & Gao, M. (2021). A multimodal deep neural network considering both structured time series and unstructured clinical records for intensive care unit mortality prediction. Health Informatics Journal, 27(2), 1-16.
Zhu, Z., Wang, S., & Zhang, Y. (2018). Bidirectional LSTM networks with supervised attention mechanisms for physiological time series modeling in critical care. Artificial Intelligence in Medicine, 92, 44-53.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Clinical and Translational Medicine
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
