International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | A Privacy-Preserving Federated Deep Learning Framework for Early ICU Deterioration Prediction Using Multivariate Time-Series Data |
|---|---|
| ABSTRACT | Early prediction of Intensive Care Unit (ICU) admission is essential for timely clinical intervention, optimal resource allocation, and improved patient outcomes. While deep learning models have shown superior predictive performance for critical care tasks, their practical adoption remains constrained by concerns regarding patient data privacy and limited interpretability of model decisions. To address these challenges, this study proposes an explainable and privacy-preserving deep learning framework for early ICU admission prediction using multivariate clinical time-series data derived from Electronic Health Records (EHRs). The proposed framework leverages Long Short-Term Memory (LSTM) networks to model complex temporal dependencies in physiological signals. To safeguard patient confidentiality and comply with healthcare data regulations, a federated learning paradigm is employed, enabling collaborative model training across multiple institutions without centralized data sharing. Furthermore, explainable artificial intelligence (XAI) techniques, specifically SHAP and LIME, are integrated to provide both global and patient-specific interpretability, thereby enhancing clinical transparency and trust. Extensive experiments conducted on publicly available ICU datasets demonstrate that the proposed framework achieves superior predictive performance compared to traditional machine learning models and centralized deep learning approaches, while maintaining strict privacy preservation. The results indicate that combining federated learning with explainable deep learning offers a scalable, ethical, and clinically trustworthy solution for AI-driven ICU decision support systems. |
| TITLE | |
| AUTHOR | SHAMSIYA PARVEEN Assist. Professor, Dept. of CSE (Data Science), SSIT, Tumakuru, India |
| VOLUME | 182 |
| DOI | DOI: 10.15680/IJIRCCE.2026.1403020 |
| pdf/20_A Privacy-Preserving Federated Deep.pdf | |
| KEYWORDS | |
| References | 1. Y. Deng, S. Liu, Z. Wang, Y. Wang, Y. Jiang, and B. Liu, “Explainable time series deep learning models for the prediction of mortality, prolonged length of stay and 30 day readmission in intensive care patients,” Front. Med., vol. 9, p. 933037, Sep. 2022, doi:10.3389/fmed.2022.933037. 2. M. Mesinovic, P. Watkinson, and T. Zhu, “Explainable machine learning for predicting ICU mortality in myocardial infarction patients using pseudo dynamic data,” Sci. Rep., vol. 15, p. 27887, Jul. 2025, doi:10.1038/s41598-025-13299-3. 3. M. Z. Fuzail, I. Ud Din, S. Ahmed, A. Alhumam, and A. H. Khan, “Optimizing sepsis mortality prediction using hybrid federated learning and explainable AI framework,” Sci. Rep., 2026. 4. Y. Benhamou, S. Kalyan, and S. Kumar, “Federated learning for ICU mortality prediction: balancing accuracy and privacy in a multi hospital setting,” bioRxiv, Aug. 2025. 5. W. Fathy, G. Emeriaud, and F. Cheriet, “A comprehensive review of ICU readmission prediction models: from statistical methods to deep learning approaches,” Artif. Intell. Med., vol. 165, 2025. 6. T. Tunduny and B. Shibwabo, “Explainable AI approaches in federated learning: systematic review,” JMIR AI, vol. 1, p. e69985, Feb. 2026. 7. Y. Ive et al., “AI assisted prediction of unplanned intensive care admissions with SHAP and LIME explanations,” Nat. Digit. Med., 2025. 8. C. Stylianides, “AI advances in ICU with an emphasis on sepsis prediction,” MDPI 2504 4990, vol. 7, 2025. 9. X. Li, J. Gu, Z. Wang, Y. Yuan, B. Du, and F. He, “XAI for in hospital mortality prediction via multimodal ICU data,” AAAI, 2023. 10. S. Chen et al., “Interpretable machine learning model for early prediction of 30 day mortality in ICU patients,” arXiv, Jun. 2025. |
