Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{210533,
        author = {ABDUL RAHEEM F and AKSHAYA PRIYA T and BALARUBAN R and KAMALESH S and GOMATHI LAKSHMI},
        title = {Explainable Multimodal AI for Early ICU Deterioration Prediction},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {6},
        pages = {1202-1208},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=210533},
        abstract = {Detecting early signs of deterioration in patients within Intensive Care Units can be a complex task. Threshold-based alerting methods may struggle to account for the relationship between different clinical parameters. In this work we propose an Explainable Multimodal AI framework for prediction of early deterioration based on physiological measurements, clinical notes and patient mobility/positioning. We use Machine Learning, Deep Learning, Biomedical Language Models and Computer Vision to extract meaningful features from each modality. These extracted features are then fused using a Multimodal approach to produce risk scores. The Shapley Additive exPlanations (SHAP) method was used to determine which key factors were driving the predictions. An interactive dashboard is presented showing the risk levels, factors driving risk and trends in patient data. This framework will provide healthcare professionals with a tool to help detect high risk patients in ICU earlier. This framework should be seen as a clinical decision support tool and not a replacement for medical expertise.},
        keywords = {ICU deterioration prediction, multimodal AI, explainable AI, clinical decision support, deep learning, SHAP, biomedical language models, patient monitoring.},
        month = {November},
        }