Critical Care Unit (CCU) patients often benefit from being referred to dietitians for various reasons. This can help improve recovery time, resulting in more effective utilisation of valuable resources within the NHS (National Health Service) in the United Kingdom. However, said resources are often in high demand with scarce availability. Therefore, in this paper we propose an AI-based dashboard that can help clinicians automatically identify such patients, thereby reducing workload as well as cognitive load on clinical staff. We have trained various machine learning classifiers using various physiological measures of CCU patients and have identified a Support Vector Machine (SVC) classifier as the best performing model (AUC: 0.78). Our investigation shows promise results that significantly improve quality of patient care within the NHS. In future we intend to undertake more extensive evaluation of the dashboard developed as well as extend this work to paediatric patients.
Soomro, K., Pimenidis, E., & McWilliams, C. (2022). Supporting patient nutrition in critical care units. In Engineering Applications of Neural Networks: EANN 2022: Engineering Applications of Neural Networks (128-136). https://doi.org/10.1007/978-3-031-08223-8_11