Abstract: In recent decades, modern healthcare systems need to respond to the challenges of constantly increasing patient demand, a growing number of complex diseases, and patients with multimorbidity. This results in constrained hospital resources and limited financing from governments to effectively deliver care to patients. For operational decision-making, the timely forecasting of patient arrivals and resource demand is of utmost importance. Although the importance of forecasting has been recognized, analysing, monitoring, and forecasting multivariate time series in healthcare delivery systems remain challenging. Experts must become involved in system key performance measurement, resulting in significant resources being allocated to analyse and monitor processes but ultimately missing the timely nature of forecasting. This work develops a new hierarchical recurrent neural network (RNN) model to provide forecasts for a comprehensive resource allocation problem. First, a monitoring framework is developed to provide insightful analyses of operational difficulties. Also, a new deep learning framework is designed to leverage the derived univariate time series distributions and capture correlations at multiple aggregation levels using a hierarchical RNN framework to produce simultaneous forecasts of the timing and magnitudes of resource distribution.
In healthcare operations, resource allocation significantly influences healthcare delivery efficiency and patient wait-time, for which cloud computing platforms have been adopted and developed in hospitals. This review first identifies cloud-based AI techniques to aid healthcare operations literature and then uses a systematic dual-faceted framework to systematically review cloud-based AI applications in healthcare operations literature from three perspectives: type of AI techniques and methods, applications of AI in healthcare operations, and dimensions used to separate healthcare operations research problems. The findings reveal that (i) the cloud platform has been mainly adopted in healthcare as a cost-effective and efficient data storage and sharing solution, (ii) few studies have investigated the cloud platform’s value in AI-based decision-making optimization, (iii) cloud-based AI techniques are ignition-infrastructure to drive healthcare transformation, which justifies the need of more studies that develop and deploy cloud-based AI techniques to address healthcare operations optimization problems.
Keywords: AI in healthcare operations, Hospital resource optimization, Cloud-based hospital management, Predictive analytics in healthcare, Healthcare cloud computing, AI hospital resource planning, Smart hospital infrastructure, Machine learning in operations management, Cloud-enabled healthcare analytics, Real-time hospital data management, AI-powered patient flow optimization, Dynamic bed allocation system, Intelligent staff scheduling, Healthcare logistics AI, Operational efficiency in hospitals.
Copyright Form