|M.Sc Student||Carmeli Boaz|
|Subject||Real-Time Optimization of Patients Flow in Emergency|
|Department||Department of Industrial Engineering and Management||Supervisor||Professor Avishai Mandelbaum|
|Full Thesis text|
Emergency Departments (EDs) are hectic, highly stochastic environments that deal with human lives under severe resource restrictions. ED personnel must provide quality clinical service and maintain an acceptable level of patient satisfaction while using limited operational resources.
In this work we consider the required features and main characteristics of a real-time ED monitoring-and-control system. We then focus on two specific applications, namely i) monitoring the real-time ED load and ii) optimizing internal ED patient flow through real-time control.
A good real-time monitoring-and-control system provides a holistic view of the entire ED operation, emphasizing information collection, analysis and display. We analyze the ED operation from multiple dimensions and viewpoints, e.g., taking clinical, operational, and service-level aspects into account. We focus on monitoring approaches and optimization techniques that can be deployed and used within a real-time ED monitoring-and-control system.
We developed an innovative load monitoring and measurement approach based on a neural networks paradigm. We thus enable adaptation of the load function into a specific ED setting, using subjective load perception provided by a specific user or a user group.
We analyzed service policies to optimize the ED patient flow by addressing the following question: Which patient should a physician treat next? For that, we provide an optimal control, based on a fluid model analysis and discrete event simulation. Deploying the resulting service policies within a real-time monitoring-and-control system would enable ED management and staff to improve overall ED operations.