A dynamic sequential decision-making model on MRI real-time scheduling with simulation-based optimization.

Area of Science:

• Operations Research
• Healthcare Management
• Medical Imaging Technology

Background:

• Magnetic Resonance Imaging (MRI) is crucial for diagnostics but faces scheduling inefficiencies due to inherent uncertainties.
• These uncertainties in patient arrival, scanning, and preparation times lead to significant delays and increased operational costs in MRI centers.

Purpose of the Study:

• To develop and evaluate a real-time decision-making framework for MRI scanner job assignment and sequencing.
• To optimize MRI operations by overriding fixed appointment schedules using real-time data.

Main Methods:

• A sequential decision-making framework combined with a simulation-based solution method was employed.
• The system utilizes real-time data on waiting patients, MRI scanner status, and dynamic scanning time information.
• A ranking and selection method was used to optimize simulation replications for efficiency and accuracy.

Main Results:

• The proposed method reduced patient waiting time by 21.7% and improved MRI scanner utilization by 23.0%.
• Compared to offline methods, an optimality gap of 13.6% was observed.
• In simulations, the method outperformed current policies, reducing average patient wait time by 17.93 minutes and improving scanner utilization by 7.20%.

Conclusions:

• Real-time rescheduling significantly enhances MRI operational efficiency.
• The proposed simulation-based approach provides a robust and accurate method for dynamic MRI scheduling.
• This optimization strategy offers substantial improvements in patient throughput and resource utilization within diagnostic imaging.