An enhanced grey wolf optimizer boosted machine learning prediction model for patient-flow prediction.

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Operations Research

Background:

• Hospitals increasingly use AI and big data to improve outpatient services and reduce wait times.
• Current AI systems often fall short of expectations due to factors like environment, patient, and physician behaviors.
• Accurate patient flow prediction is crucial for effective medical resource management and timely patient access.

Purpose of the Study:

• To develop an advanced patient-flow prediction model to address limitations in current hospital resource management.
• To forecast patient medical requirements by considering dynamic patient flow patterns and objective rules.
• To enhance the efficiency and accuracy of hospital outpatient service optimization.

Main Methods:

• Proposed a novel optimization algorithm, SRXGWO (Sequence, Cauchy, and Directional Mutation Grey Wolf Optimizer).
• Integrated SRXGWO with Support Vector Regression (SVR) to create the SRXGWO-SVR patient-flow prediction model.
• Validated SRXGWO's performance through benchmark function experiments and compared it with twelve other algorithms.

Main Results:

• The SRXGWO-SVR model demonstrated superior prediction accuracy and lower error rates compared to seven other models.
• SRXGWO algorithm showed high performance in ablation and peer algorithm comparison tests.
• The model effectively forecasts patient flow dynamics and medical needs using training and testing datasets.

Conclusions:

• The SRXGWO-SVR model offers a reliable and efficient solution for patient-flow forecasting in hospitals.
• This system can significantly aid hospitals in optimizing medical resource management and improving operational efficiency.
• Enhanced patient flow prediction contributes to better hospital outpatient service quality and reduced patient wait times.