Background: Scheduling patient appointments in hospitals is complicated due to various types of patient examinations, different departments and physicians accessed, and different body parts affected.
Objective: This study focuses on the radiology scheduling problem, which involves multiple radiological technologists in multiple examination rooms, and then proposes a prototype system of computer-aided appointment scheduling based on information such as the examining radiological technologists, examination departments, the patient's body parts being examined, the patient's gender, and the patient's age.
Methods: The system incorporated a stepwise multiple regression analysis (SMRA) model to predict the number of examination images and then used the K-Means clustering with a decision tree classification model to classify the patient's examination time within an appropriate time interval.
Results: The constructed prototype creates a feasible patient appointment schedule by classifying patient examination times into different categories for different patients according to the four types of body parts, eight hospital departments, and 10 radiological technologists.
Conclusion: The proposed patient appointment scheduling system can schedule appointment times for different types of patients according to the type of visit, thereby addressing the challenges associated with diversity and uncertainty in radiological examination services. It can also improve the quality of medical treatment.
Keywords: Data mining; appointment scheduling system; decision tree; patient appointments; stepwise multiple regression analysis.