An Optimized Decision Support System for Maintenance Strategy for Radiology Devices

Abstract

Radiology devices are essential for accurate disease diagnosis, necessitating effective maintenance practices. Optimizing maintenance strategies is vital to improving the availability and efficacy of healthcare systems, especially in settings with limited resources. This study introduces an optimized decision support system for selecting the most appropriate maintenance strategy for medical radiology devices. The maintenance strategy was classified into three categories: predictive, preventive, and corrective. The approach integrates the Multi-Attributive Border Approximation Area Comparison (MABAC) method, a Multi-Criteria Decision-Making (MCDM) technique, with machine learning algorithms. The methodology utilizes key input factors—age ratio, failure rate, preventive maintenance frequency, downtime ratio, and service costs—to determine the optimal maintenance plan. Five different types of radiology devices were tested. Results demonstrate that the integration of the MABAC and the support vector machine algorithm could effectively prioritize maintenance strategies, contributing to improved patient safety and healthcare outcomes.