Artificial Intelligence-Driven Decision Support System with Multi-Machine Learning Algorithms for Predicting Diabetes: A Case Study at Alhada Armed Forces Hospital, Saudi Arabia.

Abstract

Diabetes causes severe health complications and high mortality rates worldwide, making early prediction and intervention essential to prevent life-threatening outcomes. This study, conducted at Alhada Armed Forces Hospital in Taif, Saudi Arabia, leverages artificial intelligence to develop a machine learning models for predicting diabetes risk. Using a comprehensive dataset of 1,200 patients and incorporating key clinical features such as systolic and diastolic blood pressure, pulse rate, temperature, respiratory rate, weight, height, SpO2 levels, random and fasting blood glucose (RBS and FBS), and urine output. This research evaluates multiple machine learning models, such as Random Forest, Support Vector Machine, and Gradient Boosting. Each model is assessed for accuracy, F-score, precision, recall, and AUC-ROC, focusing on identifying high-impact risk factors to facilitate personalized clinical recommendations. By enhancing predictive capabilities in a real-world healthcare setting, this AI-driven decision support system provides clinicians with accurate, timely insights, ultimately aiming to improve patient outcomes and optimize healthcare resources.