DEVELOPMENT OF AN EXPERT SYSTEM FOR PERSONALIZED INSULIN DOSAGE PREDICTION

Abstract

Diabetes mellitus is a chronic metabolic disorder that requires continuous monitoring and accurate insulin administration to maintain optimal blood glucose levels. Improper insulin dosing can lead to severe complications, including hypoglycemia and hyperglycemia, affecting patient health and quality of life. This study presents the development of an expert system for personalized insulin dosage prediction aimed at supporting clinical decision-making and improving diabetes management. The proposed system integrates medical knowledge, patient-specific parameters, and rule-based reasoning to generate individualized insulin dosage recommendations. Key input variables include blood glucose levels, carbohydrate intake, body weight, age, physical activity, insulin sensitivity, and previous insulin administration records. The expert system employs a knowledge base developed in consultation with diabetes specialists and incorporates inference mechanisms to evaluate patient conditions and determine appropriate dosage adjustments. By analyzing real-time and historical patient data, the system provides personalized recommendations that adapt to changing physiological and lifestyle factors. The developed model seeks to reduce human error, enhance treatment accuracy, and support self-management among diabetic patients. Furthermore, the system offers explainable recommendations, enabling healthcare professionals and patients to understand the rationale behind dosage predictions. Experimental evaluation demonstrates the effectiveness of the proposed approach in generating reliable insulin dosage suggestions aligned with established clinical guidelines. The expert system has the potential to improve glycemic control, reduce diabetes-related complications, and increase patient adherence to treatment plans. This research highlights the role of artificial intelligence and expert systems in advancing personalized healthcare solutions and contributing to more efficient and patient-centered diabetes management practices.