Uncertainty-Aware Decision Support for Hiv/Aids Treatment Using Fuzzy Logic

Abstract

The treatment of HIV/AIDS presents significant challenges due to uncertainties in clinical data, patient adherence variability, and evolving medical guidelines. Traditional crisp decision models often fail to capture these uncertainties, leading to abrupt and sometimes unrealistic treatment recommendations. This study proposes an uncertainty-aware decision support system based on a Mamdani Fuzzy Inference System (FIS) to enhance clinical decision-making for HIV/AIDS treatment. The system incorporates three input variables—CD4 count, viral load, and adherence level and generate an output variable representing treatment urgency. Membership functions and a comprehensive rule base of 27 rules enable the system to process uncertain or imprecise inputs, while the centroid defuzzification method provides a crisp urgency score. Simulation results demonstrate that the fuzzy model yields nuanced and clinically interpretable recommendations, balancing biological and behavioral factors to produce smooth transitions across urgency levels. By addressing the limitations of crisp threshold models, the proposed system provides a robust, transparent, and flexible tool for improving treatment prioritization in HIV/AIDS care.