Data-Driven Expert System for Tuberculosis (TB) Diagnosis Using the Forward Chaining Method

Abstract

Tuberculosis (TBC) is a disease caused by Mycobacterium tuberculosis, one of the oldest known diseases affecting humans. While it primarily affects the lungs, about one-third of cases involve other organs, underscoring the importance of early detection and accurate diagnosis. To address this, a data-driven expert system has been developed to assist in diagnosing tuberculosis and providing relevant information to users. An expert system is a form of intelligent software that leverages data and expert knowledge to solve complex problems. In this study, the Forward Chaining method is applied, utilizing a rule-based approach to process data and conclusions from known facts. This method iteratively matches facts to rules, deriving new insights until a conclusion is reached or no further matches are found. If the premise satisfies the conditions (evaluated as TRUE), the system generates a decision. The system is designed to simplify the recognition of tuberculosis symptoms by analyzing user-provided data to produce accurate diagnostic results and actionable solutions. Findings indicate that the data-driven approach enhances the system's ability to provide precise diagnoses and recommendations, ensuring reliability and effectiveness. This work demonstrates the value of integrating data-driven methodologies in expert systems to improve healthcare delivery, particularly in the early detection and management of tuberculosis.