Smarthealth Sleman: Aplikasi Prediksi Penyebaran Tuberkulosis Dengan Kejadian Anemia Berbasis Artificial Intelligence

The need to find TB cases early and be able to provide accurate and reliable predictions of TB spread requires applications to support government programs in TB prevention and control. Through the development of the "SmartHealth Sleman" application by combining the power of two algorithms, Naive Bayes Classification and K-Nearest Neighbors, this study is expected to be able to predict the risk of TB spread and anemia incidence in TB patients in real time. The existence of an application with visualization integration and a website-based interface is expected to be a tool in decision-making at the level of health services and local government policies. Research objectives: to develop AI-based predictive applications capable of mapping the risk of TB spread and anemic incidence, to apply and compare the performance of the Neive Bayes Classification and K- Nearest Neighbors in the context of local data, and to provide spatial visualizations and interactive dashboards to support medical decision-making and government policies. The research method used is a spatial analysis approach that analyzes data based on geographical location and factors that affect the spread of TB and applies and compares the performance of Neive Bayes Classification and K-Nearest Neighbors in the context of local data related to AI-based predictions that are able to map the risk of TB spread and anemia incidence in Sleman Regency. Data on TB patients was obtained from the TB 03 register unit of the Sleman Regency Health Office covering 25 work areas of the health center in 2020-2024 including age, gender, type of TB, TB classification, number of cases, location (puskesmas and sub-districts), occupation, education level, distance of the patient's home to health services. Determination of anemia status was carried out through examination of hematological profiles using Sysmex SN1000R. The creation of an AI-based "SmartHealth Sleman" application was used to assess the predictive risk of TB spread with the incidence of anemia presented in the form of a mapping (ArcView GIS version 3.1) for spatial visualization and an interactive dashboard that is expected to support medical decision-making and government policies.