DEVELOPMENT AND VALIDATION OF A TUBERCULOSIS CASE PREDICTION MODEL IN THE STATE OF GOIÁS (2025-2026) WITH A FOCUS ON COMPUTATIONAL PATHOLOGY

• DOI: https://doi.org/10.22533/at.ed.1595162524035

• Palavras-chave: Tuberculosis, Computational Pathology, Prediction, Goiás, Time Series, Public Health.

• Keywords: Tuberculosis, Computational Pathology, Prediction, Goiás, Time Series, Public Health.

• Abstract: Introduction: Globally, tuberculosis (TB) is a public health issue. In Goiás, in 2023, there was a 9.42% increase in TB cases compared to 2022. The approach of computational pathology is an essential tool for analyzing data that can contribute to controlling the Tb epidemic. Objective: Develop and validate a model statistical capable of to predict the number of TB notifications for 2025 and 2026. Methods: A computational algorithm was developed to construct a time series model for predicting notifications. This model provided better estimates of AIC, BIC, MSE, and RMSE compared to an self-adjusting model from an R Studio software library. The model developed in this study was applied to TB notification data in Goiás from 2001 to 2023. Results: A gradual increase in TB notifications in Goiás was estimated for 2025 and 2026, with peaks observed in January, March, September, and October. Additionally, a decrease in TB notifications was noted in February, July, and December for 2025, and in February, June, and December for 2026. Conclusion: These findings can significantly contribute to public health planning and decision-making aimed at controlling TB in the region.