NIAR-Saúde presents paper on responsible AI at SBCAS 2026

NIAR-Saúde took part in the 26th Brazilian Symposium on Computing Applied to Health (SBCAS 2026), held from June 1–4, 2026, in Ouro Preto (MG), at UFOP’s Arts and Conventions Center. SBCAS is one of the country’s leading forums bringing together researchers from the computing and health fields.

On the occasion, the group presented the paper “Responsible AI for Public Health: A Methodological Illustration with a Forecasting Model applied to Respiratory Hospitalizations on SUS Data”, published in the symposium proceedings. The work uses data from Brazil’s Unified Health System (SUS) to illustrate, in practice, how to build forecasting models for respiratory hospitalizations with a focus on methodological rigor.

The study is one of the first works to apply the framework developed by NIAR for assessing responsibility in machine learning models. As an illustrative example, it presents fairness analyses across different Brazilian federative units, along with auditing and explainability procedures that allow a better understanding of the model’s behavior and its potential impacts.

By prioritizing aspects beyond predictive performance — such as transparency, reproducibility, and bias mitigation — the work reinforces NIAR-Saúde’s commitment to developing trustworthy artificial intelligence solutions to support public health. The paper is authored by Ramon G. Pereira, Luís Eduardo Limas Brito, Italo Avelar, Matheus Carvalho, Marisa Vasconcelos, Michele A. Brandão, and Wagner Meira Jr.

The full version of the work is available in the SBCAS 2026 proceedings and can be accessed on the publications page.