Work

SALUS: A Sustainable Framework for Aligning Brazilian Public Health Image Databases in Emergency Care

Year: 2026
Type: preprint
Abstract: Brazilian public health image databases are fragmented and poorly integrated, which compromises both evidence-based policy evaluation and emergency clinical care. These databases are pre-existing, autonomous, distributed, and heterogeneous, and the lack of interoperability prevents the reconstruction of a patient's therapeutic itinerary, the measurement of health outcomes, and the rapid retrieval of imaging exams when a patient arrives unconscious at an emergency unit. A patient identified by a single CPF may have imaging exams scattered across SIA/SUS, SIH/SUS, e-SUS APS, and municipal hospital systems, each with its own numbering and retention policy. This pre-project addresses this gap by proposing SA-LUS, a sustainable framework that aligns health images through sequential numbering and temporal compaction. Inspired by the SALUS v0 prototype, the framework renames images into continuous sequences, detects gaps, duplicates , and inconsistent padding, moves the oldest images to a historical archive, and generates consolidated DOCX and PDF reports with the most recent exams first. The SALUS program itself will be developed during the doctoral research and released as open source, so that public institutions can adopt, audit, and extend it without relying on foreign cloud infrastructure. Its methodological feasibility will be demonstrated through a proof of concept on a public medical image benchmark with temporal labels.
Language: English
Open Access status: green
OpenAlex ID: W7225869357
Cites:
Cited by:
Related to: 10