CHYMERA: AI-Powered System for Automated and Explainable Quality Assessment of Chest Radiographs

Chest radiography is one of the most widely used diagnostic tools in clinical practice, yet a significant proportion of images suffer from technical deficiencies that compromise their diagnostic value. Common errors such as patient rotation, incomplete lung field inclusion, or the presence of artifacts can lead to unnecessary radiation exposure, reduced workflow efficiency, and inaccurate diagnoses. Studies indicate that between 5 % and 15 % of chest X-rays are technically inadequate, directly affecting care quality and increasing healthcare costs. Mandatory hospital audits often reveal these issues, but lack objective tools to address them systematically. In this context, CHYMERA (CHest x-raY Machine LEarning Audit) emerges as a comprehensive AI-powered solution designed to automate and standardize the quality control of chest radiographs. Leveraging deep learning techniques and modular components for anatomical analysis, segmentation, artifact detection, and technical classification, CHYMERA delivers interpretable, reproducible, and real-time assessments. Its integration into clinical workflows through an intuitive interface, along with its audit-ready capabilities-such as performance tracking by operator or session-make it a strategic asset for hospitals aiming to reduce variability, comply with regulatory standards, and enhance patient safety. This multidimensional approach not only improves technical image quality but also optimizes healthcare resources and strengthens trust in radiology services.