Improving MRI Safety Through AI: The Role of the STAID System in Supporting Healthcare Staff

Abstract

The aim of this study is to integrate artificial intelligence (AI) into magnetic resonance imaging (MRI) workflows to enhance patient safety and reduce clinical risks. The STAID (Stretcher and Temperature AI Detection) system was developed as a Raspberry Pi 5-based model equipped with computer vision and thermal detection technologies to monitor body temperature and detect the presence of stretchers, for the most part incompatible with the MRI environment. The system was tested on an image sample, achieving an accuracy of 80.85% in stretcher recognition with an average inference time of 0.12 seconds. Results indicate that STAID is easily integrated into clinical workflows, with positive acceptance from healthcare staff and patients, though some concerns remain regarding privacy and the reduction of human interaction. The introduction of STAID demonstrates AI's potential to improve risk management in MRI, paving the way for further adoption of AI technologies in clinical settings. Future prospects include deeper integration with hospital systems and the development of training protocols for personnel, contributing to a safer, more technologically advanced healthcare environment.