Intelligent ergonomic Rating for Radiation protection Aprons Design Induced by Artificial Neural neTworks
Duration and funding: 1 year – 90K€
IRRADIANT uses physics-guided learning models to develop lighter, more ergonomic lead aprons for radiation protection, adapted to all genders and morphologies, thereby reducing the risk of musculoskeletal disorders (MSDs). These disorders, caused by prolonged postures and repetitive strain, impact the health of medical staff, reducing their comfort and performance at work.
CONTEXT
The use of lead aprons is essential for protecting medical staff from radiation exposure during X-ray procedures, particularly in radiology and operating rooms. However, their heavy and rigid design places significant strain on the shoulders, back, and joints, increasing the risk of musculoskeletal disorders. These issues, caused by prolonged postures and repetitive strain, can negatively affect the well-being, comfort, and performance of healthcare professionals.
This project aims to develop lighter, more ergonomic lead aprons that accommodate all body types and genders. Leveraging cutting-edge technologies such as artificial intelligence and motion capture systems, the project enables detailed analysis of personnel’s postures and movements during operations, ensuring an optimized fit and reduced physical strain.
CHALLENGES
The evaluation of musculoskeletal disorders (MSDs) in the operating room is complex due to the limitations of motion capture technologies and the configuration of the capture environment. Motion capture systems offer high precision but require a controlled environment and markers that may interfere with movement. RGB cameras, while flexible and non-invasive, lack reliability for detailed biomechanical analysis, particularly due to occlusions caused by multiple people interacting nearby. Inertial measuring unit (IMU) sensors, though suited to real-world conditions, suffer from drift errors and limited accuracy. Additionally, the diversity of surgeons’ morphologies, genders, and health conditions complicates the design of a universal system. Adaptive learning to these parameters is essential for ensuring a reliable assessment, often requiring a combination of multiple technologies.
RESULTS
Coming Soon