Fast Occupational Upper-Limb Radiation Dose Prediction Using Machine Learning and Monte Carlo Simulation

Key Points

• The aim is to develop a rapid and scalable method for predicting upper-limb radiation doses using machine learning.
• Utilized ensemble machine learning techniques for dose estimation.
• Compared predictions against full Monte Carlo simulations in interventional environments.
• Assessed speed and scalability of the proposed method.
• Achieved dose estimation in around 10 ms with high precision.
• Indicated that machine learning provides a feasible alternative to traditional Monte Carlo simulations.
• Supported radiation protection optimization in image-guided procedures.

Abstract