Predictive modeling of eye lens dose in interventional radiology: a polynomial regression approach to cumulative fluoroscopy dose.

Objectives: This study aimed to investigate the relationship between cumulative fluoroscopy dose and eye lens radiation dose among interventional radiologists, and to develop a predictive model to enhance occupational radiation safety. Methods: We collected data from interventional radiologists, focusing on cumulative fluoroscopy dose during procedures and corresponding eye lens doses. A quadratic polynomial regression model was developed to assess the non-linear relationship between cumulative fluoroscopy dose and eye lens dose. The study involved the use of machine-generated cumulative dose data and personal eye lens dosimeters. Results: The quadratic polynomial regression model effectively captured the non-linear relationship for cumulative doses >20 Gy, enabling precise dose prediction at higher exposure levels where cataract risks escalate. However, the model showed limited accuracy for doses ≤ 20 Gy. This model allowed for more precise prediction of eye lens dose, particularly at higher exposure levels where the risks of radiation-induced cataracts increase significantly. Conclusions: The quadratic polynomial regression model serves as a potentially valuable tool for real-time monitoring in high-exposure scenarios (>20 Gy), supporting radiation safety protocols in clinical practice. Integration into routine hospital systems may enhance radiation protection protocols and inform policy development, aligning occupational dose monitoring practices with international safety standards.