Statistical-Based Modeling Strategy for Entrance Skin Dose Estimation in Patient Undergoing Body Interventional Radiology

抄録

Nowadays, interventional radiology (IR) plays an important role in a wide range of invasive diagnostic imaging and minimally invasive imaging-guided therapeutic procedures. However, the radiological risk associated with unwanted side effects in patients is a concern. Radiation-induced erythema, skin burns, and epilation have been reported in the literature. For safety reasons, risk assessment in IR patients is quantified with the entrance skin dose (ESD) exceeding 2 Gy is unacceptable. However, direct dose measurement for each patient is labor-intensive and impractical because the detector probe may cause perturbation on the X-ray image during a fluoroscopy-guided procedure. The purpose of this study is to develop a statistical-based model for ESD estimation in the body-IR procedures by using simple quantities derived from the DICOMRDSR data. As a result, a simplified ESD calculation model is introduced without complicated parameters generally used in conventional calculation methods, such as a backscatter factor and a mass-energy absorption coefficient. This user-friendly ESD calculation method makes it easy for patients and medical staff to calculate the dose they received by themselves after ending each IR examination. The models are classified into two groups of body-IR procedures: 1) the vascular procedure includes TACE and TAE, and 2) the non-vascular procedure includes PTBD and PCD. The study provides that the simplified regression models are sufficient to estimate the patient entrance skin dose for both groups of body IR procedures, accounting for a 95% confidence interval.