From whole-body dosimetry to local dose assessment: a 3-y experience in interventional practice

This 3-y retrospective study analyzed 180 paired dosimetry records from five interventional radiologists and five cardiologists, comparing whole-body Hp(10) with eye-lens Hp(3) and extremity Hp(0.07) doses. Moderate-to-strong linear correlations were observed (r = 0.86-0.91, R2 = 0.73-0.83), indicating that chest-level dosemeters may support the estimation of localized exposure. Coefficients of variation were ~30% for k1 and 27% for k2, reflecting heterogeneous exposure conditions. Variance-component analysis showed that inter-operator variability accounted for 20%-25% of total variance, versus 75%-80% residual within-operator variability, likely reflecting differences in workload, case mix, positioning, exposure geometry, shielding, and measurement uncertainty. The eye-lens conversion coefficient was higher in cardiologists (1.8) than in the non-neurointerventional radiology/neurointerventional radiology group (0.68), suggesting that operator-group-specific conversion coefficients may be more appropriate than a single universal coefficient. Empirically derived conversion coefficients provide a valuable complementary approach for the retrospective estimation of localized dose exposure when dedicated dosimetry is unavailable.