Purpose: Effective dose (E) is useful as a dose index for patient exposures in interventional radiology; therefore, the authors estimated E from the kerma-area product (P(KA)) utilized during hepatic embolization interventional radiology cases performed at a cancer center and determined the variation of such doses over a representative patient population.
Methods: A single-center, IRB-approved retrospective study was performed to estimate doses from consecutive hepatic embolization procedures performed during 2006. Organ doses E and E/P(KA) were determined from patient height, weight, P(KA), procedure geometry factors, beam quality, the PCXMC Monte Carlo model, and the International Commission on Radiological Protection organ weighting factors.
Results: One hundred thirteen patients were included in the study population, 72 males and 41 females, with a median age of 63 yr (29-89 yr), weight of 79 kg (42-111 kg), height of 170 cm (147-188 cm), and P(KA) of 233 Gy cm2 (9-1020 Gy cm2). E was directly correlated with P(KA) r2 = 0.8 (p < 0.01), with a median E/P(KA) of 0.18 mSv Gy(-1) cm(-2) (0.12-0.33 mSv Gy(-1) cm(-2)). The E/P(KA) ratio was inversely and exponentially correlated with weight r2 = 0.9 (p < 0.001). The median E (mSv) for the study patient population was 44 mSv (2.0-255 mSv).
Conclusions: Values of E can be estimated utilizing patient-specific and procedure-specific parameters. The strong inverse correlation of E/P(KA) with patient weight allows simple estimation of E from P(KA) and patient weight. There is a wide variation in effective dose in oncologic hepatic embolizations with doses up to an order of magnitude higher than diagnostic imaging of the abdomen by CT radiology. Variation is likely due to patient geometry, clinical technique factors, and procedure complexity.