An Independent Comparison of VARSKIN v2.1 and Geant4 for Skin Dose Calculations

VARSKIN is becoming more commonplace in nuclear medicine departments to calculate skin dose from droplet or surface contamination due to its accessibility and ease of use. VARSKIN uses simplified dose kernels, which give dose estimates quickly; however, the disadvantage is that some accuracy may be compromised. The aim of this article was to independently compare the VARSKIN (v2.1) skin dose module against Geant4 for typical skin contamination scenarios. Realistic skin contamination scenarios were modeled in Geant4 including droplet and surface contamination. Three droplet volumes (10, 30, and 50 μL) were considered as well as an infinitely thin disk source (A = 1 cm2) to model surface contamination. Four basal depths (70, 140, 220, and 370 µm) were considered in combination with two glove layers defined by 0.1 mm and 0.2 mm of natural rubber (ρ = 0.92 g cm-3) as well as an absent glove layer. The Geant4 models were then replicated as closely as possible using VARSKIN (v2.1) skin dose module. Forty-four (44) radionuclides were considered giving a total of 2,112 direct comparisons of instantaneous skin dose rates between VARSKIN (v2.1) and Geant4. Most data (86%) were within ±50% (relative) difference between VARSKIN and Geant4 with a median difference of only +4.0% for all data. However, the 95% confidence interval was wide at -29% to +210%, showing a large positive skew. This skewness is predominantly caused by large relative differences observed for photon emitting radionuclides including 57Co, 51Cr, 67Ga, 123I, 125I, 129I and 111In. However, the differences were low in absolute terms for these radionuclides. Despite large relative differences between VARSKIN and Geant4 for certain radionuclides and geometries, the differences were low in absolute terms. Overall, we conclude that VARSKIN (v2.1) is a reliable tool for skin dose assessment for the radionuclides and geometries studied in this article.