LAUSR

Introduction Prolonged exposures in interventional radiology can lead to increased skin doses (Dskin) and to determinist side effects. The increasing frequency of Rotational mode (RM) acquisitions implies significant Dskin rises. The Dskin calculation method for a fluorography can be performed manually or with dedicated software. However, currently, such methods do not exist for RM. The aim of this study is to evaluate the feasibility of Dskin calculation with the PCXMC 2.0 Rotation software (STUK) for RM in interventional neuroradiology, then to validate this estimation by using a manual method combined with PCXMC use and fluorography. Methods This study is performed for an Allura Xper FD20/10 (Philips) system. The available RM are the XperCT (207°/rotation, 622 images) and the Propeller (240°/rotation, 122 images). For these acquisitions, mean dose is calculated by PCXMC for each organ, but this software is not able to obtain a reliable Dskin, which is averaged over the whole phantom skin. The objective is to determine the possible existence of a factor to convert mean organ dose during the acquisition to Dskin. Considering neuroradiology interventions, the mean brain dose (Dbrain) was chosen as convertible to Dskin for different points on the skin. Hence several measurements were carried out on an anthropomorphic ART phantom (Radiology Support Devices Inc.) with Thermo Luminescent Dosimeters (TLDs) which were calibrated at radiological energies and with a homogeneity below 5%. In a second phase, a manual calculation based method was created to estimate Dskin with a dedicated software which cannot consider RM: a trigonometric estimation was performed to evaluate the number of projections per rotation that could contribute to Dskin at each point. Fluorographies were then simulated on PCXMC for these points, based on each corresponding number of projections. Then, Dbrain for each fluorography was converted in RM Dskin. Experimental TLDs measurements were carried out in order to validate this method. Results For the XperCT, the posterior Dskin measured with TLDs is 34,53 mGy, 2,67 mGy for the Propeller. For these RM, the first Dskin estimations from Dbrain lead to the Dskin measured with TLDs within a 20% accuracy interval. Conclusion In addition to the Dskin determination with PCXMC for the two RM in interventional neuroradiology, its estimation with the fluorography conversion method seems possible.