03:54 PM Abstract No. 369 Comparison of measured peak skin dose to scanner-reported dose indices during CT-guided interventions

Abstract

CT-guided interventional procedures, in particular ablations and complex orthopedic procedures, can at times require the use of substantial radiation doses. A theoretical model for skin dose based on the definitions of CTDIvol and CTIDw has been published, but not validated experimentally. The aim of this study was to validate this model and benchmark against a typical mix of helical and axial scanning during CT-guided procedures. 68 patients were enrolled in this IRB-approved study. Radiochromic film was calibrated using standard methods and used to measure patient skin doses for consecutive ablation procedures performed on a single CT scanner. Procedural dose indices, including CTDIvol, were collected and compared to the measured skin doses using a paired t-test and Bland-Altman analysis. CTDIvol separated by axial and helical imaging was used to estimate skin doses using a published theoretical model, which were compared to measured doses. All data was log transformed for hypothesis testing. When a theoretical model was applied to the data, predicted skin doses were significantly different from measured skin doses according to a two-tailed t-test (p < 0.0001). However, skin dose was not significantly different from uncorrected total CTDIvol according to a two-tailed t-test (p = 0.29). Bland-Altman analysis revealed a bias of -3.75 for percent difference (standard deviation [SD] = 24.0) with limits of agreement (LoA) of [-50.7, 43.2] between measured skin dose and CTDIvol (measured skin dose – CTDIvol) and a bias of -18.1 (SD = 22.4) with LoA of [-62.0, 25.7] between measured skin dose and skin dose as calculated using a theoretical model. CTDIvol alone was a better predictor of skin dose than dose estimated using a theoretical model. This is likely a consequence of how CTDI is measured, with scatter contributing much of the signal, rather than primary radiation. Also, patient positioning during CT-guided procedures varies substantially, and patients are often off-center, meaning the skin surface may not match well to the surface of a phantom centered in the gantry. CTDIvol is reported by all CT scanners and is a simple index for monitoring skin dose.