Variability of the discrepancy between manufacturer and measured CTDI100 values by scanner type, acquisition parameters and phantom size

Abstract

According to the EU directive 2013/59, the computed tomography (CT) equipments shall be able to inform the practitioner of parameters for assessing the patient dose. The aim of this study was to measure the accuracy of the CTDI₁₀₀ with respect to the values stipulated in the manufacturer manual for different manufacturers/models of CT and to assess the impact of acquisition parameters on CTDI₁₀₀ accuracy. Ten CT from 4 different manufacturers were included in the study. The discrepancy with the manufacturer value CTDI₁₀₀ was checked with a dosimeter with traceable calibration. Manufacturer values for the CTDI₁₀₀ were derived from the equipment's manual. The impact of the kVp, CT model, primary collimation and phantom size on the discrepancy of the CTDI₁₀₀, was assessed by a four-way ANOVA. All the factors had a statistically significant impact on CTDI₁₀₀ discrepancy (P < 0.05). In a head to head comparison, the greater discrepancies were found on average for 80 kV (15.4 ± 10.7%), sixteen channels CT (10.3 ± 5.1%) and for thick collimations (13.2 ± 6.8%), whilst no significant differences were found between head and body phantoms. For sixteen channels CT and acquisition protocols involving the use of low kV or thick primary collimation, the discrepancies with the manufacturer value CTDI₁₀₀ can be higher than 20%, which is the suspension level indicated in the EC RP N.162 Publication. This suggests the need of individual calibration of CT X-ray tubes by the manufacturers and the necessity of including this check in the quality control programs for CT.

Keywords: CTDI(100) accuracy; Computed tomography; Dosimetry.