CT dose estimation using Gafchromic XR-CT film: Models comparison Ying-Lan Liao a , Hui-Chuan Kao b , Keh-Shih Chuang a , Chih-Ping Chen c , Hui-Yu Tsai b, * a Department of Biomedical Engineering and Environmental Science, National Tsing Hua University, Hsin-Chu 300, Taiwan b Department of Medical Imaging and Radiological Sciences, Chang Gung University, 259, Wen-Hwa 1st Road, Kwei-Shan, Taoyuan 333, Taiwan c Atomic Energy Council, New Taipei city 234, Taiwan article info Article history: Received 5 November 2010 Received in revised form 12 July 2011 Accepted 21 August 2011 Keywords: Radiochromic film Gafchromic XR-CT Dose response Energy dependence Radiation dosimetry abstract Previous studies have established four dose response models for Gafchromic films. However, there have been no reports comparing the four dose response models for XR-CT film; and the XR-CT film responses in phantoms at different positions remain unclear. The purpose of this study was to compare the dose response models and to determine the most suitable one for XR-CT film. The correction factors for the energy dependencies of films inside the head and body dosimetry phantoms were assessed. The most appropriate dose response model for Gafchromic XR-CT film was D film ¼ a (netPVk) 2 þ b (netPVk). The XR-CT films could be used to measure CTDI values in phantoms with appropriate correction factors. Moreover, the doses estimated from XR-CT films used in the Taiwan CT survey agreed with the data measured using the ionization chamber. The Gafchromic XR-CT film could be used to measure CTDI values with the appropriate dose response model and correction factors. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction The CT dose index (CTDI) is the most commonly used dose descriptor for CT. It represents the integral of the dose profile along a line perpendicular to the tomographic plane divided by the product of the nominal tomographic section thickness and the number of tomograms produced in a single axial scan (Shope et al., 1981). It can be measured with various dosimeters, such as, ioni- zation chambers, thermoluminescent dosimeters (TLD) (Tsai et al., 2003), silicon photodiodes (Mori et al., 2005), and radiochromic films (Gorny et al., 2005; Rampado et al., 2010). Several types of Gafchromic film from International Specialty Products with high sensitivity for the diagnostic energy range have been investigated extensively for use in dose measurement and verification (Gorny et al., 2005; Rampado et al., 2006, 2010). The XR-CT film is designed for measurement of CT beam width. The dose response and energy dependence model of XR-CT film have been established with reflective optical density (ROD) conversion model (Butson et al., 2006a, 2006b, 2008). Several other studies (Blair and Meyer, 2009; Rampado et al., 2006, 2010) have established three dose response models for XR-QA film and XR-RV2 film depending on the differences of pixel values. However, the dose response model for XR-CT film with pixel value conversion for CT scan conditions was not assessed. The purpose of this study was to compare the dose response models used in previous studies for XR-CT film, determine the most suitable dose response model, and assess the energy dependence free in air and the film responses in the CTDI head and body phantoms at central and peripheral positions. In addition, the film doses of the Taiwan CT survey were obtained with the dose response model. 2. Materials and methods 2.1. Dose response models Gafchromic Ò XR-CT film (International Specialty Products, Wayne, NJ) was used to estimate the CT dose free in air and in phantoms. Four dose response models were used to convert the XR- CT film responses to doses. The first model (Rampado et al., 2010) was D film ¼ aðnetPV  kÞ 2 þbðnetPV  kÞ; (1) where the net pixel value (netPV) was the ratio of the pixel value of the unexposed film to that of the exposed film (PV unexp /PV exp ) multiplied by the k factor of 28,000. The ratio of PV unexp /PV exp ranged from 1 to 2, corresponding to doses from 0 to 97.5 mGy; so the value of 28,000 was chosen for the term k. The second model (Rampado et al., 2006) was D film ¼ cðexp½netPV=k  1 1Þ; (2) * Corresponding author. Tel.: þ886 3 2118800x3619. E-mail address: hytsai@mail.cgu.eud.tw (H.-Y. Tsai). Contents lists available at SciVerse ScienceDirect Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas 1350-4487/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.radmeas.2011.08.018 Radiation Measurements 46 (2011) 2052e2055