Patient dose estimation in cardiac computed tomography with a 320 detector row scanner based on Monte Carlo simulation

s / Physica Medica 30 (2014) e16ee44 e28 used, where the latter two were considered to represent infinitely long head and body phantoms, respectively. The efficiencies were estimated as the ratios of CTDIIEC and D(0) in 150 mm long phantoms to CTDI∞ calculated within the infinitely long phantoms. Beam widths studied ranged from 60 to 300 mm. Results: For beams of width 80 mm, the CTDIIEC,w was more efficient in terms of recording a greater proportion of the absorbed radiation, but for wider beams >80 mm the D(0)w was more efficient. The ratios for CTDIIEC,w were approximately constant over all beam widths, where the values for the head and body phantoms were 82.2 ± 0.9% and 75.7 ± 0.7%, respectively. The ratios for D(0)w in the head and body phantoms increased from 73% to 94% and 71% to 92%, respectively, with beam width as the beam extended beyond the phantom edges, respectively. Conclusion: As most CBCT machines utilize a beamwidth >80 mm, results suggest that the D(0) approach is the better option for CBCT dosimetry. This involves fewer measurements than CTDIIEC,w and is more practical for regular use in the medical environment. PATIENT DOSE ESTIMATION IN CARDIAC COMPUTED TOMOGRAPHY WITH A 320 DETECTOR ROW SCANNER BASED ON MONTE CARLO SIMULATION M. Cros , J. Geleijns , R.M.S. Joemai , I. Hern andez-Gir on , A. Calzado , M. Dewey , M. Salvad o . Unitat de Fisica M edica, Universitat Rovira i Virgili (URV), Spain; Radiology Department, Leiden University Medical Center (LUMC), The Netherlands; Departamento de Radiologia, Universidad Complutense de Mardrid (UCM), Spain; Department of Radiology, Charite. Universit€ atsmedizin Berlin, Germany Purpose: To estimate organ doses and effective dose for CT calcium scoring, CT coronary angiography and CT myocardial perfusion according to the internationally recognized recommendations of the ICRP. Materials and methods: Monte Carlo simulations based on the EGS4 code were used to model the Aquilion ONE CT scanner and the two ICRP computational phantoms representing the standard female and male patient models. Dose calculations were performed for the cardiac CT acquisition protocol used in a multicenter study for the assessment of the calcium score, coronary arteries CTA and CT myocardial perfusion (CORE320). All acquisitions were performed as axial volumetric scans, covering the entire heart in a single (low heart rate 65 bpm) or dual (high heart rate 66 bpm) heart beats. This was done by positioning the phantoms in the isocenter (common practice in MC calculations) and by centering their hearts to the rotation axis (clinical practice). The effect of removing the arms from the phantom was also evaluated since the ICRP phantoms have their arms stretched along the body whereas in clinical practice the arms are removed from the scanned volume. Results and discussion: Effective dose averaged for females and males was 1.9 mSv for coronary calcium score; 5.1 mSv for coronary CTA including rest cardiac perfusion at low heart rates, 9.7 mSv at high heart rates; and 4.3mSv for stress cardiac perfusion at low heart rates, 8.5 mSv at high heart rates. No significant differences in effective dose were observed for centering versus off-centering for both phantoms. Removing the arms from the scan region led to an increase of effective dose of about 6% to 8%. For coronary CTA and rest cardiac perfusion and using the most relevant configuration (phantom off-centered without arms) at low heart rate, the highest organ doses averaged for females and males were 16 mGy in breast, 15.5 mGy in heart and 11 mGy in lung. The k-factors used in other studies led to underestimations of effective dose by a factor of 2 and 3 for an average male and for an average female patient, respectively, compared to this study results. A NEW OPTICAL PHOTON TRANSPORT MONTE CARLO CODE FOR MODELLING PARALLELAND FOCUSED-ELEMENT SCINTILLATION DETECTOR ARRAYS AND ITS USE FOR EXAMINATION OF THE FULL MTF RESPONSES OF THICK SEGMENTED CSI(TL) SCINTILLATORS Mohammad Amin Mosleh-Shirazi , Zinat Zarrini-Monfared , Sareh Karbasi , Ali Zamani . a Physics Unit, Department of Radiotherapy and Oncology, Iran; Department of Medical Physics and Engineering, Shiraz University of Medical Sciences, Shiraz, Iran Purpose: Arrays of thick segmented crystalline scintillators are considered useful x-ray converters in electronic portal imaging and cone-beam megavoltage CT for radiotherapy verification. They consist of a 2D matrix of scintillators separated by optically opaque, reflective septal walls and can offer a high quantum efficiency and acceptable spatial resolution. However, it has been shown using ionizing-radiation-only simulations, that in such parallel-detector-element arrays, obliquely incident x-rays reduce modulation transfer function (MTF) performance and degrade spatial resolution in peripheral areas. A focused geometry has therefore been proposed. The full performance of such a focused geometry is still unclear. The aim of this work was to investigate the imaging performance of such arrays using a detector model that supported light transport, as well as x-rays and electrons. Methods: To simulate x-ray and electron transport, we used the MCNP Monte Carlo code to model the required array geometries added to a validated model of a 6 MV linac head. Then, the transport of the generated optical photons wasmodelled using ScintSim2, an optical Monte Carlo code we wrote in MATLAB for simulation of paralleland focused-element scintillation detector arrays. We present ScintSim2 and report the results of using it to examine the full (ionizing and optical radiation) MTF responses of focusedand parallel-element geometries, for a large array of 3 3 mm2 CsI(Tl) elements with thicknesses of 10, 40 and 60mm. A slit geometry was simulated to obtain the line spread function (LSF) of the energy deposited in the elements (radiation-only) or the number of optical photons incident on an underlying screen (full simulation). Then, for each case, a composite LSF was calculated and Fourier transformed to obtain the MTF. Results: Differences were observed between the radiation-only and full MTF performances of the arrays. At the Nyquist frequency, focused elements provided an increase of up to 8 times in peripheral-area full MTF values. Further, the light exiting thicker scintillators exhibited a more forward-directed distribution. Conclusion: The observed differences in the results obtained with and without optical photon simulation justify the additional effort of including light transport when optimizing these promising thick segmented detectors.