A Monte Carlo investigation of cumulative dose measurements for cone beam computed tomography (CBCT) dosimetry

Many studies have shown that the computed tomography dose index (CTDI100) which is considered as a main dose descriptor for CT dosimetry fails to provide a realistic reflection of the dose involved in cone beam computed tomography (CBCT) scans. Several practical approaches have been proposed to overcome drawbacks of the CTDI100. One of these is the cumulative dose concept. The purpose of this study was to investigate four different approaches based on the cumulative dose concept: the cumulative dose (1) f(0,150) and (2) f(0,8) with a small ionization chamber 20 mm long, and the cumulative dose (3) f(100)(150) and (4) f(100)(infinity) with a standard 100 mm pencil ionization chamber. The study also aimed to investigate the influence of using the 20 and 100 mm chambers and the standard and the infinitely long phantoms on cumulative dose measurements. Monte Carlo EGSnrc/BEAMnrc and EGSnrc/DOSXYZnrc codes were used to simulate a kV imaging system integrated with a TrueBeam linear accelerator and to calculate doses within cylindrical head and body PMMA phantoms with diameters of 16 cm and 32 cm, respectively, and lengths of 150, 600, 900 mm. f(0,150) and f(100)(150) approaches were studied within the standard PMMA phantoms (150 mm), while the other approaches f(0,8) and f(100)(8) were within infinitely long head (600 mm) and body (900 mm) phantoms. CTDI infinity values were used as a standard to compare the dose values for the approaches studied at the centre and periphery of the phantoms and for the weighted values. Four scanning protocols and beams of width 20-300 mm were used. It has been shown that the f(0,infinity) approach gave the highest dose values which were comparable to CTDI infinity values for wide beams. The differences between the weighted dose values obtained with the 20 and 100 mm chambers were significant for the beam widths <120 mm, but these differences declined with increasing beam widths to be within 4%. The weighted dose values calculated within the infinitely long phantoms with both the chambers for the beam widths <= 140 were within 3% of those within the standard phantoms, but the differences rose to be within 15% at wider beams. By comparing the approaches studied in this investigation with other methodologies taking into account the efficiency of the approach as a dose descriptor and the simplicity of the implementation in the clinical environment, the f(0,150) method may be the best for CBCT dosimetry combined with the use of correction factors.