Physical validation of a Monte Carlo-based, phantom-derived approach to computed tomography organ dosimetry under tube current modulation

Abstract

Purpose: To physically validate the accuracy of a Monte Carlo-based, phantom-derived methodology for computed tomography (CT) dosimetry that utilizes organ doses from precomputed axial scans and that accounts for tube current modulation (TCM).

Methods: The output of a Toshiba Aquilion ONE CT scanner was modeled, based on physical measurement, in the Monte Carlo radiation transport code MCNPX (v2.70). CT examinations were taken of two anthropomorphic phantoms representing pediatric and adult patients (15-yr-old female and adult male) at various energies, in which physical organ dose measurements were made using optically stimulated luminescence dosimeters (OSLDs). These exams (chest-abdomen-pelvis) were modeled using organ dose data obtained from the computationally equivalent phantom of each anthropomorphic phantom. TCM was accounted for by weighting all organ dose contributions by both the relative attenuation of the phantom and the image-derived mA value (from the DICOM header) at the same z-extent (cranial-caudal direction) of the axial dose data.

Results: The root mean squares of percent difference in organ dose when comparing the physical organ dose measurements to the computational estimates were 21.2, 12.1, and 15.1% for the uniform (no attenuation weighting), weighted (computationally derived), and image-based methodologies, respectively.

Conclusions: Overall, these data suggest that the Monte Carlo-based dosimetry presented in this work is viable for CT dosimetry. Additionally, for CT exams with TCM, local attenuation weighting of organ dose contributions from precomputed axial dosimetry libraries increases organ dose accuracy.

Keywords: Monte Carlo; anthropomorphic phantom; computed tomography; organ dose; tube current modulation.

Figure 1

Stylized computational phantoms with equivalent cross‐sectional areas and below their normalized AV _average values. (Row 1, varying outer body shape with equivalent area. Row 2, varying inner organ shape with equivalent area. Row 3, varying inner organ size). [Color figure can be viewed at wileyonlinelibrary.com]

Figure 2

The 15‐yr‐old female anthropomorphic phantom assembled on the CT scanner bed. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 3

Comparison of the mA distribution (WF multiplied by Average mA) for each of the three dosimetry methods on the UF15F phantom (120 kVp exam shown). Note: Smaller Z values correspond to the top (cranial–caudal direction) of the phantom.

Figure 4

Comparison of the mA distribution (WF multiplied by Average mA) for each of the three dosimetry methods on the UFADM phantom (120 kVp exam shown). Note: Smaller Z values correspond to the top (cranial–caudal direction) of the phantom.