Influence of electron density spatial distribution and X‐ray beam quality during CT simulation on dose calculation accuracy

Ahmad Nobah,Nada Tomic,Belal Moftah,Slobodan Devic

doi:10.1120/jacmp.v12i3.3432

Abstract

Impact of the various kVp settings used during computed tomography (CT) simulation that provides data for heterogeneity corrected dose distribution calculations in patients undergoing external beam radiotherapy with either high‐energy photon or electron beams have been investigated. The change of the Hounsfield Unit (HU) values due to the influence of kVp settings and geometrical distribution of various tissue substitute materials has also been studied. The impact of various kVp settings and electron density (ED) distribution on the accuracy of dose calculation in high‐energy photon beams was found to be well within 2%. In the case of dose distributions obtained with a commercially available Monte Carlo dose calculation algorithm for electron beams, differences of more than 10% were observed for different geometrical setups and kVp settings. Dose differences for the electron beams are relatively small at shallow depths but increase with depth around lower isodose values.PACS numbers: 87.57.Q‐, 87.55.D‐

Highlights

81 Nobah et al.: Electron density spatial distribution errors if the heterogeneity corrections are used when the treatment planning system converts the Hounsfield Unit (HU) to electron density through a stored calibration curve.The impact of electron density (ED) variation with computed tomography (CT) acquisition parameters has been reported for high-energy photon beams in the literature by investigating its dosimetric effect by using different heterogeneous phantoms.[9,10] Cozzi at al.[9] reported on the impact of an average HU to electron density calibration table provided by the manufacturer on computed doses when a custom calibration could not be included in the treatment planning system (TPS)
Since the real patient geometry may not be quite similar to distribution of tissue equivalent substitutes within the HU to ED calibration phantom, we investigated the impact of the geometrical arrangement of electron density plugs within the RMI phantom on possible dose errors for both high-energy electron and photon radiotherapy beams. (It should be noted that some other commercially available treatment planning systems use physical density instead of electron density for their heterogeneity corrections and Monte Carlo simulations.) Since the results reported in this work compare relative dose distribution values at selected points and, due to the fact that there is a fairly linear relation between physical and electron density,† conclusions reached in this paper could be extended to the dose calculations using other treatment planning systems
We report on the impact of the various kVp settings used during CT simulation, which provides data for accurate heterogeneity corrected dose distributions in patients undergoing external beam radiotherapy with either high-energy photon or electron beams

Summary

Introduction

81 Nobah et al.: Electron density spatial distribution errors if the heterogeneity corrections are used when the treatment planning system converts the HU to electron density through a stored calibration curve.The impact of electron density (ED) variation with CT acquisition parameters has been reported for high-energy photon beams in the literature by investigating its dosimetric effect by using different heterogeneous phantoms.[9,10] Cozzi at al.[9] reported on the impact of an average HU to electron density calibration table provided by the manufacturer on computed doses when a custom calibration could not be included in the treatment planning system (TPS). Guan at al.[10] investigated variations of dose per Monitor Unit (MU) versus different CT scanning parameters (kV, mAs) and different HU–ED curves for photon beams They concluded that the dose/MU varies more with different kV and mAs for small field size 18 MV beams (smaller than 10 cm × 10 cm) just beyond the high-density bones. A smaller kV imaging beam quality will generate higher HU, which leads to a higher effective atomic number (Zeff) for given voxels. This will, in turn, result in a larger pair production fraction of the 18 MV treatment photon beam inside bone. Guan et al concluded that the dose/MU from different HU–ED curves agrees to within 2%

Methods

Results

Discussion

Conclusion