Elsevier

Physica Medica

Volume 60, April 2019, Pages 83-90
Physica Medica

Original paper
Volumetric modulated arc therapy planning based on virtual monochromatic images: Effect of inaccurate CT numbers on dose distributions

https://doi.org/10.1016/j.ejmp.2019.03.022Get rights and content

Highlights

  • CT numbers in the low energy VMI can be inaccurate for high density materials.

  • HU values in the VMI77keV is less affected by the scanning protocols than 120 kVp image.

  • The effect of the inaccurate CT numbers on PTV was more prominent in AXB than AAA.

  • Maximum dosimetric error in VMAT planning based on the VMI50keV was 0.5 Gy.

  • The dosimetric error due to the inaccurate HU estimation may be clinically insignificant.

Abstract

Purpose

Though virtual monochromatic images (VMIs) at low energy levels can improve image quality, the measured Hounsfield unit (HU) values can be inaccurate. We assessed the dosimetric error due to inaccurate HU estimation in volumetric modulated arc therapy (VMAT) planning.

Methods

Based on the VMIs at 50 keV (VMI50keV), 77 keV (VMI77keV) and single-energy CT (SECT) image for a phantom with different sizes, lookup tables (LUTL and LUTS) were created. Using an anthropomorphic phantom (head and spine regions), VMAT plans were generated based on VMI50keV, VMI77keV and SECT using the corresponding LUTL, and then, the doses were re-calculated using LUTS. For clinical cases, 30 VMAT plans (prostate, brain, and spine cases) were generated based on VMI50keV and VMI77keV.

Results

In the anthropomorphic phantom study, the difference in the dosimetric parameters for planning target volume (PTV) in the VMAT plan based on the VMI77keV was smallest (within 0.1 Gy) among three types of treatment planning approach. In clinical cases, in general, the differences of the 3-dimensional gamma passing rate and dosimetric parameters in the treatment plans based on the VMI50keV were larger than those in the VMI77keV. Especially for brain cases, the difference for PTV was more prominent when AXB was used (the maximum difference was 0.5 Gy) than AAA.

Conclusions

The dosimetric error due to the inaccurate HU estimation was larger in the VMIs at low energy levels. This may be clinically insignificant, but should be avoided in the VMAT treatment planning.

Introduction

In modern radiotherapy, the volumetric modulated arc therapy (VMAT) irradiation technique, which involves continuously varying the gantry rotation speed, multi-leaf collimator (MLC) pattern, and dose rate during delivery, has become increasingly common in clinical practice [1], [2], [3]. High doses to the target, while minimizing the doses to organs at risk (OAR) is to be expected of any highly conformal radiotherapy. For such sophisticated treatment techniques, computed tomography (CT) simulations play an important role in achieving accurate target delineation and evaluating doses for the targets as well as surrounding OARs.

Recent advancements in CT scanner technology have led to the development and clinical use of dual-energy CT (DECT) that utilizes two different photon beams for its operation; one of the advantage of the DECT scanner is that it can be used to generate virtual monochromatic images (VMIs) at a specified photon energy levels for both high and low photon energies [4]. Several researchers have demonstrated that VMIs obtained at low energy levels (40–60 keV) provided significantly higher image quality for various tumor sites compared with VMIs at high energy levels (70–77 keV), where the energy levels were equivalent to the conventional single polychromatic images (i.e., at 120 kVp) [5], [6], [7]; this improvement in image quality has the potential for improving inter- and/or intra-observer variations when determining gross tumor volume in radiotherapy treatment planning.

However, some previous studies showed that the CT numbers (Hounsfield unit; HU) in the case of VMIs at low energy level could be considerably inaccurate, especially for high-density materials; in addition, the accuracy of the HU values varied based on the volume of the surrounding tissues [8], [9]. In general, the CT numbers (Hounsfield unit, HU) are converted into electron and physical density values by using lookup tables (LUTs) for dose calculation in the radiotherapy treatment planning processes to account for inhomogeneous environments in patient anatomy. It is a common practice to register the LUT acquired using a commercial phantom for a corresponding CT scanner in a treatment planning system; thus, the use of VMIs obtained at low energy levels for VMAT planning might induce unexpected dosimetric errors during the dose calculation process because of the inaccurate estimation of HU values. To the best of our knowledge, no previous studies have investigated the effect of inaccurate HU estimation on dose distributions in VMAT planning.

Our study consisted of two components. First, we assessed the variance of HU measurements in the VMIs at 50 keV (VMI50keV; high contrast image) and 77 keV (VMI77keV; equivalent HU values for a conventional 120 kVp CT image) and single-energy CT (SECT) images with different phantom sizes. Second, we compared the effect of the inaccuracy of HU estimation for VMAT planning in an anthropomorphic phantom and patients.

Section snippets

Generation of LUTs

We used a 16.5-cm-deep multi-energy phantom (Gammex RMI 1472; Gammex RMI, Middleton, WI), which consisted of inner (20 cm in diameter) and outer (40 cm in width and 30 cm in height) sections composed of a water-equivalent material to simulate large and small patient sizes. The phantom contained cylindrical holes (2.85 cm in diameter) to enable the placement of various rods of the reference materials to be tested. In our study, 16 reference materials for a tissue characterization phantom (Gammex

Results

Fig. 2 shows the LUTs (electron density) generated from VMI50keV, VMI77keV and SECT image. In particular, for high-density materials (electron density > 1.1), VMI50keV provided higher HU values than VMI77keV and SECT. Furthermore, the differences in the absolute (relative) HU value in the case of VMI50keV between the large and small phantoms were 179.9 (14.5%), 240.4 (12.3%), and 389.2 (14.5%) HU for the reference materials of CB2-50%, SB3, and Al, respectively, while these differences in the

Discussion

In this study, we demonstrated that the HU values in VMI50keV and SECT image varied depending on the volume of the surrounding materials and the scanning parameters, especially in the case of high-density materials (Fig. 2, Fig. 3). Consequently, in our anthropomorphic phantom study, inaccurate HU estimation in the case of VMI50keV and SECT induced a larger dose deviation than the VMI77keV case (Fig. 4). In theory, the VMIs are less affected by the beam-hardening effect, which is recognized as

Funding

This study was supported by JSPS KAKENHI Grant (Grant-in-Aid for Young Scientists (B) 17K15816).

Conflict of interest

None.

References (16)

There are more references available in the full text version of this article.

Cited by (11)

  • Improvement of image quality for pancreatic cancer using deep learning-generated virtual monochromatic images: Comparison with single-energy computed tomography

    2021, Physica Medica
    Citation Excerpt :

    In further studies, image quality should be compared between deep-leaning-based VMIs and VMIs generated from various DECT systems for the same patient. Moreover, VMIs are less affected by the beam hardening effect, which induces inaccurate CT numbers and can improve the dose calculation accuracy [34]. In principle, deep-learning-based VMIs can be generated at a given energy level (i.e. 77 or 140 keV), and the generated VMIs have the potential for improving the dose calculation accuracy.

  • Stereotactic body radiation therapy planning for liver tumors using functional images from dual-energy computed tomography

    2020, Radiotherapy and Oncology
    Citation Excerpt :

    In our clinical practice, all image acquisitions were performed using a single-source DECT scanner with rapid kilovolt peak switching (Revolution HD; GE Medical Systems, Milwaukee, WI). Images reconstructed from the DECT have several advantages compared with the conventional single-energy CT images, including: (1) reduced the beam hardening effect [16]; (2) higher image quality [17]; and (3) minimal effect of the contrast-enhancing agent on dose distributions [18]. The DECT scanner switches the tube kilovoltage between 140 and 80 kVp in less than 0.5 ms, which enables the acquisition of the projection data with the high and low tube kilovoltage at very similar locations [11].

  • Dual-energy computed tomography image-based volumetric-modulated arc therapy planning for reducing the effect of contrast-enhanced agent on dose distributions

    2021, Medical Dosimetry
    Citation Excerpt :

    First, the number of patients was limited, and only NPC patients were investigated. Second, the VNC can be generated at an energy level of only 70 keV using a rapid kilovolt peak switching DECT system; several investigators considered the VMI at 77 keV to be equivalent to single-energy CT images.14,22 Third, there are various types of DECT systems, such as dual-source CT and dual-layer CT systems4, and a further study is required to verify whether the VNC can be an effective treatment planning approach for reducing the effect of CE agent on dose distributions.

View all citing articles on Scopus
View full text