Clinical implementation of Dosimetry Check™ for TomoTherapy® delivery quality assurance

Abstract Purpose The delivery quality assurance (DQA) of intensity‐modulated radiotherapy (IMRT) plans is a prerequisite for ensuring patient treatments. This work investigated the clinical usefulness of a new DQA system, Dosimetry Check™(DC), on TomoTherapy®‐based helical IMRT plans. Methods The DQA was performed for 15 different TomoTherapy®‐based clinical treatment plans. In Tomotherapy® machines, the couch position was set to a height of 400 mm and the treatment plans were delivered using QA‐Treatment mode. For each treatment plan, the plan data and measured beam fluence were transferred to a DC‐installed computer. Then, DC reconstructed the three‐dimensional (3D) dose distribution to the CT images of the patient. The reconstructed dose distribution was compared with that of the original plan in terms of absolute dose, two‐dimensional (2D) planes and 3D volume. The DQA results were compared with those performed by a conventional method using the cheese phantom with ion chamber and radiochromic film. Results For 14 out of the 15 treatment plans, the absolute dose difference between the measurement and calculation was less than 3% and the gamma pass rate with the 3%/3 mm gamma evaluation criteria was greater than 95% for both DQA methods. The P‐value calculated using Wilcoxon signed‐rank test was 0.256, which implies no statistically significance in determining the absolute dose difference between the two methods. For one treatment plan generated using the 5.0 cm field width, the absolute dose difference was greater than 3% and the gamma pass rate was less than 95% with DC, while the DQA result with the cheese phantom method passed our TomoTherapy® DQA tolerance. Conclusion We have clinically implemented DC for the DQA of TomoTherapy®‐based helical IMRT treatment plans. DC carried out the accurate DQA results as performed with the conventional cheese phantom method. This new DQA system provided more information in verifying the dose delivery to patients, while simplifying the DQA process.


| INTRODUCTION
Modern radiotherapy has become more complicated in its quest to deliver a highly conformal dose to a defined target volume, while sparing organs at risk near the target volume. The TomoTherapy ® (Accuray, Sunnyvale, CA, USA) is one of the modern radiotherapy systems allowing a continuous dose delivery in a helical fashion around the anatomical site to be treated. The quality assurance (QA) of dose delivery using TomoTherapy ® is a prerequisite for ensuring patient treatments.
For TomoTherapy ® -based intensity-modulated radiotherapy (IMRT), the current delivery quality assurance (DQA) process consists of comparing measured versus calculated doses in a phantom using an ionization chamber and a film 1,2 or using detector array devices. [3][4][5][6] These devices have generally provided such accurate DQA results in terms of low absolute dose difference or high gamma pass rate. However, these DQA processes are also time-consuming and laborious to TomoTherapy ® users since a separate DQA plan corresponding to each treatment plan needs to be created. Moreover, heavy devices are required on the treatment couch of the TomoTherapy ® system to perform the DQA measurements. It must be underlined that the DQAs performed with the above-mentioned methods allow the dose comparison between the measurement and calculation only in the small region or the plane, where the measurement tools (i.e., the ionization chamber, film, detector array, etc.) are positioned, and require much work in cases where the volume to be treated extends beyond their physical dimensions. Therefore, with the current TomoTherapy ® DQA modalities, it is hard to identify the region accurately inside the patient body where non-negligible dose difference between the measurement and calculation is present.
Recently, various new systems have been commercially released for DQA of patient treatment plans using modern radiotherapy modalities. 5,7 These systems use log-files of the beam irradiation or measured beam fluence to reconstruct the dose distribution on the CT images of patients, thereby reconstructing the dose to the target and surrounding normal structures. These systems make the DQA analysis available not only in a point dose and two-dimensional (2D) planar dose distribution but also in three-dimensional (3D) volumetric dose distribution inside the patient body. Therefore, users can compare the dose distribution between the measurement and calculation in more detail compared with the traditional DQA methods.
The purpose of this work is to investigate the clinical suitability of a new commercial DQA system, Dosimetry Check™(DC, MathResolutions, LLC., Columbia, MD, USA), on TomoTherapy ® -based IMRT plans. The DQA tests have been performed for TomoTherapy ®based clinical helical treatment plans covering various treatment sites. For the same treatment plans, the DQA process was also car- ried out using the traditional cheese phantom method. We compared the DQA results obtained with these two different methods in absolute dose difference and gamma pass rate of 2D planar dose distribution.

2.A | TomoTherapy ®
The TomoTherapy ® unit is designed to provide intensity-modulated radiotherapy delivery with flattening-filter free 6 MV photon beam and binary 64 multileaf collimators (MLCs). 8 In our clinic, two different TomoTherapy ® units were used for patient treatments: TomoTherapy ® HD and TomoTherapy ® Hi-ART. TomoTherapy ® HD provides both helical and TomoDirect™ modes, while the TomoTherapy Hi-ART provides only helical mode. These two TomoTherapy ® units clinically used three different field widths of 1.0, 2.5, and 5.0 cm, which were defined by jaws along the longitudinal direction.
TomoTherapy ® -based IMRT treatment plans are created by its own integrated treatment planning system (TPS). 2 The TomoTherapy ® TPS provides inverse planning capability in the optimization process and determines the leaf positions for all the gantry angles and couch positions. The inverse planning process is carried out until all the dose constraints are satisfied or have been optimized. The final dose calculation is performed with a convolution/superposition algorithm.

2.B | Dosimetry Check™
DC (version 5.2.4) is a software that carries out DQA by reconstructing 3D dose distribution on the CT images of a phantom or patient. 9 For the DQA of TomoTherapy ® treatment plans, DC uses the measured beam fluence, i.e., the sinogram, as the radiation source for the dose reconstruction. The beam fluence is recorded by the TomoTherapy ® MVCT detector positioned in opposite side to the linear accelerator/target. The current DC software reconstructs the 3D dose distribution based on pencil-beam (PB) algorithm or collapsed-cone convolution (CCC) algorithm, while it only used PB algorithm in the previous version. In this work, we reconstructed the dose distribution in DC using CCC algorithm with a 5 mm grid size.
DC carries out the DQA analysis using two different modes: pretreatment dosimetry mode and in vivo dosimetry mode. In the pretreatment dosimetry mode, no material is present inside the treatment unit bore except for the treatment couch. In the in vivo dosimetry mode, the phantom or patient is positioned inside the treatment bore. In this work, we tested the pretreatment dosimetry mode for the clinical application of DC. Since the in vivo dosimetry mode in DC was not commissioned in our institution, we excluded to test this mode in this work.

2.C | Dose delivery verification
Before using it for clinical DQA, it is required to evaluate whether DC performs the dose reconstruction accurately as measured by the ionization chamber or not. By performing this process, we could reflect the output of the TomoTherapy ® treatment units to the DCbased dose reconstruction. For this test, we used a TomoTherapy ®based IMRT delivery verification plan, which was created on a cylindrical Solid Water™ phantom (i.e., the cheese phantom) by the TomoTherapy ® factory and used in acceptance test procedure (ATP) during the TomoTherapy ® treatment machine installation. Three different IMRT verification plans were created using the 1.0, 2.5, and 5.0 cm field widths. Each verification plan was generated to deliver uniform dose to the cylindrical target positioned at the center of the phantom as shown in Fig. 1. The prescription dose to the cylindrical target was 10 Gy in five fractions, where 95% of the target volume receiving at least 10 Gy.
The dose distribution to the cheese phantom with each verification plan was reconstructed using DC in the pretreatment dosimetry mode. The actual dose delivered to the cylindrical phantom was also measured using an Exradin A1SL air-

2.D | Patient-specific DQA for TomoTherapy ®based helical IMRT plans
The patient-specific DQA test with DC was performed for 15 different TomoTherapy ® -based helical treatment plans, which were randomly selected from our institution's patient list. Seven treatment plans were delivered on the TomoTherapy ® HD unit and other eight treatment plans were delivered on the TomoTherapy ® Hi-ART unit. The treatment sites of these clinical plans were prostate, brain, head and neck, lung, abdomen and spine. Most of the treatment plans were generated using the 2.5 cm field width except for one abdomen treatment plan, which was created using the 5.0 cm field width.
In the TomoTherapy ® treatment machine, a couch position was set to a height of 400 mm and each plan was delivered using QA-Treatment mode. The treatment plan was archived and the recorded beam fluence from the TomoTherapy ® MVCT detector was saved as .xml format. Since the operating software version of our TomoTherapy ® units did not support the DC automation system, the beam flu-  The DQA process for each treatment plan was also performed

3.B | Patient-specific DQA results for
TomoTherapy ® -based helical IMRT plans  TomoTherapy ® -based IMRT plans. [13][14][15][16][17][18] These studies reported that the DC-calculated 3D volumetric gamma pass rates for the patientspecific DQAs were greater than 90%. Mezzenga et al. 16 reported is necessary likely to the cheese phantom method, thereby reducing the DQA preparation time and requiring less human resources. In a treatment room, only couch positioning and beam irradiation time slot is required, which is around 10-15 min per treatment plan including the measured beam fluence data transfer to a DC-installed computer. Including the dose reconstruction processing time, we could complete overall process of the DC-based DQA analysis within 1 h per treatment plan. Even though automation program of DC was not used in this work because of the technical issue, we expect that the automation program will accelerate the overall DQA process since the measured beam fluence transfer as well as dose reconstruction process can be automatically performed.
In this work, since DC installed in our institution was primarily used for the DQA before the patient treatment, it was only commissioned on the pretreatment dosimetry mode. As mentioned by Mezzenga et al. 16 and Gimeno et al., 18  Currently, DC is used as the primary resource of the DQA for TomoTherapy ® -based helical IMRT plans in our institution.

CONF LICT OF I NTEREST
Conflict of interest relevant to this article was not reported.