Evaluation of the relationship between phantom position and computed tomography dose index in cone beam computed tomography when assuming breast irradiation

Abstract This study aims to investigate the influence of the phantom position on weighted computed tomography dose index (CTDIw) in cone beam computed tomography (CBCT) when assuming breast irradiation. Computed tomography dose index (CTDI) was measured by the x‐ray volume imaging of CBCT using parameters for image‐guided radiation therapy (IGRT) in right breast irradiation. The measurement points of CTDI ranged from 0 (center) to 16 cm in the right–left (RL) direction, and from 0 (center) to 7.5 cm in the anterior–posterior (AP) direction, which assumed right breast irradiation. A nonuniform change exists in the relative value of CTDIw when the phantom deviated from the isocenter of CBCT. The CTDIw was ~30% lower compared with the value at the isocenter of CBCT when the phantom deviated 7.5 and 16 cm at the AP and RL directions, respectively. This study confirmed the influence of the phantom position on the CTDI values of CBCT. The CTDI measured at the isocenter of CBCT overestimates that measured at the irradiation center of the breast.


| INTRODUCTION
In recent years, image-guided radiation therapy (IGRT) with cone beam computed tomography (CBCT) using an electronic portal imaging device has been extensively used in radiotherapy. Many studies have reported that IGRT with CBCT is useful for improving setup errors and observing changes in tumor volume. [1][2][3][4] IGRT can compensate for position misalignment by capturing images before irradiation. However, because CBCT uses radiation, its radiation exposure may cause cancer and other health problems. Several reports exist that correlate CBCT doses with cancer risks. [5][6][7][8][9][10][11] Therefore, knowing and maintaining the IGRT dose as low as possible is essential. One of the dose indexes for CBCT is CT dose imdex (CTDI). Generally, the center of the torso is adjusted to the gantry isocenter in diagnostic CT, and the CTDI is measured by placing a phantom at the gantry isocenter. In CBCT for IGRT, the center of the patient torso may not be at the isocenter of CBCT, for example, the isocenter of CBCT is located within the planning target volume and the center of the trunk is not adjusted to the isocenter of CBCT in the breast irradiation. Therefore, the center of the torso of the patient does not coincide with the center of CBCT. No detailed report exists on the influence of phantom location when measuring CTDI although certain studies 12,13 are available on dosimetry of CBCT for IGRT. ICRP Publication 103 dictates that the effective dose can be easily estimated by multiplying the dose length product (DLP) using the conversion factor. 14 DLP is calculated from weighted CTDI (CTDI w ) and longitudinal direction irradiation range. In breast irradiation, the center of CBCT is at the center of the radiotherapy site, and the center of the torso is not at the center of CBCT. Estimating the effective dose from DLP derived from CTDI w measured at the isocenter of CBCT may not be suitable in this situation because it is not the radiation dose received at the actual location. This study examined the influence of phantom positioning on CTDI values when assuming breast irradiation.

2.A | X-ray volume imaging CBCT system
The CTDI of CBCT using x-ray volume imaging (XVI) was measured.
The XVI is a device attached to the Elekta Synergy (Elekta, Stockholm, Sweden). The system comprises an x-ray tube and a flat panel detector. These components are perpendicular to the gantry of the linear accelerator. Moreover, the XVI is available for CBCT, fluoroscopy, and radiography.

2.B | Phantom and dosimeter
A cylindrical acrylic phantom was used for the CTDI measurement.
This phantom had a diameter of 32 cm, a longitudinal length of 15 cm, a hole in the center, and four holes (top, bottom, left, and right) in the periphery that were located 1 cm inside the surface of the phantom.
The dosimeters used were ACCU-GOLD+ (Radcal, Monrovia, CA, USA) and an ionization chamber (10X6-3CT: Radcal) that has an active length 10 cm. Figure 1 shows the dosimeters used in this study.

2.C | Acquisition parameters
This study measured the CTDI of CBCT at the location that assumed right breast irradiation. One of the methods for reducing breast dose from CBCT in IGRT is to irradiate the x rays from the back. In the system used in this study, the collimator of the gantry head, which is placed vertically to CBCT imaging system, protrudes toward the isocenter. Thus, the collimator may interfere with the patient which limits the methods of dose reduction compared with the left breast irradiation when CBCT is performed on the right breast by irradiating x rays from the back of the patient. In the facility of the first author, CBCT is not performed by irradiating x rays from the back of the patient for right breast irradiation in consideration of patient safety. Therefore, the dose of CBCT in right breast irradiation is expected to be higher compared with the left breast irradiation.
Thus, CBCT dose was investigated when assuming right breast irradiation.
The acquisition parameters were a tube voltage of 100 kV, a tube current of 20 mA per frame, 5.5 frames per second collection, and irradiation angle of XVI from 265°to 110°with clock-wise direction. The collimator was S20, which yielded an axial field-ofview of 27 cm, and no additional filter was attached (F0). The maximum diameter of the reconstruction was 270 mm, and the longitudinal x-ray beam width was 276.7 mm.

2.E | CTDI w quantification
The measured air kerma was multiplied by the calibration factor and atmospheric correction factor to obtain CTDI 100 (CTDI 100 ) values.
CTDI 100 is the value measured by inserting a pencil-type ionization chamber dosimeter with a length of 100 mm into the measurement hole of the phantom for CTDI measurement. The calibration factor was obtained by comparing it with a dosimeter that was traceable to a Japanese national standard dosimeter. The measurement at each point was performed thrice, and the mean value of these measurements was used as the measured value. The extended uncertainty of the calibration factor was 6%, and the CTDI 100 was calculated from the following equation: where K a is the air kerma, k TP is the atmospheric correction factor, and N is the calibration factor of the ionization chamber. | 263 CTDI w was calculated from the following equation: where CTDI 100,c is the CTDI 100 at the center of the phantom and CTDI 100,p is the average of CTDI 100 at the peripheral of the phantom. CTDI w measurements were repeated by displacing the phantom along with the RL and AP directions from the isocenter of CBCT.
The CTDI 100 relative value against the value measured at the isocenter of CBCT at each measurement point was calculated using the following equation: Relative value of CTDI 100 % ð Þ¼ðCTDI 100 =CTDI 100, iso Þ Â 100, where CTDI 100,iso is the CTDI 100 measured by placing the phantom at the isocenter of CBCT. The relative CTDI w value was then calculated by displacing the phantom against the value measured at the isocenter of CBCT by using the following equation: where CTDI w,iso is the CTDI w measured by placing the phantom at the isocenter of CBCT.
where K ai is the initial air kerma, (μ en /ρ) breast is the mass energy absorption coefficient for breast (50% glandular tissue and 50% adipose tissue), and (μ en /ρ) air is the mass energy absorption coefficient for air. 15 3 | RESULTS Figure 4 shows the CTDI 100 at the measurement hole in the center of the phantom, the average of the CTDI 100 at the four measurement holes at the peripheral, and the CTDI w . The positive values in RL direction mean the phantom was moved from the isocenter to right direction, and those in AP direction mean the phantom was moved from the isocenter to posterior direction. The coefficient of variation of CTDI w measured at the isocenter of CBCT was 0.57%.
Nonuniform changes in CTDI w were observed as the phantom moved from the isocenter of CBCT. The relative values of CTDI w changed from 69.5% to 100.9% as the phantom deviated from the isocenter of CBCT along both the 90°and 180°directions. When the phantom deviated >5 cm or >10 cm along the AP and RL directions from the isocenter of CBCT, the relative value of CTDI w reduced by >10% compared with that obtained at the isocenter of CBCT. The CTDI w was~30% lower than that obtained at the isocenter of CBCT when the phantom deviation was the largest, which was 7.5 and 16 cm at the AP and RL directions, respectively. The relative doses of CTDI w were reported to be different according to adjusted to the isocenter. 5,8,9,19,20 It is believed that the effective dose can be more precisely estimated by multiplying the relative value of CTDI w by the CTDI w measured at the isocenter of CBCT although CTDI is the radiation dose output of a CT scanner and cannot be used to compare patient dose.
Alvarado et al. 21 reported organ doses when the center of the breast and the center of the trunk were located at the isocenter of CBCT, and the absolute dose of right breast were reduced by 10 to 23% when the center of the breast was located at the isocenter of CBCT compared with those when the center of the trunk was located at the isocenter of CBCT.
Their results showed similar trend with our results that the absorbed dose for the right breast decreased as the distance from the center of the phantom torso center and the isocenter of CBCT was enlarged, which were similar to CTDI results.
Note that this study has several limitations. The results of this study are for a single acquisition parameter for a specific device.
Moreover, the results obtained in this study depend on the dose

| CONCLUSION
The measured CTDI w , which assumes breast irradiation, decreases by~30% when the phantom deviates from the isocenter of CBCT compared with that measured at the isocenter of CBCT. The relative values of CTDI w can be used as correction factors to estimate the CTDI w of the actual measurement location that improves the estimation accuracy of effective dose when the patient is not set at the isocenter of CBCT.
This study showed that the dose of CBCT in IGRT should reflect the patient's position. This may enable understanding the dose more accurately in actual situations.

AUTHOR CONTRI BUTIONS
Hiroyuki Ueno was involved in conceptualization, data curation, investigation, visualization, and writing-original draft. Kosuke Matsubara and Akihiro Takemura were involved in supervision and writing-review and editing. Masato Hizume and Sayuri Bou were involved in writing-review and editing.

CONF LICT OF I NTEREST
The authors declare no conflict of interest.

D A T A A V A I L A B I L I T Y S T A T E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.