Technical note: No increase in effective dose from half compared to full rotation pelvis cone beam CT

Abstract Purpose To image the abdomen of a patient with a gantry mounted imaging system of a linear accelerator, different cone beam computed tomography (CBCT) protocols are available. The whole‐body dose of a full rotation abdomen CBCT and a half rotation CBCT was compared. In our clinic, both CBCT protocols are used in daily routine work. Methods With an adult anthropomorphic Alderson phantom, the whole‐body dose per CBCT scan was measured with thermoluminescence dosimeters. The half rotation CBCT was applied such that the gantry mounted X‐ray source rotated around the right side of the phantom. The 183 measurement locations covered all ICRP recommended critical organs (except the gonads). The effective dose was calculated with the mean organ dose and the corresponding tissue weighting factors. A point‐by‐point dose comparison of both protocols was conducted. Results The effective dose was 5.4 mSv ±5% and 5.0 mSv ±5% (estimated type B 1σ) for the full and the half rotation CBCT respectively. There was no significant difference (α = 0.05) in the effective dose within the precision of the measurement (1σ = 5%). The half rotation CBCT displayed an inhomogeneous dose distribution in a transversal phantom slice in contrast with the full rotation CBCT. In the imaging region, the mean dose was (20.5 ± 3.4) mGy and (19.2 ± 7.4) mGy (measured type A 1σ) for the full and the half rotation CBCT respectively. Conclusion The half compared to the full rotation CBCT displays a smaller field‐of‐view in a transversal slice and no significant difference in the effective dose. Hence, the full rotation CBCT is favorable compared to the half rotation CBCT. However, by using the half rotation protocol, critical volumes in the patient can be spared compared to the full rotation protocol.


Cone beam computed tomography (CBCT) is widely used in clinics
for patient positioning in a radiation treatment session. The advantage of image-guided radiation therapy (IGRT) is the application of more conformal plans and therefore, a reduction of the irradiated volume. The additional dose received by the patient 1 raises concerns about late effects such as second primary cancers. Hence, the quantification of CBCT dose is an important issue.
The TrueBeam linear accelerator (Varian Medical Systems, Palo Alto, CA, USA) is equipped with a gantry mounted on-board imager (OBI) capable of performing CBCT scans. To image the abdomen, the pelvis or the pelvis spotlight protocol is available. For the pelvis protocol, the CBCT is acquired with a full rotation of the X-ray source around the patient. The pelvis spotlight protocol uses a half rotation of the X-ray source around the patient. The weighted CT dose index (CTDI w ) given by Varian, is 1/3 lower for the half compared to the full rotation CBCT. Hence in clinical practice, the spotlight protocol is used more frequently because of the potential dose reduction compared to the full rotation protocol. The resulting spotlight CBCT has a smaller field-ofview in a transversal slice compared to the full rotation protocol.
In the last years, many studies have been conducted about CBCT dose for different protocols. [2][3][4][5][6][7][8][9] Some of them evaluated only the CTDI values. 3,9 Kan et al., 10 Cheng et al., 6 and H€ alg et al. 7 determined the effective dose (ED) for pelvis protocols using the OBI of a Varian linear accelerator. In another paper, 2 Monte Carlo (M.C.) dose calculations of different pelvis CBCT protocols were performed for real patient geometries.
With the OBI (version 2.5.28.0) of a TrueBeam linear accelerator, the absorbed dose for the latest pelvis and pelvis spotlight CBCT protocol was measured. Both CBCT protocols are used in-house in daily routine work. The 183 measurement locations in an anthropomorphic Alderson phantom were equipped with thermoluminescence dosimeters (TLDs). A combination of Li:Mg,Ti (TLD100) and Li:Mg,Cu,P (TLD100H) chips was used to automatically correct for the variation in response with radiation energy of the TLDs. 11 This allowed a more accurate determination of the whole-body absorbed dose (1r = AE5%) compared to previous studies. The ED values of both pelvis protocols were calculated by the determination of the mean doses to critical structures and the ICRP guidelines. 12 Furthermore, the absorbed dose of the pelvis and the pelvis spotlight CBCT was compared on a point-by-point basis.

| ME TH ODS AND MATERIALS
All measurements and detector readouts were performed according to a strict protocol to ensure the consistency of the measurements.

2.A | Imaging modality and whole-body dose measurement
The two evaluated CBCT pelvis protocols were provided by the vendor (see Table 1). For both protocols the X-ray source was operated at 125 kVp, which resulted in a mean photon energy of 64 keV. 9 The CBCT measurements were done at Hirslanden Medical Center in Aarau, Switzerland.
The absorbed dose of the CBCTs was measured using an adult  Table 1 and Fig. 2

2.B | TLD dose evaluation
The whole-body dose of the two CBCT protocols was measured with a combination of TLD100 and TLD100H chips. The two TLD types show a different response with radiation energy compare to 60 Co. 13 For the dose measurements, a TLD100H chip was put on top of a TLD100 chip. The doses measured by the TLD100 and TLD100H were evaluated by using individual calibration factors determined with a 6 MV nominal beam energy irradiation applied with a TrueBeam linear accelerator. The individual energy correction factor for TLD100 and TLD100H was found by the ratio of the TLD100 divided by the TLD100H dose. Hence, each single TLD was corrected with a specific energy correction factor displaying a random error. The finale dose was calculated with the mean of the corrected TLD100 and TLD100H dose. A detailed description of the TLD dose and energy measurement is given by Hauri et al. 11 All T A B L E 1 Acquisition parameters for the two different kV CBCT protocols given by Varian (version 2.5.28.0). In our clinic, both protocols are used in daily routine work in the current form.

Pelvis
Pelvis spotlight  ED ¼ X

2.C | Dose comparison
where w R is the radiation weighting factor with w R ¼ 1 for photon irradiations. D T;R is the mean absorbed dose from radiation R in tissue T. w T is the tissue weighting factor for tissue T with P T w T ¼ 1.
R is the equivalent dose for tissue T. Furthermore, the absorbed dose of the two different CBCT protocols was evaluated by a point-by-point comparison.

| RESULTS
The estimated type B standard deviation (1r) for a CBCT dose point measurement was AE5%. A more detailed error analysis can be found in Hauri et al. 11 In Table 2 the effective organ doses are displayed for both CBCT protocols. Furthermore, the number of measurement points to calculate the mean organ dose is shown. The ED was determined to 5.4 mSv AE5% and 5.0 mSv AE5% (type B 1r) for the full and the half rotation CBCT respectively. The 95% confidence interval of the difference between the two ED values contained zero (0.4 AE 0.7 mSv).
Hence, there was no significant difference (a = 0.05) of the ED values between the two protocols within the precision of the measurement (AE5%).  The ICRP weighting factors wT 12 and the determined equivalent dose HT of the organ T for both CBCT protocols (OBI version 2.5.28.0). Furthermore, the number of measurement points to calculate the mean absorbed organ dose is shown. The mean red bone marrow and bone surface doses were determined from the same 33 measurement locations. The effective dose for the pelvis and the pelvis spotlight was calculated to (

| DISCUSSION
With the OBI of a TrueBeam Varian linear accelerator, a full rotation and a half rotation CBCT protocol (Table 1)  should not be projected to a pediatric patient, since the absorbed dose in a young patient is likely to be higher. 8 The higher dose is caused by less attenuation of the X-ray beam in the pediatric compared to the adult patient. Furthermore, in an adolescent patient critical structures are closer to the imaging region compared to an adult patients.
In the image region, a dose reduction by a factor of ten for critical points can be achieved by using the half compared to the full rotation CBCT. To accomplish a dose reduction for a critical volume, the starting angle of the half rotation CBCT acquisition is crucial [see Fig. 2(b)]. Nevertheless, there is no significant reduction in ED for the half compared to the full rotation CBCT.
According to Gardner et al., 4 the contrast-to-noise ratio of the pelvis compared to the spotlight protocol is significantly higher.
Hence, the full rotation CBCT provides better soft tissue contrast compared to the half rotation protocol.
The ED is a function of the mean organ dose. Therefore, the ED values depend on the distribution of the measurement location in the phantom. This study represents a conservative estimation of the ED since 1/5 of the TLDs were distributed in the imaging region representing 1/10 of the body. The type B uncertainty of the dose measurement (1r = AE5%) was propagated to the ED. This is rather an overestimation of the ED uncertainty since H T was calculated by the mean of multiple measurement locations ( Table 2).
The higher photoelectric crosssection of bones compared to tissue is reported to cause a few times higher dose to bones then to nearby tissue. 2,5 For the TLDs distributed in the human bony material of the Alderson phantom used in this work, no higher dose was notice compared to the surrounding soft tissue. Hence, the equivalent dose to bones could be underestimated.

CONCLUSI ON
The half compared to the full rotation CBCT displays a smaller fieldof-view and no significant difference in the ED. Hence, the full rotation CBCT is favorable compared to the half rotation CBCT. However, by using the half rotation protocol critical points in the patient can be spared compared to the full rotation protocol.

ACKNOWLEDGMENT
This work was funded by the grant KFS-3249-08-2013 from the Swiss Cancer League.

CONFLI CT OF INTEREST
The authors declare no conflict of interest.