Medical Dosimetry ] (2016) ]]]–]]]

Medical Dosimetry
journal homepage: www.meddos.org

Dosimetric comparison of lung stereotactic body radiotherapy

treatment plans using averaged computed tomography and
end-exhalation computed tomography images: Evaluation of the effect
of different dose-calculation algorithms and prescription methods
Takamasa Mitsuyoshi, M.D., Mitsuhiro Nakamura, Ph.D., Yukinori Matsuo, M.D., Ph.D.,
Nami Ueki, M.D., Ph.D., Akira Nakamura, M.D., Ph.D., Yusuke Iizuka, M.D., Ph.D., Wambaka
Ange Mampuya, M.D., Ph.D., Takashi Mizowaki, M.D., Ph.D., and Masahiro Hiraoka, M.D., Ph.D.
Department of Radiation Oncology and Image-Applied Therapy, Graduate School of Medicine, Kyoto University, Kyoto, Japan

A R T I C L E I N F O A B S T R A C T

Article history: The purpose of this article is to quantitatively evaluate differences in dose distributions calculated using
Received 8 January 2016 various computed tomography (CT) datasets, dose-calculation algorithms, and prescription methods in
Received in revised form
stereotactic body radiotherapy (SBRT) for patients with early-stage lung cancer. Data on 29 patients with
13 April 2016
early-stage lung cancer treated with SBRT were retrospectively analyzed. Averaged CT (Ave-CT) and
Accepted 16 July 2016
expiratory CT (Ex-CT) images were reconstructed for each patient using 4-dimensional CT data. Dose
distributions were initially calculated using the Ave-CT images and recalculated (in the same monitor
Keywords:
units [MUs]) by employing Ex-CT images with the same beam arrangements. The dose-volume
Early lung cancer
parameters, including D95, D90, D50, and D2 of the planning target volume (PTV), were compared
Stereotactic body radiotherapy
CT dataset for dose calculation between the 2 image sets. To explore the influence of dose-calculation algorithms and prescription
Dose-calculation algorithm methods on the differences in dose distributions evident between Ave-CT and Ex-CT images, we
calculated dose distributions using the following 3 different algorithms: x-ray Voxel Monte Carlo
(XVMC), Acuros XB (AXB), and the anisotropic analytical algorithm (AAA). We also used 2 different dose-
prescription methods; the isocenter prescription and the PTV periphery prescription methods. All
differences in PTV dose-volume parameters calculated using Ave-CT and Ex-CT data were within
3 percentage points (%pts) employing the isocenter prescription method, and within 1.5%pts using the
PTV periphery prescription method, irrespective of which of the 3 algorithms (XVMC, AXB, and AAA) was
employed. The frequencies of dose-volume parameters differing by 41%pt when the XVMC and AXB
were used were greater than those associated with the use of the AAA, regardless of the dose-
prescription method employed. All differences in PTV dose-volume parameters calculated using Ave-CT
and Ex-CT data on patients who underwent lung SBRT were within 3%pts, regardless of the dose-
calculation algorithm or the dose-prescription method employed.
& 2016 American Association of Medical Dosimetrists.

Introduction
Lung cancer is the leading cause of cancer-related deaths
worldwide, and the number of cases of early-stage lung cancer is
This research was supported in part by a Grant-in-Aid for Scientific Research
from the Ministry of Education, Culture, Sports, Science, and Technology, Japan
(Grant no. 25253078).
Reprint requests to Mitsuhiro Nakamura, Ph.D., Department of Radiation
Oncology and Image-applied Therapy, Graduate School of Medicine, Kyoto Uni-
versity, 54 Shogoin-Kawaharacho, Sakyo-ku, Kyoto 606-8507, Japan.
E-mail: m_nkmr@kuhp.kyoto-u.ac.jp
expected to increase.1 The standard of care for patients with early-
stage lung cancer is surgical resection; however, some patients are
medically unable to undergo any type of surgery because of
advanced age or the presence of comorbidities.2,3 Recently, stereo-
tactic body radiotherapy (SBRT), which features delivery of hypo-
fractionated radiotherapy using a stereotactic reference system,
has come to play an increasingly important role as a nonsurgical
treatment for early-stage lung cancer.4 Several prospective multi-
center clinical trials of SBRT have reported high control rates and
favorable outcomes (comparable to those of surgery).5,6 In multi-
center clinical trials, it is wise to standardize treatment processes,

http://dx.doi.org/10.1016/j.meddos.2016.07.003
0958-3947/Copyright Ó 2016 American Association of Medical Dosimetrists
2 T. Mitsuyoshi et al. / Medical Dosimetry ] (2016) ]]]–]]]
including the dose-prescription method and the dose-calculation
algorithm, among institutions.
Several studies have evaluated the effects of different dose-
prescription methods and dose-calculation algorithms on radia-
tion dose distributions and dose-volume histograms (DVHs).7-9
However, few studies have explored the effects on DVHs of using
different computed tomography (CT) images in dose calcula-
tions.10,11 Free breathing (FB) expiratory phase CT (Ex-CT) and
averaged CT (Ave-CT) images derived from 4-dimensional CT
(4D-CT) data are commonly used in clinical practice to perform
dose calculations, but the optimal image dataset for such work
remains unclear. Both image sets have advantages and disadvan-
tages. Ave-CT images embrace all respiratory phases. However,
motion-blurring artifacts may influence the calculated dose dis-
tributions. Ex-CT images, on the contrary, restrict the acquisition of
CT data to a single phase of breathing, thereby reducing motion
artifacts. However, respiration-driven tumor motion is not taken
into account.
We sought to quantitatively assess differences in dose distri-
butions and DVH data obtained using Ave-CT and Ex-CT images for
dose calculations, with the beam arrangements and monitor units
(MUs) held constant. In addition, we evaluated the effects of
different dose-calculation algorithms and dose-prescription meth-
ods on differences in dose distributions calculated using the
2 types of images.
Methods
Patients and tumor characteristics
Twenty-nine patients with solitary lung tumors who under-
went SBRT were included in the analysis. Tumors were located in
the upper lobe in 7 patients, in the middle lobe in 3, and in the
lower lobe in 19. The respiration-driven amplitudes of tumor
motion were fluoroscopically measured before treatment. The
median amplitude was 15.1 mm (range: 3.5 to 46.1 mm). In 2 of
the 29 cases, the motion amplitude was 4 40 mm.
4D-CT and target delineation
The details of the SBRT planning and treatment processes used
in our institution have been described previously.12,13 Each 4D-CT
scan was obtained using a Varian Real Time Position Management
Respiratory Gating System (ver. 1.7; Varian Medical Systems, Palo
Alto, CA, USA) and a LightSpeed 16 RT CT Scanner with 16 rows of
detectors (General Electric Medical Systems, Waukesha, WI, USA);
the slice thickness was 2.5 mm in the axial cine mode. The tube
voltage and current were set to 120 kV and 100 mA, respectively.
The scan length was that of the entire lung. The gantry rotation
time was 0.7 s. An interscan delay of 2 s was selected to prevent
marker vibration caused by couch movement. Next, 4D-CT slice
and respiratory motion data were transferred to an Advantage 4D
Workstation (General Electric Medical Systems) and imported to
an iPlan RT Image System (Brainlab AG, Feldkirchen, Germany).
The 4D-CT datasets were reconstructed by reference to the
respiratory phase and then binned into 10 respiratory cycle phases
(0% to 90%, with 0% representing maximum inspiration). Ex-CT
images were defined as images taken at the midpoints between
consecutive inhalation peaks. Ave-CT images were generated by
averaging 10-phase CT datasets.
Each internal target volume (ITV) was determined using
maximum intensity projection (MIP) datasets derived from
4D-CT images with reference to a 10-phase CT image dataset.
When the delineated ITV did not adequately embrace the extent of
tumor motion (as revealed fluoroscopically), the ITV was manually
expanded based on x-ray fluoroscopy evaluation.12 Each PTV was
determined by adding a uniform margin of 5 mm to the ITV, to
allow for setup uncertainties and errors in mechanical accuracy.
The tumor characteristics of the PTVs are shown in Table 1. The
Ave-CT values of, and the PTVs derived using, Ave-CT data were
lower than those obtained when Ex-CT data were employed in
26 patients.
Field setup, dose prescription, and dose-calculation algorithms
We used 6 to 8 noncoplanar static x-ray beams of 6 MV. All
plans were calculated using 2 different dose-prescription meth-
ods: isocenter and PTV periphery prescription. When employing
the isocenter prescription method, we added a leaf margin of
5 mm to the PTV. When using the PTV periphery prescription
method, each leaf margin was arranged to fit the 70% isodose line
to the 90% level of the PTV (the acceptance criterion was 88% to
92%), and the dose received by 2% of the PTV was required to be o
107%.
Dose distributions were calculated using 3 different algorithms.
The first was the XVMC running iPlan RT Dose (ver. 4.5.3; Brainlab
AG) and featuring heterogeneity correction. The grid size and the
mean variance were set to 2.3
2.3
2.5 mm3 and 2%, respec-
tively. The second was the Acuros XB (AXB), and the third the
anisotropic analytical algorithm (AAA) with a grid size of 2.5
2.5

2.5 mm3; both algorithms are implemented in the software of
the Eclipse treatment planning system (ver. 11.0.31; Varian Medical
Systems). Dose distributions were initially calculated using Ave-CT
images employing the 2 prescription methods and all of the XVMC,
AXB, and AAA (thus, 6 calculations for each case). Next, dose
distributions were recalculated using Ex-CT data, the same target
volumes, the same beam arrangements, and the same fields
and MUs.
Dose-volume histogram evaluation
The differences in PTV dose distribution and DVH data evident
when Ave-CT and Ex-CT data were used were measured in each
case. The DXX indicates the dose received by XX% of the PTV (thus,
D95, D90, D50, and D2). We also evaluated the effects of tumor size,
motion amplitude, and CT target value on differences in dose
distributions by calculating Pearson's product-moment correlation
coefficients. Differences in dose-volume parameters were com-
pared via analysis of variance.
Results
We calculated the differences in PTV dose-volume parameters
(D95, D90, D50, and D2) by subtracting the parameters derived using
Ave-CT data from those obtained employing Ex-CT data collected
Table 1
Tumor characteristics. Data are shown as medians (with ranges) or as numbers of
patients
Median Number or (range)
Tumor location (upper/middle/lower) 7/3/19
Amplitude (mm) 15.1 (3.5 to 46.1)
PTV
Volume (cm3) 64.7 (15.6 to 115.7)
CT value
Ave-CT image (HU) 521 (777 to 228)
Ex-CT image (HU) 506 (768 to 166)
Ave-CT image ¼ image generated by averaging 10 computed tomography images;
Ex-CT image ¼ image taken at the midpoint between consecutive inhalation peaks;
HU ¼ Hounsfield unit.
 T. Mitsuyoshi et al. / Medical Dosimetry ] (2016) ]]]–]]] 3

Table 2
Means, standard deviations, minima, and maxima of differences in PTV dose-volume parameters calculated using the XVMC, AXB, and AAA, and the isocenter prescription
(A) and PTV periphery prescription (B) methods

XVMC AXB AAA

Mean SD Min Max Mean SD Min Max Mean SD Min Max

(A) Isocenter prescription

△D95 (%pts) 0.3 0.8 2.1 1.7 0.5 0.7 2.4 0.6 0.3 0.6 2.1 0.4
△D90 (%pts) 0.1 1.2 2.7 1.8 0.3 0.6 2.3 0.5 0.1 0.6 2.1 0.3
△D50 (%pts) 0.1 1.1 2.9 1.8 0.3 0.5 1.7 0.5 0.2 0.4 1.3 0.3
△D2 (%pts) 0.0 0.9 2.3 1.2 0.2 0.3 0.9 0.6 0.1 0.4 1.3 0.8

(B) PTV periphery prescription

△D95 (%pts) 0.1 0.6 1.1 1.3 0.2 0.5 1.1 0.6 0.1 0.3 0.5 0.5
△D90 (%pts) 0.0 0.5 1.0 0.8 0.1 0.4 1.1 0.5 0.0 0.3 0.5 0.5
△D50 (%pts) 0.0 0.6 1.0 1.2 0.0 0.5 1.4 0.8 0.0 0.3 0.8 0.7
△D2 (%pts) 0.1 0.6 1.3 1.2 0.3 0.4 1.4 0.3 0.1 0.4 0.8 1.4

max ¼ maximum; min ¼ minimum; SD ¼ standard deviation. △DXX indicates the DXX calculated using Ex-CT data minus the DXX calculated employing Ave-CT data.

using the same beam arrangements and MUs. Table 2 shows the distribution to a tumor associated with 46.1 mm of motion
means, standard deviations, minima, and maxima of dose-volume amplitude is shown in Fig. 2.
parameter differences in the 29 cases, calculated using either the We calculated correlation coefficients between the differences
isocenter or PTV periphery prescription method. When the former in dose-volume parameters on the one hand, and tumor sizes,
method was employed, the mean differences of all evaluated motion amplitudes, and target CT values on the other hand. The
parameters ranged from 0.5%pts to 0%pts, regardless of the only factor correlating strongly with the observed differences was
algorithm used. The standard deviations of the XVMC data were the amplitude of motion; the correlation coefficients are shown in
significantly greater than those of the AXB and AAA figures (p o Table 4. When the AAA and the isocenter prescription method
0.05) (Table 2A). When the PTV periphery prescription method were used, a strong correlation was evident only between the PTV
was used, the differences in dose-volume parameters and their D50 and motion amplitude (R ¼ 0.72). The isocenter prescription
standard deviations were smaller than those obtained employing method tended to robustly correct the dose-volume parameters
the isocenter prescription method (Table 2B). The standard devia- for motion amplitude (more so than did the PTV periphery
tions associated with use of the XVMC were significantly larger prescription method). The absolute values of the correlation
than those of the AXB and AAA figures (p o 0.05). Regardless of coefficients were o 0.32 and o 0.52 in terms of tumor sizes and
the dose-calculation algorithm used, the differences in dose- target CT values, respectively.
volume parameters were within 3%pts and 1.5%pts when the
isocenter and PTV periphery prescription methods, respectively,
were employed. The differences in the PTV D90 calculated using Discussion
the isocenter and PTV periphery prescription methods to evaluate
Tumor location and shape can vary during treatment, both
XVMC-derived data were not significant, but the differences
because of breathing motion and anatomical changes.14 The
obtained when the AXB and AAA were used were significant
accuracy of a calculated dose distribution is thus affected, as is
(p o 0.05).
treatment efficacy. Correct assessment of such changes and appro-
Of the 29 cases, the frequencies of dose-volume parameter
priate adaptation of treatment is crucial if radiotherapy is to be
differences Z 1%pt are shown in Table 3. Such differences were
safe and effective. Using 4D-CT imaging, respiration-driven lung
more common when the isocenter rather than the PTV periphery
tumor motion can be incorporated into dose calculations.15,16 The
prescription method was used, regardless of the algorithm
doses absorbed during various stages of breathing can be
employed. The frequencies of such differences in terms of the
algorithm used were in the order XVMC, AXB, and AAA, regardless
of the method of prescription.
Figure 1 shows a histogram of the PTV D90 differences when
the PTV periphery prescription was used. The differences were
within 0.5%pts in 20, 25, and 28 cases when the XVMC, AXB, and
AAA, respectively, were employed. Only a single difference 4 1%pt
was evident when AXB was used. An example of the dose

Table 3
The frequencies of D95, D90, D50, and D2 PTV differences over 1%pt

Isocenter prescription PTV periphery prescription

XVMC AXB AAA XVMC AXB AAA

△D95 6 7 2 3 1 0
△D90 14 5 3 1 1 0 Fig. 1. A histogram showing differences in planning target volume (PTV) D90s.
△D50 11 5 2 5 2 0 Dose-volume data were calculated using the x-ray Voxel Monte Carlo (XVMC),
△D2 1 0 1 3 1 1 Acuros XB (AXB), and anisotropic analytical algorithm (AAA), and the PTV
periphery prescription method. Each difference in D90 was calculated as the D90
△DXX indicates the DXX calculated using Ex-CT data minus the DXX calculated derived from expiratory computed tomography (Ex-CT) data minus that derived
employing Ave-CT data. from averaged CT (Ave-CT) data.
4 T. Mitsuyoshi et al. / Medical Dosimetry ] (2016) ]]]–]]]

Fig. 2. Dose distributions calculated using the XVMC, AXB, and AAA, and data from Ave-CT and Ex-CT images, shown in the sagittal direction. (A) XVMC (Ave-CT), (B) XVMC
(Ex-CT), (C)AXB (Ave-CT), (D) AXB (Ex-CT), (E) AAA (Ave-CT), and (F) AAA (Ex-CT). The colored lines are isodose lines. (Color version of figure is available online.)

individually computed, summed, and calculated for a single images obtained during FB, MIP images, and Ave-CT images, on
breathing phase using deformable registration imaging. However, patients with lung cancer undergoing SBRT; the dose-calculation
the method is complex and is seldom used in clinical practice. algorithm employed was the AAA. The dose-volume parameters
Instead, other 4D-CT-derived image sets (Ave-CT or Ex-CT) are derived using Ave-CT images were similar to those obtained with
routinely employed in dose calculations. The optimal image set the aid of FB images. Compared to Ave-CT images, MIP images may
remains controversial. either over- or underestimate the target volume when the target lies
A few studies have explored the effects of using different CT close to a high-density tissue, such as the liver. The cited authors
datasets in dose calculations on dose distributions and DVH patterns. concluded that Ave-CT images were suitable for use in both planning
Yamazaki et al.10 used a superposition algorithm to measure variations and dose calculation when lung SBRT was contemplated. In this study,
in calculated doses (caused by respiration) in patients who underwent in addition to Ave-CT images, we used Ex-CT images as the single-
lung SBRT. The total dose variation at the isocenter was within ⫾ 2%pts phase datasets instead of mid-ventilation CT (MidV-CT). Indeed, MidV-
when CT images taken at inhalation and exhalation were compared. CT is defined as the breathing phase, where the tumor is located
This is clinically important. However, recently, dose prescription has closest to the mathematical time-weighted mean tumor position
come to be defined as the dose covering the PTV. It is thus necessary obtained by averaging the 10 tumor positions derived from all phases
to consider not only the dose at the isocenter but also the dose-volume of the 4D-CT.17 Based on this definition, we assumed that the degree of
parameters of the entire PTV, and in great detail. Tian et al.11 compared motion artifacts in MidV-CT is between Ave-CT and Ex-CT. Differences
the dosimetric characteristics of treatment plans calculated using in the dose-volume parameters of PTVs obtained using Ex-CT and
Ave-CT images were within 3%pts in this study (Table 2). We,
therefore, think that the differences would have been within the same
Table 4 range if MidV-CT were used.
Correlation coefficients between motion amplitudes and differences in dose-
volume parameters calculated using the XVMC, AXB, and AAA, and the isocenter
Our study differs from others in that we evaluated differences
prescription and PTV periphery prescription methods in dose-volume parameters using 3 dose-calculation algorithms
and 2 dose-prescription methods, and compared the results
Isocenter prescription PTV periphery prescription
obtained. It is well known that the choice of dose-calculation
algorithm affects dose distribution.7,8 It is of clinical importance to
XVMC AXB AAA XVMC AXB AAA
further evaluate differences in the dose-volume parameters
△D95 0.13 0.64 0.57 0.11 0.30 0.49 obtained when various algorithms are used to evaluate Ave-CT
△D90 0.35 0.60 0.55 0.24 0.37 0.51 and Ex-CT images.
△D50 0.42 0.59 0.72 0.20 0.43 0.48 Turning to the dose-calculation algorithms employed, differ-
△D2 0.19 0.13 0.49 0.52 0.54 0.16
ences in dose-volume parameters Z 1%pt were most frequently
△DXX indicates the DXX calculated using Ex-CT data minus the DXX calculated observed in the order XVMC, AXB, and AAA. This is because the
employing Ave-CT data. AAA (compared to XVMC and AXB) tends to underestimate the
 T. Mitsuyoshi et al. / Medical Dosimetry ] (2016) ]]]–]]]
doses required by areas that differ greatly in density, such as the
PTV periphery.7 Also, the variance of the XVMC was a contributory
factor.18 The mean variance was set to 2% in the present study. This
means that differences in dose-volume parameters calculated
using XVMC may be larger than those obtained with the aid of
AXB or the AAA.
The most important finding of the present study is that all
differences in dose-volume parameters (D95, D90, D50, and D2)
attributable to the use of 2 different CT datasets for dose calcu-
lations were within 3%pts regardless of the dose-calculation
algorithm and dose-prescription method employed. Motion ampli-
tude influenced such differences, especially when the isocenter
prescription method was employed. The CT datasets used for dose
calculation in multicenter trials of lung cancer SBRT have not yet
been defined. Differences in dose distributions caused by the use
of various CT datasets during dose calculations should certainly be
considered, but are unlikely to be large.
Conclusions
All differences in the dose-volume parameters (D2, D50, D90,
and D95) of PTVs obtained using Ex-CT and Ave-CT images, in dose
calculations during lung SBRT planning, were within 3%pts,
regardless of the dose-calculation algorithm (XVMC, AXB, or
AAA) used or the dose-prescription method (isocenter or PTV
periphery) employed.
References
1. National Lung Screening Trial Research Team. Results of initial low-dose
computed tomographic screening for lung cancer. N. Engl. J. Med.
368:1980–91; 2013.
2. Ginsberg, R.J.; Rubinstein, L.V. Randomized trial of lobectomy versus limited
resection for T1 N0 non-small-cell lung cancer. Lung cancer study group. Ann.
Thorac. Surg. 60:615–22; 1995.
3. NCCN Clinical Practice Guidelines in Oncology website. Non-small cell lung cancer,
version 2.2013. Available at: http://www.nccn.org/professionals/physician.
5
4. Chi, A.; Liao, Z.; Nguyen, N.P.; et al. Systemic review of the patterns of failure
following stereotactic body radiation therapy in early-stage non-small-cell lung
cancer: clinical implications. Radiother. Oncol. 94:1–11; 2010.
5. Timmerman, R.; Paulus., R.; Galvin, J.; et al. Stereotactic body radiation therapy
for inoperable early stage lung cancer. J. Am. Med. Assoc. 303:1070–6; 2010.
6. Nagata, Y.; Hiraoka, M.; Shibata, T.; et al. Stereotactic body radiation therapy for
T1N0M0 non-small cell lung cancer: first report for inoperable population of a
phase II trial by Japan Clinical Oncology Group (JCOG 0403). Int. J. Radiat. Oncol.
Biol. Phys. 84:S46; 2012.
7. Tsuruta, Y.; Nakata, M.; Nakamura, M.; et al. Dosimetric comparison of Acuros
XB, AAA, and XVMC in stereotactic body radiotherapy for lung cancer. Med.
Phys. 41:081715; 2014.
8. Mampuya, W.A.; Matsuo, Y.; Nakamura, A.; et al. Differences in dose-volumetric
data between the analytical anisotropic algorithm and the x-ray voxel Monte
Carlo algorithm in stereotactic body radiation therapy for lung cancer. Med.
Dosim. 38:95–9; 2013.
9. Oku, Y.; Takeda, A.; Kunieda, E.; et al. Analysis of suitable prescribed isodose line
fitting to planning target volume in stereotactic body radiotherapy using
dynamic conformal multiple arc therapy. Pract. Radiat. Oncol. 2:46–53; 2012.
10. Yamazaki, R.; Onimaru, R.; Katoh; et al. Influence of respiration on dose
calculation in stereotactic body radiotherapy of the lung. Radiol. Phys. Technol.
7:284–9; 2014.
11. Tian, Y.; Wang, Z.; Ge, H.; et al. Dosimetric comparison of treatment plans based
on free breathing, maximum, and average intensity projection CTs for lung
cancer SBRT. Med. Phys. 39:2754–60; 2012.
12. Nakamura, M.; Narita, Y.; Matsuo, Y.; et al. Geometrical differences in target
volumes between slow CT and 4D CT imaging in stereotactic body radiotherapy
for lung tumors in the upper and middle lobe. Med. Phys. 35:4142–8; 2008.
13. Matsugi, K.; Nakamura, M.; Miyabe, Y.; et al. Evaluation of 4D dose to a moving
target with Monte Carlo dose calculation in stereotactic body radiotherapy for
lung cancer. Radiol. Phys. Technol. 6:233–40; 2013.
14. Keall, P.J.; Mageras, G.S.; Balter, J.M.; et al. The management of respiratory
motion in radiation oncology: report of AAPM Radiation Therapy Committee
Task Group No. 76. Med. Phys. 33:3874–900; 2006.
15. Keall, P. 4-dimensional computed tomography imaging and treatment plan-
ning. Semin. Radiat. Oncol. 14:81–90; 2004.
16. Rietzel, E.; Chen, G.T.Y.; Choi, N.C.; et al. Four-dimensional image-based treat-
ment planning: target volume segmentation and dose calculation in the
presence of respiratory motion. Int. J. Radiat. Oncol. Biol. Phys. 61:1535–50;
2005.
17. Oechsner, M.; Odersky, L.; Berndt, J.; et al. Dosimetric impact of different CT
datasets for stereotactic treatment planning using 3D conformal radiotherapy
or volumetric modulated arc therapy. Radiat. Oncol. 249:1–10; 2015.
18. Fippel, M.; Laub, W.; Huber, B.; et al. Experimental investigation of a fast Monte
Carlo photon beam dose calculation algorithm. Phys. Med. Biol. 44:3039–54;
1999.