Practical Radiation Oncology (2014) 4, e67–e73

www.practicalradonc.org

Original Report

Accuracy and precision of cone-beam computed tomography guided intensity modulated radiation therapy Matthew W. Sutton MS a , Jonas D. Fontenot PhD a,b,⁎, Kenneth L. Matthews II PhD a , Brent C. Parker PhD a,b , Maurice L. King MD b , John P. Gibbons PhD a,b , Kenneth R. Hogstrom PhD a,b a

Department of Physics and Astronomy, Louisiana State University and Agricultural and Mechanical College, Baton Rouge, Louisiana b Mary Bird Perkins Cancer Center, Baton Rouge, Louisiana Received 10 December 2012; revised 15 January 2013; accepted 18 February 2013

Abstract Purpose: To assess the accuracy and precision of cone-beam computed tomography (CBCT)guided intensity modulated radiation therapy (IMRT). Methods and Materials: A 7-field intensity modulated radiation therapy plan was constructed for an anthropomorphic head phantom loaded with a custom cassette containing radiochromic film. The phantom was positioned on the treatment table at 9 locations: 1 “correct” position and 8 “misaligned” positions along 3 orthogonal axes. A commercial kilovoltage cone-beam computed tomography (kV-CBCT) system (VolumeView, Elekta AB, Stockholm, Sweden) was then used to align the phantom prior to plan delivery. The treatment plan was delivered using the radiation therapy delivery system (Infinity; Elekta AB) 3 times for each of the 9 positions, allowing film measurement of the delivered dose distribution in 3 orthogonal planes. Comparison of the planned and delivered dose profiles along the major axes provided an estimate of the accuracy and precision of CBCT-guided IMRT. Results: On average, targeting accuracy was found to be within 1 mm in all 3 major anatomic planes. Over all 54 measured dose profiles, the means and standard errors of the displacement of the center of the field between the measured and calculated profiles for each of the right-left, anterior-posterior, and superior-inferior axes were + 0.08 ± 0.07 mm, + 0.60 ± 0.08 mm, and + 0.78 ± 0.16 mm, respectively. Agreement between planned and measured 80% profiles was less than 0.4 mm on either side along the right-left axis. A systematic shift of the measured profile of slightly less than 1 mm in anterior and superior directions was noted along the anterior-posterior and superior-inferior axes, respectively.

Conflicts of interest: Mary Bird Perkins Cancer Center receives research funding from Elekta, Ltd. However, Elekta, Ltd. did not participate in the following: study design; collection, analysis, and interpretation of data; writing of the manuscript; or decision to submit the manuscript for publication. ⁎ Corresponding author. Mary Bird Perkins Cancer Center, 4950 Essen Ln, Baton Rouge, LA 70809. E-mail address: [email protected] (J.D. Fontenot). 1879-8500/$ – see front matter © 2014 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.prro.2013.02.006

e68

M.W. Sutton et al

Practical Radiation Oncology: January-February 2014

Conclusions: Submillimeter targeting accuracy can be achieved using a commercial kV-CBCT IGRT system. © 2014 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.

Introduction The use of commercial image guided radiation therapy (IGRT) systems has become a routine component of clinical radiation oncology at many facilities. In particular, on-board kilovoltage cone-beam computed tomography (kV-CBCT) systems are routinely used to assist in localization and targeting of the tumor volume. 1 Such developments in localization have improved daily targeting accuracy and offer the possibility of reducing margins needed to account for interfraction motion. Numerous reports have indicated the ability of a variety of commercial image guidance systems to achieve submillimeter targeting accuracy, 2-5 and these results help to support the rationale of reducing setup margins. However, the image guidance system is but 1 component within a complex chain of events needed to treat radiation therapy patients. The radiation therapy process is also critically dependent on the performance of the computed tomography (CT) simulator, the treatment planning system, and individual components of the linear accelerator delivery system (eg, the multileaf collimator, x-ray diaphragms, gantry, treatment table, etc.). Each of these necessary components also possesses device-specific performance measures with individual tolerances. So-called “end-to-end tests,” which involve subjecting a phantom geometry to all steps of the treatment process, provide an effective means of characterizing the aggregate performance of all treatment chain elements needed for clinical delivery. Most commonly, a simple test object (such as a radiopaque sphere) is aligned to the treatment isocenter using the IGRT system and irradiated with a circular or rectangular field from several gantry positions. A portal image is acquired at each gantry position using either film or the electronic portal imager, and the accuracy of the treatment delivery system is assessed by comparing the position of the test object within the field borders. 6-10 A more comprehensive approach to end-to-end tests involves the use of humanoid phantoms. After positioning the phantom at the treatment isocenter using the IGRT system, the dose delivered from all planned fields is then measured within the phantom geometry. The accuracy of the treatment delivery system is assessed by comparing the measured dose distribution with the planned distribution. The latter approach allows for a more complete assessment of the treatment process by allowing for delivery of increasingly complex plans under conditions that more closely reflect the clinical use of the system; however, because of the laborious nature of these measurements, relatively few such studies are reported in the literature. Vinci et al 11 previously reported the dosimetric accuracy

of conformal therapy using a stereoscopic x-ray image guidance system (Novalis ExacTrac; BrainLAB Inc, Westchester, IL). More recently Wang et al 12 reported the dosimetric accuracy of intensity modulated radiation therapy (IMRT), a new commercial digital linear accelerator equipped with a kV-CBCT image guidance system (TrueBeam STx; Varian Medical Systems, Palo Alto, CA). However, as the results of end-to-end tests are dependent on the treatment delivery system, the applicability of their results to the other commercial kV-CBCT IGRT manufacturer (Elekta AB, Stockholm Sweden) was unknown. Moreover, the study by Wang et al examined a variety of planning approaches, resulting in comparatively few measurements for each plan type and resulting in a limited assessment of the reproducibility (eg, precision) of the treatment delivery approach. The purpose of the present study was to assess the dosimetric accuracy and precision of CBCT-guided IMRT by subjecting an anthropomorphic phantom to repeated end-to-end testing. Our examination was extended by introducing intentional offsets in order to test the ability of the IGRT system to correct for patient misalignment.

Methods and materials To assess the delivery accuracy of CBCT-guided IMRT, a treatment plan was developed for the brain of an anthropomorphic head phantom. The phantom was positioned at 9 locations; 1 “correct” position and 8 “misaligned” positions along 3 orthogonal axes. The kVCBCT was used to align the phantom prior to each delivery. The treatment plan was delivered 3 times for each of the 9 positions, allowing film measurement of the delivered dose distribution in 3 orthogonal planes. The resulting accuracy of the system was assessed by comparing the planned dose to the delivered dose measured using film. Further details are provided below.

Phantom Treatments were planned and delivered to a cylindrical planning target volume (PTV) located in the cranium of a CIRS (Computerized Imaging Reference Systems, Norfolk, VA) Model 605 radiosurgery head phantom (see Fig 1). The phantom is composed of tissue-equivalent materials that represent brain, bone, spinal cord, vertebral disks, and soft tissue to within 1% of the true linear attenuation coefficient in the energy range of 50 keV-25 MeV. A removable tissue-equivalent cubical cassette

Practical Radiation Oncology: January-February 2014

Accuracy and precision of CBCT-guided IMRT

e69

within the dynamic range of the film. Planning objectives for organs at risk were based on typical objectives used in our clinic. The plan was optimized with direct machine parameter optimization, using a maximum of 35 total segments, a minimum segment area of 2 cm 2, and a minimum of 2 monitor units per segment. The final dose was calculated using the collapsed cone superpositionconvolution algorithm with a 2 × 2 × 2 mm 3 dose calculation grid. Additional details of the treatment planning process and results are described elsewhere. 13 The treatment plan and reference structures were then transferred to a commercial record and verify system (MOSAIQ, Version 2.0; Elekta AB) for delivery on a commercial radiation therapy system (Infinity; Elekta AB). Figure 1 (Left) The Computerized Imaging Reference Systems Model 605 radiosurgery head phantom used for the present study. The phantom contains a 6.35 × 6.35 × 6.35 cm 3 plastic block that can be loaded with film prior to irradiation.

inside the brain allows for placement of film in 2 anatomic planes. Dose distributions in 3 orthogonal planes can be measured by rotating the film cassette and repeating the dose delivery. Films are held firmly in place by 4 solid Delrin rods (4.1-mm diameter) that serve as fiducials for accurate registration of planned and measured dose distributions. 11

Simulation and treatment planning Imaging and simulation were conducted using typical clinical procedures at our institution. The phantom was aligned and affixed to the table of a GE Discovery CT simulator (General Electric Healthcare, Waukesha, WI) using an S-frame immobilization mask and head support (Med-Tec, Orange City, IA). Spherical radiopaque markers were placed at the intersection of the room lasers along the anterior and lateral surfaces of mask. A helical CT scan (0.7 × 0.7 × 1.25 mm 3 voxels, 140 kVp, 380 mA) was acquired for the purpose of radiation therapy treatment planning. The treatment planning CT was transferred to the Philips Pinnacle treatment planning system (version 9.0; Philips Healthcare, Fitchburg, WI). A cylindrical PTV (2cm diameter, 2-cm height with the long axis in the superior-inferior direction) representing a centrally located cranial lesion was contoured and centered within the film cassette. Delineated organs at risk included the spinal cord, optic nerves, optic chiasm, brainstem, orbits, and lenses of the eyes. A 7-field IMRT plan was constructed with coplanar beams placed approximately 51 degrees apart (0, 51, 102, 154, 205, 257, and 308 degrees) with the collimator angle set to 0 degrees. The plan was optimized to deliver 6000 cGy in 20 fractions to the PTV, which was chosen so that a fractional dose (300 cGy) produced an optical density

Phantom setup and treatment delivery Delivered doses were measured in each of 3 orthogonal anatomic planes (axial, sagittal, and coronal) using the film cassette of the head phantom. The accuracy of dose delivery was measured in 3 separate sessions. Each measurement session consisted of dose measurements in all 3 anatomic planes at each of the 9 initial phantom positions. Prior to phantom setup at each position, a piece of unexposed film was placed inside the film cassette of the head phantom with the film oriented along the anatomic plane of interest. An initial localization scan was then acquired using the head protocol of the kV-CBCT system (VolumeView; Elekta AB) with the film loaded inside the phantom. The additional imaging dose from our CBCT head protocol was measured to be less than 0.3 cGy, which is insignificant compared with the fractional treatment dose of 300 cGy. The kV-CBCT was automatically registered to the planning CT using the soft tissue (gray value) algorithm of the software. Because the treatment table (iBEAM evo; Elekta AB) cannot correct for rotational misalignments, rotational errors greater than 2 degrees were manually corrected. Otherwise, the suggested table shifts were applied, thus establishing the “correct” alignment of the phantom to the isocenter. At this point the phantom was either left in position (in the case of zero offset) or intentionally misaligned by ± 5 mm in the lateral, longitudinal, and vertical directions (in the case of the 8 offset positions) by manually moving the couch (see Fig 2). Shifts were measured using rulers affixed to the treatment table. After shifting the phantom, kV-CBCT image guidance was used to reposition the phantom. Again, the kV-CBCT was automatically registered to the planning CT using gray value algorithm. The resulting suggested table shifts were remotely applied to restore the patient to the proper treatment position. Following alignment, the treatment plan was delivered to the phantom. After irradiation, the film was removed and stored in a light-proof container for 24 hours.

e70

M.W. Sutton et al

Practical Radiation Oncology: January-February 2014

Imaging Reference Systems, Inc) at 100 cm source-to-axis distance at a depth of 10 cm and 5 cm backscatter. The absolute dose output was measured with a TN30006 Farmer Chamber (PTW Freiburg, Freiburg, Germany) using reference institutional output conditions. Based on repeated irradiations of the phantom in a constant setup, the spatial precision of the film-phantom measurement system used in this work has previously been reported as less than 0.32 mm at the 95% confidence interval. 11

Data analysis

Figure 2 Illustration of the 9 positions at which the anthropomorphic phantom was initially aligned. The phantom was either initially aligned to isocenter (eg, zero offset corresponding to position 0) or intentionally misaligned at 1 of 8 offset points corresponding to the corners of a 10-mm cube centered about the isocenter. The intentional offsets served to simulate incorrectly positioned patients who would be correctly aligned using the image guided radiation therapy system.

Film dosimetry The dosimeter used for measurement of planar doses in this study was Gafchromic EBT2 radiochromic film (International Specialty Products, Wayne, NJ), chosen because of its tissue-equivalence, superior spatial resolution, insensitivity to ambient light, ideal dynamic range (1 cGy-1000 cGy), and elimination of the need for wet chemical processing. Each film piece was prepared for use in the phantom by loading it into a custom aluminum template that allowed the film to be cut to match the dimensions of the film cassette, and for holes to be drilled in the film in locations corresponding to the location of the 4 compression rods in the film cassette. 11 All irradiated films were digitized using a Vidar DosimetryPRO Advantage (RED) 16-bit film digitizer (Herndon, VA) and RIT113 (Radiological Imaging Technology Inc, Colorado Springs, CO) film dosimetry software (version 5.2). All films were scanned in the same orientation and position in the digitizer at a resolution of 0.178 mm. Transmitted light intensity was converted to dose using a calibration curve (described in the next paragraph). Films containing measured dose distributions were registered to planned dose distributions using a template containing the coordinates of the center of the 4 compression rods in the film cassette. 13 Calibration films were created for each measurement session by irradiating 12 unexposed film squares with a 10 cm × 10 cm field of increasing dose (0-360 cGy). Each film square was placed in Plastic Water (Computerized

Relative dose profiles along the 2 principle axes for each orthogonal anatomic plane were used to compare the planned and measured dose distributions. Agreement between planned and measured profiles was assessed by determining the positioning alignment error (Δc) and the shift in the 80% dose point (Δ80) on either side of the profile. The positional alignment error (Δc) was defined as the displacement between the center (defined as the midpoint between the 70% dose levels on each side of the profile) of the planned and measured profiles. The 70% dose value was selected as it is located near the point of steepest dose gradient. The value of Δc is a measure of the alignment error in a measurement, where positive values indicate a shift of the measured profile in the left, posterior, or inferior direction and negative values indicate a shift in the right, anterior, or superior direction relative to the planned profile. Shifts in the 80% dose point (Δ80) were defined as the displacement of the 80% (eg, 240 cGy) dose level on each side of the measured profile relative to the planned profile. The 80% dose value was selected because of its clinical relevance in the shoulder of the profile. Positive values of Δ80 indicate a measured 80% isodose that fell outside the planned value, while negative values indicate a measured 80% isodose point was inside the planned value.

Results Figure 3 shows a comparison of planned and measured dose profiles along the 3 major orthogonal axes. As a film measurement in each anatomic plane contains 2 major axes, there were 2 independent measurements for each major axis. The overall targeting accuracy, as assessed by quantifying the shift in the center of the field (Δc), along the 3 major axes is shown in Table 1. The means and standard errors of the mean, c  σc , from 54 film measurements for each of the right-left (R-L), anterior-posterior (A-P), and superiorinferior (S-I) axes were + 0.08 ± 0.07 mm, + 0.60 ± 0.08 mm, and + 0.78 ± 0.16 mm, respectively. The quality of the measured data was assessed by comparing the agreement between measured and calculated profiles along the same axis in different anatomic

Practical Radiation Oncology: January-February 2014

Accuracy and precision of CBCT-guided IMRT

e71

Table 1 Values of c  σc for each measurement session along each axis Measurement session

Axis Right-Left

AnteriorPosterior

SuperiorInferior

1 2 3 Combined

0.12 ± 0.16 0.27 ± 0.14 − 0.14 ± 0.07 0.08 ± 0.07

0.74 ± 0.13 0.61 ± 0.14 0.46 ± 0.17 0.60 ± 0.08

0.84 ± 0.11 0.15 ± 0.26 1.37 ± 0.36 0.78 ± 0.16

The value for each session was computed from 18 profiles for each axis. The combined value was computed from all 54 profiles acquired over all 3 measurement sessions. All values are in units of millimeters.

measurement session are shown in Fig 4B. Each data point represents the average value of 9 measurements, with error bars indicating the standard error of the mean for that measurement session. The clinical impact of this study was assessed using the displacement between the planned and measured ___ 80% isodose line along the major axes. Values of Δ80 in each orthogonal ___ direction are shown in Table 3. Negative values of Δ80 indicate that the measured 80% dose point was within the planned 80% isodose ___ line (ie, underdose). Conversely, positive values of Δ80 indicate that the measured 80% dose point was outside the planned 80% isodose line (ie, overdose). From Fig 3, agreement between planned and measured profiles along the R-L and A-P was less than 0.4 mm on either side. A systematic shift of the measured profile of slightly less than 1 mm in anterior and superior directions was noted along the A-P and S-I axes, respectively.

Figure 3 Representative measured and planned profiles along the (A) anterior-posterior, (B) left-right, and (C) superior-inferior axes.

planes; ie, agreement between the measured and calculated profiles should be independent of plane. Values of c  σc , stratified by anatomic plane, are shown in Table 2. Slight differences in c along the R-L, A-P, and S-I major axes in different anatomic planes were noted, but were all less than 0.21 mm and within statistical uncertainties. A graphical representation of measured values of c is shown in Fig 4. Values of c for each anatomic plane and offset position are shown in Fig 4A. Each data point represents the average value from the 3 measurement sessions, and the number contained within each data symbol corresponds to its offset position (indicated in Fig 1). Values of c for each anatomic plane and

Discussion The dosimetric accuracy and precision of CBCT-guided IMRT were assessed in an anthropomorphic phantom. The ability of the IGRT system to correct for potentially large misalignments (± 5 mm) was assessed by applying offsets of known magnitude and direction to the initial placement of the phantom. On average, targeting accuracy was found to be within 1 mm in all 3 major anatomic planes. The results of the current study are consistent with recent studies that have examined targeting and dosimetric accuracy of other IGRT systems. Using similar methods reported in our study, Vinci et al 11 reported targeting errors (c) up to 0.3, 0.9, and 1.0 mm along the R-L, A-P, and S-I directions, respectively, for the Novalis Exactrac system (BrainLAB Inc). Wang et al 12 reported targeting errors of up to 0.5, 0.4, and 0.8 mm in the R-L, A-P, and S-I directions, respectively, for the Varian TrueBeam STx system (Varian Medical Systems). Target errors reported in

e72

M.W. Sutton et al

Table 2

Practical Radiation Oncology: January-February 2014

Values of c  σc from different anatomic planes for each measurement session

Measurement session

Axis Right-Left Axial

1 2 3 Combined

0.17 ± 0.25 ± − 0.07 ± 0.11 ±

0.22 0.15 0.11 0.10

Anterior-Posterior

Superior-Inferior

Coronal

Axial

Sagittal

Coronal

Sagittal

0.07 ± 0.20 0.29 ± 0.22 − 0.20 ± 0.08 0.05 ± 0.14

0.63 0.46 0.42 0.50

0.84 ± 0.16 0.77 ± 0.16 0.51 ± 0.22 0.71 ± 0.10

0.71 0.32 1.62 0.88

0.97 ± 0.14 − 0.03 ± 0.27 1.12 ± 0.58 0.69 ± 0.36

± 0.17 ± 0.20 ± 0.22 ± 0.06

± 0.15 ± 0.43 ± 0.38 ± 0.38

The value for each session was computed from 18 profiles for each axis. The combined value was computed from all 54 profiles acquired over all 3 measurement sessions. All values are in units of millimeters.

those studies compare favorably with values observed in our study. Similar to Wang ___ et al, our study noted a greater offset of c and Δ80 in the S-I direction in both the coronal and sagittal film measurements. Potential sources of this observation are numerous within the clinical treatment chain and include the finite planning CT slice width in the S-I direction (1.25 mm), inaccuracies in the treatment planning system dose algorithm and beam model, positional inaccuracies or misalignment of the multileaf collimator and jaws, uncertainties in couch positioning, and uncertainties in the geometric accuracy of the IGRT system. Although repeat measurements were performed using the image guidance and delivery systems, the CT simulation and planning process was performed once and is also subject to uncertainty. To estimate the uncertainty of the CT simulation and treatment planning process in our study, we performed 2 additional CT simulations of the phantom, including placement of isocenter in the CT image set using the original simulation fiducial markers. The treatment plan was copied to the repeat scans and the dose was calculated. The S-I profiles were exported and compared. The maximum alignment error (eg, Δc) between profiles calculated on the initial and repeat CT simulation scans was 0.11 mm, which is small compared with the other uncertainties in this work.

Conclusions

Figure 4 Values of c measured in each anatomic plane, stratified by (A) offset position and (B) measurement session. In (A), each data point represents the average value of 3 measurements, and the number contained within each data symbol corresponds to its offset position label (see Fig 1). In (B), each data point represents the average value of 9 measurements, with error bars indicating the standard error of the mean for that measurement session.

While the results of this study are dependent upon our institutional practices (linear accelerator and IGRT quality assurance procedures, alignment procedure, the accuracy of our treatment planning system and its commissioned model), we believe our clinical practice standards fall within currently accepted standards; our current quality assurance procedures for CT simulators, linear accelerators, and kV-CBCT system are consistent with the recommendations of American Association of Physicists in Medicine (AAPM) Task Groups 66, 14 142, 15 and 179, 16 respectively. Our treatment planning system is subjected to monthly quality assurance tests, and a recent study of

Practical Radiation Oncology: January-February 2014 Table 3

Accuracy and precision of CBCT-guided IMRT

e73

___ Values of Δ80 F σ for each measurement session along each axis. All values are in units of millimeters

Measurement session

Direction

1 2 3 Combined

− 0.10 0.01 − 0.16 − 0.08

Right

Left ± 0.49 ± 0.57 ± 0.30 ± 0.47

− 0.49 − 0.59 0.02 − 0.36

± 0.85 ± 0.83 ± 0.63 ± 0.81

Anterior

Posterior

− 0.98 ± 0.61 − 0.94 ± 0.65 − 0.63 ± 0.64 0.85 ± 0.71

0.60 ± 0.30 ± 0.48 ± − 0.46 ±

IMRT patient-specific quality assurance measurements at our institution showed outcomes well within action levels recommended by the AAPM Task Group 119. 17 As such, we conclude that other institutions that have adopted AAPM quality assurance recommendations could similarly achieve submillimeter accuracy and precision of CBCTguided IMRT.

References 1. McBain CA, Henry AM, Sykes J, et al. X-ray volumetric imaging in image guided radiotherapy: the new standard in on-treatment imaging. Int J Radiat Oncol Biol Phys. 2006;64:625-634. 2. Bissonnette JP, Moseley D, White E, et al. Quality assurance for the geometric accuracy of cone-beam CT guidance in radiation therapy. Int J Radiat Oncol Biol Phys. 2008;71:S57-S61. 3. Sharpe MB, Moseley DJ, Purdie TG, Islam M, Siewerdsen JH, Jaffray DA. The stability of mechanical calibration for a kV cone beam computed tomography system integrated with linear accelerator. Med Phys. 2006;33:136-144. 4. Meyer J, Wilbert J, Baier K, et al. Positioning accuracy of cone-beam computed tomography in combination with a HexaPOD robot treatment table. Int J Radiat Oncol Biol Phys. 2007;67:1220-1228. 5. Ploquin N, Rangel A, Dunscombe P. Phantom evaluation of a commercially available three modality image guided radiation therapy system. Med Phys. 2008;35:5303-5311. 6. Fukuda A. Pretreatment setup verification by cone beam CT in stereotactic radiosurgery: phantom study. J Appl Clin Med Phys. 2010; 11:3162. 7. Guan H, Hammoud R, Yin FF. A positioning QA procedure for 2D/2D (kV/MV) and 3D/3D (CT/CBCT) image matching for radiotherapy patient setup. J Appl Clin Med Phys. 2009;10:2954.

0.53 0.56 0.55 0.55

Inferior

Superior

− 0.98 ± 0.51 0.01 ± 1.15 − 1.43 ± 1.54 − 0.77 ± 1.25

0.82 ± 0.48 0.29 ± 1.07 1.31 ± 1.59 0.92 ± 1.19

8. Kry SF, Jones J, Childress NL. Implementation and evaluation of an end-to-end IGRT test. J Appl Clin Med Phys. 2012;13:3939. 9. Mao W, Speiser M, Medin P, Papiez L, Solberg T, Xing L. Initial application of a geometric QA tool for integrated MV and kV imaging systems on three image guided radiotherapy systems. Med Phys. 2011;38:2335-2341. 10. Sykes JR, Lindsay R, Dean CJ, Brettle DS, Magee DR, Thwaites DI. Measurement of cone beam CT coincidence with megavoltage isocentre and image sharpness using the QUASAR Penta-Guide phantom. Phys Med Biol. 2008;53:5273-5275. 11. Vinci JP, Hogstrom KR, Neck DW. Accuracy of cranial coplanar beam therapy using an oblique, stereoscopic x-ray image guidance system. Med Phys. 2008;35:3809-3819. 12. Wang L, Kielar KN, Mok E, Hsu A, Dieterich S, Xing L. An end-toend examination of geometric accuracy of IGRT using a new digital accelerator equipped with onboard imaging system. Phys Med Biol. 2012;57:757-769. 13. Sutton MW. Delivery accuracy of image guided radiation therapy using Elekta Infinity's on-board imaging system. Louisiana State University: MS thesis 2011: http://etd.lsu.edu/docs/available/etd07072011-104906/unrestricted/Suttonthesis.pdf. 14. Mutic S, Palta JR, Butker EK, et al. Quality assurance for computedtomography simulators and the computed-tomography-simulation process: report of the AAPM Radiation Therapy Committee Task Group No. 66. Med Phys. 2003;30:2762-2792. 15. Klein EE, Hanley J, Bayouth J, et al. Task Group 142 report: quality assurance of medical accelerators. Med Phys. 2009;36: 4197-4212. 16. Bissonnette JP, Balter PA, Dong L, et al. Quality assurance for image-guided radiation therapy utilizing CT-based technologies: a report of the AAPM TG-179. Med Phys. 2012;39:1946-1963. 17. Mancuso GM, Fontenot JD, Gibbons JP, Parker BC. Comparison of action levels for patient-specific quality assurance of intensity modulated radiation therapy and volumetric modulated arc therapy treatments. Med Phys. 2012;39:4378-4385.

Accuracy and precision of cone-beam computed tomography guided intensity modulated radiation therapy.

To assess the accuracy and precision of cone-beam computed tomography (CBCT)-guided intensity modulated radiation therapy (IMRT)...
729KB Sizes 0 Downloads 4 Views