|Year : 2013 | Volume
| Issue : 4 | Page : 181-186
Nondosimetric quality assurance of radiotherapy treatment planning system using multi-leaf collimator beam geometry phantom
Rajesh Kumar1, Rahul Kumar Chaudhary1, Sudesh Deshpande2, Parimal Patwe2, SD Sharma1, D. A. R. Babu1
1 Bhabha Atomic Research Centre, Radiological Physics and Advisory Division, CTCRS Building, Mumbai, Maharashtra, India
2 Department of Radiation Oncology, P. D. Hinduja National Hospital and Medical Research Centre, Mumbai, Maharashtra, India
|Date of Web Publication||8-Oct-2014|
Radiological Physics and Advisory Division, Bhabha Atomic Research Centre, CTCRS Building, Anushakti Nagar, Mumbai - 400094, Maharashtra
Source of Support: None, Conflict of Interest: None
Radiotherapy treatment planning system (RTPS) plays an important role in overall treatment delivery process. Nondosimetric quality assurance (QA) of the RTPS was carried out to assess the accuracy of nondosimetric parameters of the RTPS for regular/irregular fields obtained with jaws/multi-leaf collimator (MLC) using a dedicated MLC beam geometry phantom. Simulated radiation beams of field sizes 1 Χ 2 cm, 10 Χ 10 cm and 15 Χ 15 cm were created in RTPS using jaws for combinations of (i) 0° couch and gantry angles, and (ii) 323° gantry and 204° couch angles on computed tomography (CT) images of the phantom. Digitally reconstructed radiographs (DRR) for these setups were also generated. MLC co-ordinates were set in the RTPS corresponding to preset irregular field (formed by over travel of A or B bank of leaves) and resulting leaves positions were manually adjusted to fit the structures provided in the phantom. The dimensions of known geometries were measured and compared against the actual dimensions. The variation in measured and expected values of field sizes created by jaws was within 1.8 mm. In the case of DRR for 0° couch and gantry angles, the variation ranged from − 2.7 mm to 1.7 mm and for 204° couch and 323° gantry angles it was in the range of − 1.5-1.3 mm. The maximum variation between set leaf positions and manually adjusted positions for irregular field of MLC were found in the range of 2-4 mm. Nondosimetric QA of an RTPS was carried out, and results of the test provide confidence for its safe use for clinical practice.
Keywords: Beam geometry phantom, multi-leaf collimator, nondosimetric, quality assurance, treatment planning
|How to cite this article:|
Kumar R, Chaudhary RK, Deshpande S, Patwe P, Sharma S D, Babu D. Nondosimetric quality assurance of radiotherapy treatment planning system using multi-leaf collimator beam geometry phantom. Radiat Prot Environ 2013;36:181-6
|How to cite this URL:|
Kumar R, Chaudhary RK, Deshpande S, Patwe P, Sharma S D, Babu D. Nondosimetric quality assurance of radiotherapy treatment planning system using multi-leaf collimator beam geometry phantom. Radiat Prot Environ [serial online] 2013 [cited 2020 Jun 6];36:181-6. Available from: http://www.rpe.org.in/text.asp?2013/36/4/181/142386
| Introduction|| |
Radiotherapy is one of the established modality modalities for the treatment of cancer.  The success of use of radiation for cancer treatment depends on how accurately the treatment is planned and delivered. , To ensure the accurate planning and delivery of radiation treatment, quality assurance (QA) of all the components in the radiotherapy treatment chain is essential.  Radiotherapy treatment planning system (RTPS) is one of the major components of this chain, where the radiotherapy plan is generated keeping in mind the goal of radiotherapy to deliver a highly conformal and accurate dose to the tumor while sparing surrounding normal tissues and organs at risk. Considering the role of RTPS in radiotherapy treatment process, QA of RTPS becomes very important for safe, effective, and accurate implementation of radiation in cancer treatment.  The errors may be there right from the time of commissioning of RTPS or may creep into the planning system during its clinical use, such as, if there is an upgradation in the version of treatment planning software. The new software may contain bugs, or there may be changes in data files and errors in the input/output devices.  Therefore to detect these errors in time the QA is essential.
The QA of the RTPS can be divided into two parts: (i) Dosimetric QA, and (ii) nondosimetric QA. In general, QA of treatment delivery system draws most of the attention, whereas QA of TPS is given relatively less priority. In addition, nondosimetric QA has always been given least priority. However, the nondosimetric parameters of RTPS play very important role in finalizing the treatment plan in terms of dose volume histogram, which is estimated on the basis of geometrical accuracies, placing proper conformal beam portal and visualization of digitally reconstructed radiograph (DRR) for sparing the normal tissue. Limited information/methodology is available on nondosimetric QA of RTPS. , This paper describes the nondosimetric QA methodology and test results of RTPS using a dedicated nondosimetric QA phantom.
| Materials and methods|| |
Multi-leaf collimator beam geometry phantom
The QUASAR TM multi-leaf collimator (MLC) beam geometry phantom (Modus Medical Devices Inc., London, Ontario, Canada) was used in this study for nondosimetric QA of the RTPS. As shown in [Figure 1], the main parts of the phantom are: (i) A rotable component, (ii) a base, (iii) an adjustable leveling foot, (iv) a level Indicator and (v) leveling screws. The rotating component is a divergent acrylic three-dimensional trapezoid which is mounted on a base plate to simulate 360° of gantry and couch rotations. [Figure 2] shows a cross section of rotating the part of MLC beam geometry phantom. This part of the phantom is made up of outer acrylic square, outer air cavity, and rhomboid with inner and outer face having different size of steps, the inner cavity and 1 Χ 2 cm rectangle. These designs allow the MLC to align along the air/acrylic interface, forming a symmetrical, stepped-diamond shaped MLC field. The base of the phantom also contains a Z-line marker. This phantom is designed in such a way that when it is setup to match a gantry rotation along the central axis of the beam, the edges of the trapezoid follows the divergence of the beam. The phantom has the capability to test 1 Χ 2 cm, 10 Χ 10 cm and 15 Χ 15 cm fields at 100 cm source-to-axis distance (SAD) by lining up collimator jaws or MLC along the outer edge and the inner edge of the acrylic volume, respectively. The hole in the right-hand side of the inner rhomboid can be used to verify the orientations of the scans and reconstructions because it eliminates the symmetry of the inner stepped volume. 
|Figure 1: Photograph of multi-leaf collimator geometry phantom used for nondosimetric quality assurance of radiotherapy treatment planning system|
Click here to view
|Figure 2: Cross-sectional view showing internal structure in rotating part of multi-leaf collimator beam geometry phantom|
Click here to view
Volumetric computed tomography (CT) data of the phantom was acquired using a CT scanner (Light Speed VCT, GE Healthcare, USA). Before acquisition of CT data, the phantom was placed on the flat couch, aligned with CT room lasers and leveled properly. The proper leveling was ensured by taking transverse CT images near the superior and inferior ends of the phantom and the coordinates of the two outer high contrast points on the left and right side of the phantom`s base were recorded for both superior and inferior slices. The horizontal and vertical coordinates were checked (within ±1 mm) which ensured the recommended leveling criteria. Acquired CT images were transferred to the RTPS (Eclipse v6.5, Varian Medical System, USA) through the DICOM network. In RTPS, transverse CT images were searched for the slice with the prefix high-density reference marker inside the phantom. The position of the reference marker was set as the origin in the RTPS. To confirm proper orientation of the acquired images, the transverse CT images of the phantom were analyzed to verify that the central and right Z-line wires in the base of the phantom meet at the inferior end of the Z-line.  The display accuracies of reconstructed CT data in RTPS for multiple field apertures on transverse, sagittal, coronal and oblique CT images; beam`s eye view; and DRRs for different combinations of gantry and couch rotations were determined. The phantom is designed in such a way that geometrical information of the phantom is defined at 100 cm SAD, which was verified in the RTPS before conducting the nondosimetric QA on the RTPS. The known dimensions of square and rectangular geometry (15 Χ 15 cm, 10 Χ 10 cm and 1 Χ 2 cm) available in the phantom were used to assess the RTPS accuracies for displaying these field sizes. As three-dimensional RTPS is capable of reconstructing/displaying the acquired CT data in different planes  (transverse, sagittal and coronal), known geometries in the phantom were also used to assess the ability of RTPS to accurately reconstruct the CT images in different planes. The geometrical accuracies of DRRs generated by RTPS were verified by comparing the known dimensions of the phantom and measured/displayed values from the RTPS. These tests were carried out by generating the DRRs for 0° gantry and couch angles as well as 323° gantry and 204° couch angles keeping collimator at 0°. The MLC positioning accuracy of the RTPS were tested for 0° gantry and couch angles keeping collimator at 90°. The manufacturer recommended coordinates of the MLC leaves were entered in the RTPS to create full and half pyramid geometries. The MLC leaves were manually adjusted to match with the step boundaries. The difference between supplier recommended MLC coordinates and coordinates of the manually adjusted leaf positions were recorded.
| Results|| |
[Figure 3]a shows reconstructed CT image of the Z-wire placed in the base of the phantom to test the orientation as well as proper transfer of CT images to RTPS via DICOM network. [Figure 3]b shows the image of the wire at one corner. [Figure 3]c-g represents the position of wires in different slices. It can be observed from the CT slices that images of the wire smoothly converge from one end to other, and it indicates that the images were acquired at proper orientation and transferred successfully.
|Figure 3: Reconstructed computed tomography image showing Z-wire placed in the base of the phantom and wire smoothly converge from one end to other|
Click here to view
The dimensions of the square and rectangular geometries (15 Χ 15 cm, 10 Χ 10 cm and 1 Χ 2 cm) measured by the software ruler function of the RTPS in the coronal plane has been shown in [Figure 4], while [Figure 5] presents the similar data in the three planes. It is observed that the difference between the actual and measured field dimensions is within 1.8 mm in all the three planes, while this difference in transverse CT slice is within 1 mm. The measured values for these field sizes at other than 0° gantry and couch angles have been shown in [Figure 6] and the difference between actual and measured field dimensions are found within 0.6 mm.
|Figure 4: Screenshot showing measured dimension of square and rectangular geometries (15 × 15 cm, 10 × 10 cm and 1 × 2 cm) available in phantom by the software ruler function of the radiotherapy treatment planning system in the coronal plane|
Click here to view
|Figure 5: Screenshot showing measured dimension of square and rectangular geometries (15 × 15 cm, 10 × 10 cm and 1 × 2 cm) available in phantom by the software ruler function of the radiotherapy treatment planning system in the all three plane|
Click here to view
|Figure 6: Screenshot showing measured dimension of square and rectangular geometries (15 × 15 cm, 10 × 10 cm and 1 × 2 cm) available in phantom by the software ruler function of the radiotherapy treatment planning system at other than 0°gantry and couch angles|
Click here to view
[Figure 7] shows the snapshot of the DRR generated by the RTPS for 0° couch and gantry angles. The value of the difference between actual and measured field sizes is also shown in this figure. It is observed from the data that the difference in actual and measured (from DRR) field dimensions for 0° gantry and couch angles are in the range of − 2.7-1.7 mm. For, gantry angle = 323° and couch angle = 204°, this difference was found in the range of − 1.5-1.3 mm.
|Figure 7: Screenshot of digitally reconstructed radiograph generated by the radiotherapy treatment planning system (RTPS) for 0° couch and gantry angles and measured dimension of square and rectangular geometries (15 × 15 cm, 10 × 10 cm and 1 × 2 cm) available in phantom by the software ruler function of the RTPS|
Click here to view
The snapshots of set MLC leaf positions and adjusted MLC leaf positions with full and half pyramid structures of the phantom in RTPS for 0° gantry and couch and 90° collimator settings are shown in the [Figure 8]. [Figure 8]a and b shows screen shot of the full pyramid with set leaf positions and manually adjusted leaf positions, respectively. [Figure 8]c shows screen shots of half pyramid containing set leaf positions with over travelled MLC B bank while [Figure 8]d shows screen shots of half pyramid containing adjusted leaf positions with over travelled MLC B bank. [Figure 8]e shows screen shots of half pyramid containing set leaf positions with over travelled MLC A bank while [Figure 8]f shows screen shots of half pyramid containing adjusted leaf positions with over travelled MLC A bank. The differences in the set and adjusted leaf positions for full pyramid geometry [Figure 8]a-b are given in the [Table 1]. It can be observed from this table that difference between a set and adjusted leaf positions are within 2.0 mm. The differences in the set and adjusted leaf positions for half pyramid geometry with over travelled B bank [Figure 8]c-d are given in the [Table 2]. It can be observed from this table that difference between a set and adjusted leaf positions are within 3.1 mm. The differences in the set and adjusted leaf positions for half pyramid geometry with over travelled A bank [Figure 8]e-f are given in the [Table 3]. It can be observed from this table that difference between set and adjusted leaf positions are up to 4.0 mm.
|Figure 8: Screenshot showing multi-leaf collimator (MLC) leaf positions and adjusted MLC leaf positions with full and half pyramid structures of the phantom in radiotherapy treatment planning system for 0° gantry and couch and 90° collimator settings|
Click here to view
|Table 1: Difference (in mm) between set leaf position and manually adjusted leaf position for full pyramid structure |
Click here to view
|Table 2: Difference (in mm) between set leaf position and manually adjusted leaf position for half pyramid structure formed by overtravel of the left bank (A) of leaves |
Click here to view
|Table 3: Difference (in mm) between set leaf position and manually adjusted leaf position for half pyramid structure formed by overtravel of the right bank (B) of leaves |
Click here to view
| Discussion|| |
Results of nondosimetric tests of RTPS provide confidence for the safe use of the RTPS. For advance radiotherapy techniques, pretreatment dose verification is performed to avoid the dosimetric error.  However, these tests will not detect an error accrued during the image transfer from the CT workstation to the RTPS. Therefore, periodic QA to ensure the proper transfer of CT images to RTPS will play an important role in avoiding such source of error. A module of different known electron density materials can be attached with images to be transferred from CT work station to RTPS for further improving the QA method for proper image transfer process. In this study, the proper image transfer process was tested by localizing the Z-wire, and it was found that images were transferred properly. The difference between a set and measured dimensions of the square and rectangular fields were found to be up to 1.8 mm. Similar agreement was obtained while comparing the DRR results. Slightly higher difference (about 2.7 mm) was observed for 15 Χ 15 cm field at 0° couch and gantry angles. This difference can be attributed to the blurred image that introduces uncertainty in locating the exact edges. The possible reasons for blurred image of the phantom edge may be the higher slice thickness and choice of window level/width while measuring the distances using the software ruler function. In this work, a constant window level/width was maintained throughout the experiment for the sake of uniformity in measurements. It can be seen from the data in [Table 1] where all the leaves were moved without over travel that all the positions are within 2.0 mm. However, for over travelled leaf positions the difference between a set and adjusted leaf positions is 3.1 mm and 4.0 mm. The leaf positioning error of the order of 4 mm can introduce significant error in optimizing the overall treatment plan.
| Conclusion|| |
Nondosimetric QA of RTPS was carried out using commercially available QUASAR MLC beam geometry phantom. The measurements were carried out simulating the possible treatment situations. The difference in actual and measured field sizes was found within the acceptable variation. However, the difference in set and adjusted MLC leaf positions for complex field geometry was found somewhat larger than the conventional field geometries. It is recommended that the nondosimetric tests on the RTPS should also be conducted on the frequencies similar to dosimetric tests.
| References|| |
|1.||Patil SL, Mallaiah SH, Patil RK. Antioxidative and radioprotective potential of rutin and quercetin in Swiss albino mice exposed to gamma radiation. J Med Phys 2013;38:87-92. |
|2.||Ebert MA, Harrison KM, Cornes D, Howlett SJ, Joseph DJ, Kron T, et al. Comprehensive Australasian multicentre dosimetric intercomparison: Issues, logistics and recommendations. J Med Imaging Radiat Oncol 2009;53:119-31. |
|3.||Dixon P, O'Sullivan B. Radiotherapy quality assurance: Time for everyone to take it seriously. Eur J Cancer 2003;39:423-9. |
|4.||Boyer AL, Mok E, Luxton G. In: Shiu AS, Mellenberg DE. General Practice of Radiation Oncology Physics in the 21 st Century. Madison: Medical Physics Publishing; 2000. p. 187-230. |
|5.||Jacky J, White CP. Testing a 3-D radiation therapy planning program. Int J Radiat Oncol Biol Phys 1990;18:253-61. |
|6.||Fraass BA. Quality assurance for 3D treatment planning in teletherapy: Present and future. In: Palta JR, Mackie TR, editors. AAPM Monograph Number 22. Madison, WI: Advanced Medical Publishing; 1996. p. 253-318. |
|7.||McNiven A, Kron T, Van Dyk J. A multileaf collimator phantom for the quality assurance of radiation therapy planning systems and CT simulators. Int J Radiat Oncol Biol Phys 2004;60:994-1001. |
|8.||QUASAR Manual, Modus Medical Devices Inc.; 2010. |
|9.||Partridge M, Evans PM, Mosleh-Shirazi A, Convery D. Independent verification using portal imaging of intensity-modulated beam delivery by the dynamic MLC technique. Med Phys 1998;25:1872-9. |
|10.||International Atomic Energy Agency (IAEA). Commissioning and quality assurance of computerized planning systems for radiation treatment of cancer. Technical Report Series No. 430. Vienna: IAEA; 2004. |
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8]
[Table 1], [Table 2], [Table 3]