- Email: [email protected]
Helical tomotherapy quality assurance with ArcCHECK
Medical Dosimetry 39 (2014) 159–162
Medical Dosimetry journal homepage: www.meddos.org
Helical tomotherapy quality assurance with ArcCHECK David Chapman, Rob Barnett, Ph.D., and Slav Yartsev, Ph.D., D.Sc. London Regional Cancer Program, London Health Sciences Centre, London, Ontario, Canada
A R T I C L E I N F O
A B S T R A C T
Article history: Received 26 November 2013 Accepted 3 December 2013
To design a quality assurance (QA) procedure for helical tomotherapy that measures multiple beam parameters with 1 delivery and uses a rotating gantry to simulate treatment conditions. The customized QA procedure was preprogrammed on the tomotherapy operator station. The dosimetry measurements were performed using an ArcCHECK diode array and an A1SL ion chamber inserted in the central holder. The ArcCHECK was positioned 10 cm above the isocenter so that the 21-cm diameter detector array could measure the 40-cm wide tomotherapy beam. During the implementation of the new QA procedure, separate comparative measurements were made using ion chambers in both liquid and solid water, the tomotherapy onboard detector array, and a MapCHECK diode array for a period of 10 weeks. There was good agreement (within 1.3%) for the beam output and cone ratio obtained with the new procedure and the routine QA measurements. The measured beam energy was comparable (0.3%) to solid water measurement during the 10-week evaluation period, excluding 2 of the 10 measurements with unusually high background. The symmetry reading was similarly compromised for those 2 weeks, and on the other weeks, it deviated from the solid water reading by 2.5%. The ArcCHECK phantom presents a suitable alternative for performing helical tomotherapy QA, provided the background is collected properly. The proposed weekly procedure using ArcCHECK and water phantom makes the QA process more efficient. & 2014 American Association of Medical Dosimetrists.
Keywords: Helical tomotherapy Quality assurance ArcCHECK
Introduction Because of its unique treatment characteristics, helical tomotherapy requires different quality assurance (QA) than conventional linear accelerators. These QA requirements have been set out in the American Association of Physicists in Medicine Task Group 148 report along with various procedures to test various machine parameters.1 To simplify the QA process, several authors have presented different methods of monitoring beam parameters, using combinations of film and ion chambers2,3 or diode arrays.4,5 Currently ion chambers in water tanks or solid water phantoms are routinely employed to monitor output and to determine beam quality. Film measurements, diode arrays, or multiple ion chamber measurements are used to monitor the beam profile, which is cone shaped because of the tomotherapy’s flattening filter-free design. A tomotherapy Hi-ART II unit has a built-in detector array, consisting of xenon-filled ion chambers, to allow the acquisition of megavoltage computed tomography images to check patient positioning before treatment. Previous works have listed methods of using these onboard detectors for measuring beam profiles.4,5 Because of
Reprint requests to: Slav Yartsev, Ph.D., LRCP/LHSC, 790 Commissioners Road, East London, Ontario, Canada N6AL6. E-mail: [email protected]
the geometry of the array, there is a dip in the center of the beam profile measured by the onboard detector array, so it is preferable to have an independent measurement of the actual beam profile. The detectors can also be used to monitor the beam output and energy. Although helical tomotherapy treatments are almost always delivered with a rotating gantry (the exception being treatments delivered with the TOMODirect modality6), most of the currently recommended tests of beam parameters are done with a static gantry,2-5 and separate procedures are used to test the rotational output variation. This communication proposes a new technique for performing helical tomotherapy QA using an ArcCHECK (Sun Nuclear, Melbourne, FL) diode array. The advantage of performing routine QA with ArcCHECK is twofold: it measures radiation delivery with a rotating gantry as well as multiple beam parameters in a single procedure. The cylindrical geometry of the ArcCHECK phantom allows for measurement of the beam profile as the linear accelerator rotates, unlike 2-dimensional diode arrays. As ArcCHECK diodes also measure the entrance and exit doses of incident beams, the ratio of these readings can be used to evaluate the beam energy. In addition, ArcCHECK is already used at our center for patient delivery QA, and it would be convenient to perform both patient and machine QA with the same dosimeter.
0958-3947/$ – see front matter Copyright Ó 2014 American Association of Medical Dosimetrists http://dx.doi.org/10.1016/j.meddos.2013.12.002
160
D. Chapman et al. / Medical Dosimetry 39 (2014) 159–162
Fig. 1. MLC sinogram file. Each square represents the opening time for 1 leaf for 1 projection; the darker the square, the longer the leaf is open. (1) Warm-up rotation, (2) first custom rotation (magnified on the left), (3) rotation with all leaves closed, (4) all leaves opening for 15% of each projection for a rotation, (5) all leaves closed rotation, and (6) second custom rotation. Methods and Materials To get information from all 40 cm of the tomotherapy beam width using the 21cm diameter diode array the ArcCHECK phantom was set up 10 cm above the isocenter so that a complete beam profile could be measured as the gantry rotated. The moveable red lasers are raised accordingly to facilitate positioning of the phantom. The custom QA procedure was preprogrammed on the TomoTherapy operator station, version 4.2. The procedure employed a jaw width of 2.5 cm, a pitch of 1.2, a gantry rotation period of 40 seconds, and a couch setup distance of 59.5 cm. The sinogram editor was employed to create a custom multileaf collimator (MLC) opening pattern shown in Fig. 1. A warm-up (first) rotation with all leaves closed is followed by 3 rotations with varying leaf-opening patterns separated by 2 rotations with all leaves closed. The second and sixth rotations have groups of 8 MLC leaves opening in sequence and readings from different parts of the beam profile are used to evaluate the beam symmetry and cone ratio. The fourth rotation, with all the leaves open, has every leaf opening 15%, so that the dose would be on the same scale as that of the groups of 8 leaves opening 100%. Total beam-on time for this radiation delivery is 240 seconds. The data from ArcCHECK diode measurements were transferred to an Excel file and the readings from individual diodes in the path of the beam were extracted. Figure 2 shows the ArcCHECK-measured isodose lines, with individual diodes labeled. The output was measured as the average of the readings from 2 diodes (O1 and O2) receiving dose from the fourth rotation, where the highest dose was recorded. The effective energy was determined by calculating the ratio of the measured exit (24.16 cm of water equivalent depth; diodes EL1 and EL2) and entrance (3.28 cm of water equivalent depth; diodes EH1 and EH2) doses from the openings of the 8 leaves to the left and right of the beam center. Because of the spiral arrangement of the diodes in ArcCHECK, no 2 diodes lie in the same transverse plane of the phantom. Spiral diode positioning was chosen to improve the effective detector density in any incident beam. Because of this geometry, the diodes selected for entrance and exit dose are chosen as those measuring the highest dose for the regions generated by the opening of the 8 leaves. This was then compared with the expected ratio that was measured at the same depths using the vendor-supplied cheese phantom. The symmetry was measured as the ratio of the highest diode reading from the “right” entrance dose from the MLC leaves 49 to 56 Symmetry Right (SR) and the highest diode reading from the “left” entrance dose of MLC leaves 9 to 16 Symmetry Left (SL). The cone ratio was found by averaging both these diode readings and dividing by the average of the diode readings from the groups of 8 leaves immediately to the left and right of the isocenter (EH1 and EH2). An A1SL ion chamber (Standard Imaging, Middleton, WI) was inserted into the removable ArcCHECK cavity chamber support to obtain an independent absolute
dose measurement during the same radiation delivery. The ion chamber reading was corrected for leakage, which was determined each day. The proposed ArcCHECK QA procedure was tested by measurements performed in conjunction with the routine QA tests for a period of 10 weeks. These included measurements of (1) output with a water tank, (2) beam profile and energy using solid water slabs, (3) beam profile and energy with a MapCHECK diode array (Sun Nuclear, Melbourne, FL), and (4) a rotational procedure that used the Tomotherapy Hi-Art II onboard detector array. The water tank measured the dose at the isocenter using a Farmer ion chamber at 5-cm depth in a field of 40 5 cm delivered from a static gantry and all the leaves open. MapCHECK was irradiated with the same field 4 times: 3 times with the diode array offset by 10, 0, and 10 cm in the lateral direction to measure the full 40-cm width of the beam to get information about output and beam profile and at the isocenter with 3 cm of solid water buildup to measure the beam quality. The 200 seconds rotational procedure TDAT (Accuray Inc) was delivered with the couch retracted from the treatment beam and all the leaves open. The output and energy were measured using the onboard detector array and compared with a reference measurement.
Results The comparison of the normalized ArcCHECK diode readings, A1SL ion chamber at the center of ArcCHECK, and onboard detector to the water tank output measurement is shown in Fig. 3. The water tank measurements were normalized to the manufacturer0 s recommended value and the A1SL chamber and ArcCHECK diode readings were normalized using the measurement from week 4, which corresponded to the water tank measurement closest to the nominal output. The average percentage of differences from the output values measured in liquid water compared with the ArcCHECK and A1SL results were 0.4% ⫾ 0.5% and 0.3% ⫾ 0.6%, respectively, with a maximum difference of 1.3% for the ArcCHECK and 1.2% for the A1SL ion chamber. The ArcCHECK energy reading corresponded well to the measurement in solid water slabs for 8 weeks of measurements when the average percentage of difference was 0.3% ⫾ 0.5%, which is comparable to the MapCHECK/solid water agreement of 0.3% ⫾ 0.3%. The onboard detector’s energy evaluation had an average percentage of difference of 1.0% ⫾ 0.4% from the measurements in the solid water slabs. The results for the ArcCHECK, solid water, and MapCHECK symmetry measurements are shown in Fig. 4. The ArcCHECK symmetry value was calculated as the ratio of the SL and SR diodes readings, one measuring entrance dose from MLC leaves 49 to 56 and the other from leaves 9 to 16. Both these groups of leaves correspond to 5-cm wide regions on 2 slopes of the beam profile separated by 20 cm. For more than 8 of the 10 weeks the ArcCHECK symmetry measurement was systematically about 2.5% lower (average ratio: 0.975 ⫾ 0.002) than the solid water measurements (average ratio: 1.001 ⫾ 0.002). The ArcCHECK symmetry readings on weeks 5 and 9 differed by a much greater amount ( 7% and 6%, respectively) because of the ArcCHECK ASIC chip fault related to background collection, and so the readings were not included in the plot.
Fig. 2. ArcCHECK-measured dose distribution with diode positions displayed.
D. Chapman et al. / Medical Dosimetry 39 (2014) 159–162
161
Fig. 3. Percentage of difference of output measurements by ArcCHECK (squares), A1SL ion chamber (diamonds), and onboard detectors (open circles) compared with water tank results.
As discussed in the next section, an incorrect ArcCHECK background measurement on weeks 5 and 9 caused by a hardware problem with the ArcCHECK ASIC chip compromised the diode readings, particularly for the lower dose. The average deviation of the ArcCHECK cone ratio, normalized to an arbitrary value, and the solid water cone ratio, normalized to the expected value, was 0.3% ⫾ 0.2%. The ArcCHECK cone ratio was closer to the solid water measurement than the MapCHECK measurement, which had an average deviation of 1.0% ⫾ 0.2%.
Discussion The energy was 1 of the 2 beam parameters where the ArcCHECK results deviated the most from the routine QA checks. On 8 of the 10 weeks of measurements, the energy differed from solid water results by an average of 0.3%, except for weeks 5 and 9, which had deviations of 8%. Both these weeks also corresponded to the worst symmetry measurements. On these occasions, there was a problem with the background collected before the measurement. Weeks 5 and 9 had lower background threshold measurements of 60 than the other weeks, which had correct background thresholds of 80 as recorded in the text file of the ArcCHECK measurement. The lower collected background value caused the ArcCHECK to constantly add data from background diode readings for the entire procedure as demonstrated by the higher recorded beam-on time for those 2 weeks, of 290 seconds instead of 60 seconds for the other 8 weeks. This caused an apparent absolute dose increase more important for lower recorded doses, changing the ratio of the exit to entrance doses for the energy evaluation. The error in background collection also affected the output
measurement for weeks 5 and 9, and so these measurements were not included in the plot. Weeks 5 and 9 also had symmetry measurements, which had large errors of 7% and 6%, respectively. Over the other 8 weeks, the symmetry ratio reported by ArcCHECK was 2.5% lower than that with the 2-dimensional diode array MapCHECK and with A1SL ion chamber measurements in solid water. Because the symmetry depended on 2 diodes measuring dose from different gantry angles, this could have been caused by the output variation with gantry rotation described by Flynn et al.7 and affected by differences in the diodes, which remained after the application of the array calibration provided by Sun Nuclear. This is also partly because of the inherent differences of evaluating the symmetry using a rotational procedure instead of a static gantry. When the gantry rotates, any point in the bore receives dose from various areas along the beam profile, which means that an ArcCHECK symmetry measurement collects information from various regions of the profile. With a static gantry, it is possible to compare specific smaller areas limited by the detector size in the beam profile. Using MapCHECK or ion chambers, several radiation deliveries are needed to measure the full 40 cm of the beam profile with a static gantry, whereas it can be done in a single procedure with ArcCHECK by offsetting the phantom from the isocenter. All tomotherapy patients at our center are treated with a rotating gantry, so it is better to measure beam parameters in treatment conditions. For these reasons, the ArcCHECK procedure with rotating gantry is recommended for weekly QA, whereas static beam measurements can be used less frequently, such as after major component replacement. The ArcCHECK phantom is able to check multiple beam parameters with a single procedure, which means less time for
Fig. 4. Lateral beam symmetry measurements with ArcCHECK (squares), solid water (crosses), and MapCHECK (circles) by date.
162
D. Chapman et al. / Medical Dosimetry 39 (2014) 159–162
weekly QA. It usually takes about 25 to 30 minutes to perform the ArcCHECK measurements, whereas the combination of the water tank, MapCHECK or solid water, and onboard detector measurements normally takes more than an hour. ArcCHECK is able to check many beam parameters including the output, but we continue water tank output and the onboard detector measurements. In addition, it includes an independent check of the output using the A1SL ion chamber within the insert. Performing 2 independent measurements of the output also allows a qualitative measure of the beam energy because a decrease or increase in the ion chamber output that is not matched by a change in the ArcCHECK diode reading signals a change in the beam energy. The water phantom is the most dependable check of the output, which is the most important of the beam parameters. The onboard detectors’ measurement is a quick and simple (does not require any phantom setup) independent check of the rotational output and energy.
Conclusions ArcCHECK can be employed to measure several beam parameters of a helical tomotherapy unit with rotating gantry. Only a
single phantom and a single measurement are required, so the proposed QA procedure is more efficient than using multiple phantoms for measuring various parameters of the treatment beam. Using ArcCHECK along with the water tank represents an improvement over using multiple deliveries with MapCHECK or ion chambers in solid water.
References 1. Langen, K.M.; Papanikolaou, N.; Balog, J.; et al. QA for helical tomotherapy: Report of the AAPM Task Group 148. Med. Phys. 37:4817–53; 2010. 2. Goddu, S.M.; Mutic, S.; Pechenaya, O.L.; et al. Enhanced efficiency in helical tomotherapy quality assurance using a custom-designed water-equivalent phantom. Phys. Med. Biol. 54:5663–74; 2009. 3. Fenwick, J.D.; Tomé, W.A.; Jaradat, H.A.; et al. Quality assurance of a helical tomotherapy machine. Phys. Med. Biol. 49:2933–53; 2004. 4. Langen, K.M.; Meeks, S.L.; Poole, D.O.; et al. Evaluation of a diode array for QA measurements on a helical tomotherapy unit. Med. Phys. 32:3424–30; 2005. 5. ALaei, P.; Hui, S.K.; Higgins, P.D.; et al. The use of a commercial QA device for daily output check of a helical tomotherapy unit. Med. Phys. 33:3680–2; 2006. 6. Franco, P.; Catuzzo, P.; Cante, D.; et al. TomoDirect: An efficient means to deliver radiation at static angles with tomotherapy. Tumori 97:498–502; 2011. 7. Flynn, R.T.; Kissick, M.W.; Mehta, M.P.; et al. The impact of linac output variations on dose distributions in helical tomotherapy. Phys. Med. Biol. 53:417–30; 2008.
Medical Dosimetry journal homepage: www.meddos.org
Helical tomotherapy quality assurance with ArcCHECK David Chapman, Rob Barnett, Ph.D., and Slav Yartsev, Ph.D., D.Sc. London Regional Cancer Program, London Health Sciences Centre, London, Ontario, Canada
A R T I C L E I N F O
A B S T R A C T
Article history: Received 26 November 2013 Accepted 3 December 2013
To design a quality assurance (QA) procedure for helical tomotherapy that measures multiple beam parameters with 1 delivery and uses a rotating gantry to simulate treatment conditions. The customized QA procedure was preprogrammed on the tomotherapy operator station. The dosimetry measurements were performed using an ArcCHECK diode array and an A1SL ion chamber inserted in the central holder. The ArcCHECK was positioned 10 cm above the isocenter so that the 21-cm diameter detector array could measure the 40-cm wide tomotherapy beam. During the implementation of the new QA procedure, separate comparative measurements were made using ion chambers in both liquid and solid water, the tomotherapy onboard detector array, and a MapCHECK diode array for a period of 10 weeks. There was good agreement (within 1.3%) for the beam output and cone ratio obtained with the new procedure and the routine QA measurements. The measured beam energy was comparable (0.3%) to solid water measurement during the 10-week evaluation period, excluding 2 of the 10 measurements with unusually high background. The symmetry reading was similarly compromised for those 2 weeks, and on the other weeks, it deviated from the solid water reading by 2.5%. The ArcCHECK phantom presents a suitable alternative for performing helical tomotherapy QA, provided the background is collected properly. The proposed weekly procedure using ArcCHECK and water phantom makes the QA process more efficient. & 2014 American Association of Medical Dosimetrists.
Keywords: Helical tomotherapy Quality assurance ArcCHECK
Introduction Because of its unique treatment characteristics, helical tomotherapy requires different quality assurance (QA) than conventional linear accelerators. These QA requirements have been set out in the American Association of Physicists in Medicine Task Group 148 report along with various procedures to test various machine parameters.1 To simplify the QA process, several authors have presented different methods of monitoring beam parameters, using combinations of film and ion chambers2,3 or diode arrays.4,5 Currently ion chambers in water tanks or solid water phantoms are routinely employed to monitor output and to determine beam quality. Film measurements, diode arrays, or multiple ion chamber measurements are used to monitor the beam profile, which is cone shaped because of the tomotherapy’s flattening filter-free design. A tomotherapy Hi-ART II unit has a built-in detector array, consisting of xenon-filled ion chambers, to allow the acquisition of megavoltage computed tomography images to check patient positioning before treatment. Previous works have listed methods of using these onboard detectors for measuring beam profiles.4,5 Because of
Reprint requests to: Slav Yartsev, Ph.D., LRCP/LHSC, 790 Commissioners Road, East London, Ontario, Canada N6AL6. E-mail: [email protected]
the geometry of the array, there is a dip in the center of the beam profile measured by the onboard detector array, so it is preferable to have an independent measurement of the actual beam profile. The detectors can also be used to monitor the beam output and energy. Although helical tomotherapy treatments are almost always delivered with a rotating gantry (the exception being treatments delivered with the TOMODirect modality6), most of the currently recommended tests of beam parameters are done with a static gantry,2-5 and separate procedures are used to test the rotational output variation. This communication proposes a new technique for performing helical tomotherapy QA using an ArcCHECK (Sun Nuclear, Melbourne, FL) diode array. The advantage of performing routine QA with ArcCHECK is twofold: it measures radiation delivery with a rotating gantry as well as multiple beam parameters in a single procedure. The cylindrical geometry of the ArcCHECK phantom allows for measurement of the beam profile as the linear accelerator rotates, unlike 2-dimensional diode arrays. As ArcCHECK diodes also measure the entrance and exit doses of incident beams, the ratio of these readings can be used to evaluate the beam energy. In addition, ArcCHECK is already used at our center for patient delivery QA, and it would be convenient to perform both patient and machine QA with the same dosimeter.
0958-3947/$ – see front matter Copyright Ó 2014 American Association of Medical Dosimetrists http://dx.doi.org/10.1016/j.meddos.2013.12.002
160
D. Chapman et al. / Medical Dosimetry 39 (2014) 159–162
Fig. 1. MLC sinogram file. Each square represents the opening time for 1 leaf for 1 projection; the darker the square, the longer the leaf is open. (1) Warm-up rotation, (2) first custom rotation (magnified on the left), (3) rotation with all leaves closed, (4) all leaves opening for 15% of each projection for a rotation, (5) all leaves closed rotation, and (6) second custom rotation. Methods and Materials To get information from all 40 cm of the tomotherapy beam width using the 21cm diameter diode array the ArcCHECK phantom was set up 10 cm above the isocenter so that a complete beam profile could be measured as the gantry rotated. The moveable red lasers are raised accordingly to facilitate positioning of the phantom. The custom QA procedure was preprogrammed on the TomoTherapy operator station, version 4.2. The procedure employed a jaw width of 2.5 cm, a pitch of 1.2, a gantry rotation period of 40 seconds, and a couch setup distance of 59.5 cm. The sinogram editor was employed to create a custom multileaf collimator (MLC) opening pattern shown in Fig. 1. A warm-up (first) rotation with all leaves closed is followed by 3 rotations with varying leaf-opening patterns separated by 2 rotations with all leaves closed. The second and sixth rotations have groups of 8 MLC leaves opening in sequence and readings from different parts of the beam profile are used to evaluate the beam symmetry and cone ratio. The fourth rotation, with all the leaves open, has every leaf opening 15%, so that the dose would be on the same scale as that of the groups of 8 leaves opening 100%. Total beam-on time for this radiation delivery is 240 seconds. The data from ArcCHECK diode measurements were transferred to an Excel file and the readings from individual diodes in the path of the beam were extracted. Figure 2 shows the ArcCHECK-measured isodose lines, with individual diodes labeled. The output was measured as the average of the readings from 2 diodes (O1 and O2) receiving dose from the fourth rotation, where the highest dose was recorded. The effective energy was determined by calculating the ratio of the measured exit (24.16 cm of water equivalent depth; diodes EL1 and EL2) and entrance (3.28 cm of water equivalent depth; diodes EH1 and EH2) doses from the openings of the 8 leaves to the left and right of the beam center. Because of the spiral arrangement of the diodes in ArcCHECK, no 2 diodes lie in the same transverse plane of the phantom. Spiral diode positioning was chosen to improve the effective detector density in any incident beam. Because of this geometry, the diodes selected for entrance and exit dose are chosen as those measuring the highest dose for the regions generated by the opening of the 8 leaves. This was then compared with the expected ratio that was measured at the same depths using the vendor-supplied cheese phantom. The symmetry was measured as the ratio of the highest diode reading from the “right” entrance dose from the MLC leaves 49 to 56 Symmetry Right (SR) and the highest diode reading from the “left” entrance dose of MLC leaves 9 to 16 Symmetry Left (SL). The cone ratio was found by averaging both these diode readings and dividing by the average of the diode readings from the groups of 8 leaves immediately to the left and right of the isocenter (EH1 and EH2). An A1SL ion chamber (Standard Imaging, Middleton, WI) was inserted into the removable ArcCHECK cavity chamber support to obtain an independent absolute
dose measurement during the same radiation delivery. The ion chamber reading was corrected for leakage, which was determined each day. The proposed ArcCHECK QA procedure was tested by measurements performed in conjunction with the routine QA tests for a period of 10 weeks. These included measurements of (1) output with a water tank, (2) beam profile and energy using solid water slabs, (3) beam profile and energy with a MapCHECK diode array (Sun Nuclear, Melbourne, FL), and (4) a rotational procedure that used the Tomotherapy Hi-Art II onboard detector array. The water tank measured the dose at the isocenter using a Farmer ion chamber at 5-cm depth in a field of 40 5 cm delivered from a static gantry and all the leaves open. MapCHECK was irradiated with the same field 4 times: 3 times with the diode array offset by 10, 0, and 10 cm in the lateral direction to measure the full 40-cm width of the beam to get information about output and beam profile and at the isocenter with 3 cm of solid water buildup to measure the beam quality. The 200 seconds rotational procedure TDAT (Accuray Inc) was delivered with the couch retracted from the treatment beam and all the leaves open. The output and energy were measured using the onboard detector array and compared with a reference measurement.
Results The comparison of the normalized ArcCHECK diode readings, A1SL ion chamber at the center of ArcCHECK, and onboard detector to the water tank output measurement is shown in Fig. 3. The water tank measurements were normalized to the manufacturer0 s recommended value and the A1SL chamber and ArcCHECK diode readings were normalized using the measurement from week 4, which corresponded to the water tank measurement closest to the nominal output. The average percentage of differences from the output values measured in liquid water compared with the ArcCHECK and A1SL results were 0.4% ⫾ 0.5% and 0.3% ⫾ 0.6%, respectively, with a maximum difference of 1.3% for the ArcCHECK and 1.2% for the A1SL ion chamber. The ArcCHECK energy reading corresponded well to the measurement in solid water slabs for 8 weeks of measurements when the average percentage of difference was 0.3% ⫾ 0.5%, which is comparable to the MapCHECK/solid water agreement of 0.3% ⫾ 0.3%. The onboard detector’s energy evaluation had an average percentage of difference of 1.0% ⫾ 0.4% from the measurements in the solid water slabs. The results for the ArcCHECK, solid water, and MapCHECK symmetry measurements are shown in Fig. 4. The ArcCHECK symmetry value was calculated as the ratio of the SL and SR diodes readings, one measuring entrance dose from MLC leaves 49 to 56 and the other from leaves 9 to 16. Both these groups of leaves correspond to 5-cm wide regions on 2 slopes of the beam profile separated by 20 cm. For more than 8 of the 10 weeks the ArcCHECK symmetry measurement was systematically about 2.5% lower (average ratio: 0.975 ⫾ 0.002) than the solid water measurements (average ratio: 1.001 ⫾ 0.002). The ArcCHECK symmetry readings on weeks 5 and 9 differed by a much greater amount ( 7% and 6%, respectively) because of the ArcCHECK ASIC chip fault related to background collection, and so the readings were not included in the plot.
Fig. 2. ArcCHECK-measured dose distribution with diode positions displayed.
D. Chapman et al. / Medical Dosimetry 39 (2014) 159–162
161
Fig. 3. Percentage of difference of output measurements by ArcCHECK (squares), A1SL ion chamber (diamonds), and onboard detectors (open circles) compared with water tank results.
As discussed in the next section, an incorrect ArcCHECK background measurement on weeks 5 and 9 caused by a hardware problem with the ArcCHECK ASIC chip compromised the diode readings, particularly for the lower dose. The average deviation of the ArcCHECK cone ratio, normalized to an arbitrary value, and the solid water cone ratio, normalized to the expected value, was 0.3% ⫾ 0.2%. The ArcCHECK cone ratio was closer to the solid water measurement than the MapCHECK measurement, which had an average deviation of 1.0% ⫾ 0.2%.
Discussion The energy was 1 of the 2 beam parameters where the ArcCHECK results deviated the most from the routine QA checks. On 8 of the 10 weeks of measurements, the energy differed from solid water results by an average of 0.3%, except for weeks 5 and 9, which had deviations of 8%. Both these weeks also corresponded to the worst symmetry measurements. On these occasions, there was a problem with the background collected before the measurement. Weeks 5 and 9 had lower background threshold measurements of 60 than the other weeks, which had correct background thresholds of 80 as recorded in the text file of the ArcCHECK measurement. The lower collected background value caused the ArcCHECK to constantly add data from background diode readings for the entire procedure as demonstrated by the higher recorded beam-on time for those 2 weeks, of 290 seconds instead of 60 seconds for the other 8 weeks. This caused an apparent absolute dose increase more important for lower recorded doses, changing the ratio of the exit to entrance doses for the energy evaluation. The error in background collection also affected the output
measurement for weeks 5 and 9, and so these measurements were not included in the plot. Weeks 5 and 9 also had symmetry measurements, which had large errors of 7% and 6%, respectively. Over the other 8 weeks, the symmetry ratio reported by ArcCHECK was 2.5% lower than that with the 2-dimensional diode array MapCHECK and with A1SL ion chamber measurements in solid water. Because the symmetry depended on 2 diodes measuring dose from different gantry angles, this could have been caused by the output variation with gantry rotation described by Flynn et al.7 and affected by differences in the diodes, which remained after the application of the array calibration provided by Sun Nuclear. This is also partly because of the inherent differences of evaluating the symmetry using a rotational procedure instead of a static gantry. When the gantry rotates, any point in the bore receives dose from various areas along the beam profile, which means that an ArcCHECK symmetry measurement collects information from various regions of the profile. With a static gantry, it is possible to compare specific smaller areas limited by the detector size in the beam profile. Using MapCHECK or ion chambers, several radiation deliveries are needed to measure the full 40 cm of the beam profile with a static gantry, whereas it can be done in a single procedure with ArcCHECK by offsetting the phantom from the isocenter. All tomotherapy patients at our center are treated with a rotating gantry, so it is better to measure beam parameters in treatment conditions. For these reasons, the ArcCHECK procedure with rotating gantry is recommended for weekly QA, whereas static beam measurements can be used less frequently, such as after major component replacement. The ArcCHECK phantom is able to check multiple beam parameters with a single procedure, which means less time for
Fig. 4. Lateral beam symmetry measurements with ArcCHECK (squares), solid water (crosses), and MapCHECK (circles) by date.
162
D. Chapman et al. / Medical Dosimetry 39 (2014) 159–162
weekly QA. It usually takes about 25 to 30 minutes to perform the ArcCHECK measurements, whereas the combination of the water tank, MapCHECK or solid water, and onboard detector measurements normally takes more than an hour. ArcCHECK is able to check many beam parameters including the output, but we continue water tank output and the onboard detector measurements. In addition, it includes an independent check of the output using the A1SL ion chamber within the insert. Performing 2 independent measurements of the output also allows a qualitative measure of the beam energy because a decrease or increase in the ion chamber output that is not matched by a change in the ArcCHECK diode reading signals a change in the beam energy. The water phantom is the most dependable check of the output, which is the most important of the beam parameters. The onboard detectors’ measurement is a quick and simple (does not require any phantom setup) independent check of the rotational output and energy.
Conclusions ArcCHECK can be employed to measure several beam parameters of a helical tomotherapy unit with rotating gantry. Only a
single phantom and a single measurement are required, so the proposed QA procedure is more efficient than using multiple phantoms for measuring various parameters of the treatment beam. Using ArcCHECK along with the water tank represents an improvement over using multiple deliveries with MapCHECK or ion chambers in solid water.
References 1. Langen, K.M.; Papanikolaou, N.; Balog, J.; et al. QA for helical tomotherapy: Report of the AAPM Task Group 148. Med. Phys. 37:4817–53; 2010. 2. Goddu, S.M.; Mutic, S.; Pechenaya, O.L.; et al. Enhanced efficiency in helical tomotherapy quality assurance using a custom-designed water-equivalent phantom. Phys. Med. Biol. 54:5663–74; 2009. 3. Fenwick, J.D.; Tomé, W.A.; Jaradat, H.A.; et al. Quality assurance of a helical tomotherapy machine. Phys. Med. Biol. 49:2933–53; 2004. 4. Langen, K.M.; Meeks, S.L.; Poole, D.O.; et al. Evaluation of a diode array for QA measurements on a helical tomotherapy unit. Med. Phys. 32:3424–30; 2005. 5. ALaei, P.; Hui, S.K.; Higgins, P.D.; et al. The use of a commercial QA device for daily output check of a helical tomotherapy unit. Med. Phys. 33:3680–2; 2006. 6. Franco, P.; Catuzzo, P.; Cante, D.; et al. TomoDirect: An efficient means to deliver radiation at static angles with tomotherapy. Tumori 97:498–502; 2011. 7. Flynn, R.T.; Kissick, M.W.; Mehta, M.P.; et al. The impact of linac output variations on dose distributions in helical tomotherapy. Phys. Med. Biol. 53:417–30; 2008.