MLC Errors Research Articles

The sensitivity of gamma-index method to the positioning errors of high-definition MLC in patient-specific VMAT QA for SBRT.

BackgroundTo investigate the sensitivity of various gamma criteria used in the gamma-index method for patient-specific volumetric modulated arc therapy (VMAT) quality assurance (QA) for stereotactic body radiation therapy (SBRT) using a flattening filter free (FFF) photon beam.MethodsThree types of intentional misalignments were introduced to original high-definition multi-leaf collimator (HD-MLC) plans. The first type, referred to Class Out, involved the opening of each bank of leaves. The second type, Class In, involved the closing of each bank of leaves. The third type, Class Shift, involved the shifting of each bank of leaves towards the ground. Patient-specific QAs for the original and the modified plans were performed with MapCHECK2 and EBT2 films. The sensitivity of the gamma-index method using criteria of 1%/1 mm, 1.5%/1.5 mm, 1%/2 mm, 2%/1 mm and 2%/2 mm was investigated with absolute passing rates according to the magnitudes of MLCs misalignments. In addition, the changes in dose-volumetric indicators due to the magnitudes of MLC misalignments were investigated. The correlations between passing rates and the changes in dose-volumetric indicators were also investigated using Spearman’s rank correlation coefficient (γ).ResultsThe criterion of 2%/1 mm was able to detect Class Out and Class In MLC misalignments of 0.5 mm and Class Shift misalignments of 1 mm. The widely adopted clinical criterion of 2%/2 mm was not able to detect 0.5 mm MLC errors of the Class Out or Class In types, and also unable to detect 3 mm Class Shift errors. No correlations were observed between dose-volumetric changes and gamma passing rates (γ < 0.8).ConclusionsGamma criterion of 2%/1 mm was found to be suitable as a tolerance level with passing rates of 90% and 80% for patient-specific VMAT QA for SBRT when using MapCHECK2 and EBT2 film, respectively.

SU-E-T-405: Robustness of Volumetric-Modulated Arc Therapy (VMAT) Plans to Systematic MLC Positional Errors

Purpose: To evaluate the dosimetric impact of systematic MLC positional errors (PEs) on the quality of volumetric-modulated arc therapy (VMAT) plans. Methods: Five patients with head-and-neck cancer (HN) and five patients with prostate cancer were randomly chosen for this study. The clinically approved VMAT plans were designed with 2–4 coplanar arc beams with none-zero collimator angles in the Pinnacle planning system. The systematic MLC PEs of 0.5, 1.0, and 2.0 mm on both MLC banks were introduced into the original VMAT plans using an in-house program, and recalculated with the same planned Monitor Units in the Pinnacle system. For each patient, the original VMAT plans and plans with MLC PEs were evaluated according to the dose-volume histogram information and Gamma index analysis. Results: For one primary target, the ratio of V100 in the plans with 0.5, 1.0, and 2.0 mm MLC PEs to those in the clinical plans was 98.8 ± 2.2%, 97.9 ± 2.1%, 90.1 ± 9.0% for HN cases and 99.5 ± 3.2%, 98.9 ± 1.0%, 97.0 ± 2.5% for prostate cases. For all OARs, the relative difference of Dmean in all plans was less than 1.5%. With 2mm/2% criteria for Gamma analysis, the passing rates were 99.0 ± 1.5% for HN cases and 99.7 ± 0.3% for prostate cases between the planar doses from the original plans and the plans with 1.0 mm MLC errors. The corresponding Gamma passing rates dropped to 88.9 ± 5.3% for HN cases and 83.4 ± 3.2% for prostate cases when comparing planar doses from the original plans and the plans with 2.0 mm MLC errors. Conclusion: For VMAT plans, systematic MLC PEs up to 1.0 mm did not affect the plan quality in term of target coverage, OAR sparing, and Gamma analysis with 2mm/2% criteria.

SU-E-T-156: Evaluating VMAT Dose Accuracy Using 3D Detector Systems:A Comparative Study with Intentional MLC Errors

Purpose The purpose of this study was to investigate the sensitivity of commercially available 3D detectors (ArcCHECK and Delta4) to detect introduced errors on the Elekta MLCi head during Volumetric Modulated Arc Therapy (VMAT). Methods Six randomly selected VMAT plans were studied including two Head and Neck, three Prostate and one SBRT lung. Baseline measurements were taken, and then the known MLC errors were applied in both IN and OUT directions individually to each MLC leaf bank in increments of 0.5 mm and 1.0 mm symmetrically and also together in increments of 0.25 mm and 0.5 mm. The measurements were analyzed using various local gamma and DTA analysis criteria and compared to baseline measurements. The ArcCHECK was evaluated both with and without the acrylic plug. Results There was considerable variation in QA passing rates for both 3D detector systems ranging from 99.9% to 60.9% for all measurements with local gamma analysis. The ArcCHECK and Delta4 recorded a 10% decrease in passing rates when MLC leafbank were shifted 1.0 mm IN and OUT directions respectively. When the acrylic plug was inserted into the ArcCHECK, there was a maximum increase in gamma pass rate of 10.4% and an average increase of 3.9% respectively. For the Head and Neck VMAT plans, both devices detected only half of the introduced errors using conventional QA metrics. Conclusion There was a poor correlation between gamma pass rates and the introduced MLC errors for the 3D detectors studied. It is recommended to use the Acrylic plug insert in ArcCHECK for highly modulated VMAT plans as it improved passing rates by 3.1%. The Delta4 showed a higher sensitivity to the introduced errors when a local gamma analysis of 2%/2 mm was used and therefore that analysis criterion is recommended to detect any possible MLC errors during VMAT delivery.

SU‐E‐T‐473: A Patient‐Specific QC Paradigm Based On Trajectory Log Files and DICOM Plan Files

Purpose:To evaluate a remote QC tool for monitoring treatment machine parameters and treatment workflow.Methods:The Varian TrueBeamTM linear accelerator is a digital machine that records machine axis parameters and MLC leaf positions as a function of delivered monitor unit or control point. This information is saved to a binary trajectory log file for every treatment or imaging field in the patient treatment session. A MATLAB analysis routine was developed to parse the trajectory log files for a given patient, compare the expected versus actual machine and MLC positions as well as perform a cross‐comparison with the DICOM‐RT plan file exported from the treatment planning system. The parsing routine sorts the trajectory log files based on the time and date stamp and generates a sequential report file listing treatment parameters and provides a match relative to the DICOM‐RT plan file.Results:The trajectory log parsing‐routine was compared against a standard record and verify listing for patients undergoing initial IMRT dosimetry verification and weekly and final chart QC. The complete treatment course was independently verified for 10 patients of varying treatment site and a total of 1267 treatment fields were evaluated including pre‐treatment imaging fields where applicable. In the context of IMRT plan verification, eight prostate SBRT plans with 4‐arcs per plan were evaluated based on expected versus actual machine axis parameters. The average value for the maximum RMS MLC error was 0.067±0.001mm and 0.066±0.002mm for leaf bank A and B respectively.Conclusion:A real‐time QC analysis program was tested using trajectory log files and DICOM‐RT plan files. The parsing routine is efficient and able to evaluate all relevant machine axis parameters during a patient treatment course including MLC leaf positions and table positions at time of image acquisition and during treatment.

TU‐C‐BRE‐08: IMRT QA: Selecting Meaningful Gamma Criteria Based On Error Detection Sensitivity

Purpose:To develop a strategy for defining meaningful tolerance limits and studying the sensitivity of IMRT QA gamma criteria by inducing known errors in QA plans.Methods:IMRT QA measurements (ArcCHECK, Sun Nuclear) were compared to QA plan calculations with induced errors. Many (&gt;24) gamma comparisons between data and calculations were performed for each of several kinds of cases and classes of induced error types with varying magnitudes (e.g. MU errors ranging from ‐10% to +10%), resulting in over 3,000 comparisons. Gamma passing rates for each error class and case were graphed against error magnitude to create error curves in order to represent the range of missed errors in routine IMRT QA using various gamma criteria.Results:This study demonstrates that random, case‐specific, and systematic errors can be detected by the error curve analysis. Depending on location of the peak of the error curve (e.g., not centered about zero), 3%/3mm threshold=10% criteria may miss MU errors of up to 10% and random MLC errors of up to 5 mm. Additionally, using larger dose thresholds for specific devices may increase error sensitivity (for the same X%/Ymm criteria) by up to a factor of two. This analysis will allow clinics to select more meaningful gamma criteria based on QA device, treatment techniques, and acceptable error tolerances.Conclusion:We propose a strategy for selecting gamma parameters based on the sensitivity of gamma criteria and individual QA devices to induced calculation errors in QA plans. Our data suggest large errors may be missed using conventional gamma criteria and that using stricter criteria with an increased dose threshold may reduce the range of missed errors. This approach allows quantification of gamma criteria sensitivity and is straightforward to apply to other combinations of devices and treatment techniques.

SU-D-105-06: A Novel QA Process Based On Monte Carlo 2nd Check and Dose Reconstruction From Machine Log File On TomoTherapy

Purpose: To develop a QA process that combines the pre-treatment plan verification with post-treatment delivery verification. Individual source of error such as dose calculator, MLC, and gantry angle can be isolated and analyzed. Methods: A Monte-Carlo (MC) package, TomoPen, has been modified for pre-treatment plan verification. During TomoTherapy treatment, information such as ion chamber measured output, gantry and couch position, and MVCT exit-detectors fluence are recorded into a log file (rawdata). The log file is retrieved after each treatment and information about the treatment delivery is extracted. MLC movement is obtained through analysis of the exit-detector fluence. The actual delivery information is fed into MC dose calculation to evaluate the overall impact. 255 treatment deliveries in 38 patient plans were analyzed with this process. Results: The average percent difference from the MC 2nd check on planned dose was −1.2%. MLC errors have strong correlation with plans. On average, MLC errors were 0.1+/−2.2%, which will produce negligible error. However, error as big as 1.7+/−0.8% was observed. It is also observed that the MLC errors did not have big variations day-to-day for the same patient plan. The LINAC output has been a big source of error. On average, the output dropped by 0.8% during each treatment, and varies day to day as target degrades. The output error for all treatment during a 7 month span has been 1.3+/−1.1%. MC reconstructed dose is found to correlated with the mean MLC and output errors. The log file analysis can be completed in less than 30 seconds and the MC calculation took 2–4 minutes. Conclusion: We have developed a QA process capable of detecting errors from the planning to treatment delivery for TomoTherapy. The advantage over the existing phantom based QA method is that it provides more information about discrepancies in planning and delivery. This study is supported in part by UVa George Amorino Pilot Grant.

Sensitivity test of a 3D diode array system for MLC and GANTRY errors in IMRT delivery

Introduction Delta4 device is designed for three-dimensional dose verification and shows excellent reproducibility, linearity and doserate independence. The purpose of this work was to verify the sensitivity of the Delta4 (ScandiDos) for small MLC and gantry errors in intensity modulated radiotherapy (IMRT). Materials and methods Ten IMRT plans (5 prostate and 5 head and neck) were used. Systematic MLC errors of (–0.2), (–0.5), (–0.7), (–1), (–2), (–3), (–5) and (–10) mm, were introduced for one leaf. Gantry errors of 0.2°, 0.5°, 0.7°, 1°, 2°, 2.5°, 3°, 4° and 5° were introduced for the same plans. The Delta4 system was used to acquire and analyze the data. All the measurements were acquired consecutively for each patient. To eliminate the repeatability errors, standard deviations (SD) of the maximum gamma (Gmax) were calculated for 10 successively beams with no offsets. Gmax values were noted for both global gamma (normalization dose) and local gamma (local detector dose) with pass/fail criteria of 2%/2 mm. Gmax was evaluated for each modified field when compared to the measured field with no introduced error (0 mm or 0°). Because we had a single leaf with a small offset we evaluated only the Gmax and not the mean gamma or the percentage of area gamma Results For local gamma evaluation (LGE), the maximum SD for Gmax was of 0.06. For global gamma evaluation (GGE), the maximum SD was of 0.02. The MLC and gantry offsets had a noticeable impact on the plan evaluation. For all the ten plans modifications in Gmax were observed starting with 0.2 mm (from 0.2 to 4.75 for LGE and from 0.13 to 2.45 for GGE) for MLC plans and with 0.2° for the gantry errors (from 0.54 to 9.23 for LGE and from 0.37 to 4.71 for GGE). Conclusion The Delta4 system seems to be a very sensitive instrument for the quality assurance of IMRT plans. In our study it detected submilimetric and subdegree errors for MLC and gantry, respectively. In clinical use these small errors may not be detected and could be hidden by the residual error between planned and delivered dose.

Dosimetric impact of multileaf collimator position errors during prostatic treatment by dynamic arctherapy

Introduction The quality of a treatment performed by intensity modulation is based on the positioning accuracy of the multileaf collimator leaves (MLC, 120 leaves). The dosimetric impact achieved by the introduction of systematic errors on the leaves positioning has been studied to assess the generated effects. Materials and methods The study was performed on 8 patients treated for prostate localization with 80 Gy prescribed dose in dynamic arc therapy. Original treatment plans clinically approved were exported in DICOM RT to be processed using Matlab ® software. Systematic errors of magnitude between 0.5 mm and 5 mm were introduced by opening the MLC leaves, or by closing and shifting. The 128 revised plans were then introduced into the treatment planning system (TPS). They were recalculated keeping the same optimization constraints than the original plan. Geometric and dosimetric index (SFRO and ICRU used index) were calculated using the Artiview ® software, Aquilab. Results Changes introduced to the original plans are significantly impacting dosimetric treatment planning validation index. Among these index, the near maximum dose D_2%(Gy) received by the target volume (PTV) differs from 12%, 15% and 20% to the original treatment plan for maximum openings, shifts and closures respectively. For the same index and amplitude errors of up to 2 mm, an offset MLC error is less penalizing ( 5%) systematic error. Other dosimetric and geometric index complete the analysis as the Dice similarity coefficient (DSC), which is more sensitive to errors introduction in treatment plans. Conclusion Dosimetric validation is based on calculated and measured doses distributions comparison, thanks to tolerances (generally 3%/3 mm). This study allows us to understand the impact of potential MLC positioning errors on planned treatment and to discuss about the validity of quality assurance tolerance values.

A study on investigating the delivery parameter error effect on the variation of patient quality assurance during RapidArc treatment

The purpose of this study is to evaluate delivery parameter errors (DPEs) and their impact on clinical dose variation with the Varian RapidArc technique. The dynalog files of 16 head-and-neck patients were retrospectively analyzed to characterize three RapidArc DPEs: dose MU, gantry angle, and MLC gap errors. A total of 64 reconstructed plans were created by creating four variants of each of the original 16 plans (three with the DPEs applied individually and one with the three DPEs combined). These reconstructed plans were compared to the original plans to evaluate the impact of the DPEs on the clinical dose distribution. The mean dose MU, gantry angle, and MLC gap error for all patients were 0.00 ± 0.00 MU, -0.36 ± 0.03°, and 0.00 ± 0.01 mm, respectively. The DPEs had no obvious dosimetric impact on any of the studied dosimetric endpoints except the parotid dose. The gantry angle error, MLC gap error, and combined DPEs changed the parotid Dmean (mean dose) and parotid V30 (volume receiving at least 30 Gy) by 1%-2%. It is feasible to use dose distributions reconstructed from dynalog file data as a quality assurance tool. The dose MU, gantry angle, and MLC errors have only minor effects on the accuracy of the delivered dose.

A real time dose monitoring and dose reconstruction tool for patient specific VMAT QA and delivery

To develop a real time dose monitoring and dose reconstruction tool to identify and quantify sources of errors during patient specific volumetric modulated arc therapy (VMAT) delivery and quality assurance. The authors develop a VMAT delivery monitor tool called linac data monitor that connects to the linac in clinical mode and records, displays, and compares real time machine parameters with the planned parameters. A new measure, called integral error, keeps a running total of leaf overshoot and undershoot errors in each leaf pair, multiplied by leaf width, and the amount of time during which the error exists in monitor unit delivery. Another tool reconstructs Pinnacle(3)™ format delivered plan based on the saved machine logfile and recalculates actual delivered dose in patient anatomy. Delivery characteristics of various standard fractionation and stereotactic body radiation therapy (SBRT) VMAT plans delivered on Elekta Axesse and Synergy linacs were quantified. The MLC and gantry errors for all the treatment sites were 0.00 ± 0.59 mm and 0.05 ± 0.31°, indicating a good MLC gain calibration. Standard fractionation plans had a larger gantry error than SBRT plans due to frequent dose rate changes. On average, the MLC errors were negligible but larger errors of up to 6 mm and 2.5° were seen when dose rate varied frequently. Large gantry errors occurred during the acceleration and deceleration process, and correlated well with MLC errors (r = 0.858, p = 0.0004). PTV mean, minimum, and maximum dose discrepancies were 0.87 ± 0.21%, 0.99 ± 0.59%, and 1.18 ± 0.52%, respectively. The organs at risk (OAR) doses were within 2.5%, except some OARs that showed up to 5.6% discrepancy in maximum dose. Real time displayed normalized total positive integral error (normalized to the total monitor units) correlated linearly with MLC (r = 0.9279, p < 0.001) and gantry errors (r = 0.742, p = 0.005). There is a strong correlation between total integral error and PTV mean (r = 0.683, p = 0.015), minimum (r = 0.6147, p = 0.033), and maximum dose (r = 0.6038, p = 0.0376). Errors may exist during complex VMAT planning and delivery. Linac data monitor is capable of detecting and quantifying mechanical and dosimetric errors at various stages of planning and delivery.

MO-D-BRB-08: BEST IN PHYSICS (THERAPY) - A Real Time Dose Monitoring and Dose Reconstruction Tool for Patient Specific VMAT QA and Delivery.

To develop a real time dose monitoring and dose reconstruction tool to identify and quantify sources of errors during patient specific VMAT delivery and QAMethods: The VMAT delivery monitor tool called Linac Data Monitor (LDM) has been developed that connects to the linac in clinical mode and displays, records and compares real-time machine parameters to the planned parameters. A new quantity called integral error keeps a running total of leaf overshoot and undershoots errors in each leaf pair multiplied by leaf width and the amount of time during which error exists in MU delivery. Another tool reconstructs pinnacle format delivered plan based on the saved machine logfile and recalculates actual delivered dose in patient anatomy. Delivery characteristics of various standard and hypofractionation VMAT plans delivered on Elekta Axesse and Synergy linacs were quantified. The MLC and gantry errors for all the treatment sites were 0.00±0.59mm and 0.05±0.31°, indicating a good MLC gain calibration. Standard fractionation plans had a larger gantry error than hypofractionation plans due to frequent dose rate changes. On average the MLC errors were negligible but larger errors of 4-6 mm and 2.5° were seen when dose rate varied frequently. Large gantry errors occurred during the acceleration and deceleration process, and correlated well with MLC errors (p<0.0001). PTV mean, minimum, maximum dose discrepancy were 0.87±0.21%, 0.99±0.59% and 1.18±0.52%. The other OAR doses were within 2.5% except a few that showed up to 5.6% discrepancy in maximum dose. Realtime displayed normalized total positive integral error (normalized to the total MUs) correlated linearly with MLC and gantry errors (p<0.001) and dosimetric discrepancy (PTVmean: p<0.01; PTVmax: p<0.067 and PTVmax: p<0.046). Errors may exist during complex VMAT planning and delivery. LDM is capable of detecting and quantifying mechanical and dosimetric errors at various stages of planning and delivery.

WE-E-BRB-07: Direct 3D Fluence Calculation from Machine Beam Parameters for VMAT Delivery Verification.

MLC dynamic log files have been clinically used for quality assurance for years. The logged machine parameters and the derived beam 2D fluence maps can be compared to the ones obtained in the treatment plans in order to evaluate the accuracy and consistency of IMRT beam deliveries. In this study, we propose a computationally efficient method, called Direct 3D Fluence Calculation or D3DFC, to extend 2D fluence map derivation to 3D fluence volume computation. The aim is to extend dynalog-based QA from fixed-gantry IMRT to rotational-gantry VMAT. D3DFC calculates the 3D volume of photon fluence distribution directly from the machine parameters (gantry angles, jaw positions, MLC positions, collimator rotation angle, MU) contained in the dynamic log files or DICOM plans, without converting to 2D fluence maps per gantry angle in order to allow higher computation speed and accuracy. For testing, results were verified with film-in-air measurements. 3D fluence volumes computed from VMAT delivery records (with artificially introduced delivery errors) are compared to ones computed from the treatment plans to determine if these delivery errors can be identified. D3DFC is implemented in MATLAB and supports the DICOM plans, Varian MLC dynamic logs and Varian Truebeam machine logs. Computation takes 10 to 20 seconds for a single-arc VMAT plan or delivery records. The results showed that 1 mm MLC errors can be clearly detected using delivery-to-plan fluence volume comparison. Direct computation from machine parameters allows higher computation speed and accuracy. These advantages are useful for beam delivery verification purposes for which (slower) full patient CT based dose computation is less necessary. The calculated 3D photon fluence volume is useful to detect and visually present the VMAT delivery discrepancies.

SU-E-T-382: An Analysis of VMAT Monthly Quality Assurance.

The purpose of this study is to quantify the change in monthly VMAT quality assurance (QA) and determine the tests to maintain consistent delivery with a baseline. VMAT monthly QA has been performed for over 14 months on two Elekta Synergy LINACs. A baseline was established at acceptance and the monthly QA results were compared to those initial values. Films were used to test the dependence on varying dose rate, gantry speed, rotational direction, and MLC speed. These parameters were tested independently and then together in a test called the synchronicity test. Ion chamber readings test a DMLC field with a varying dose rate and MLC speed. Introducing intentional errors into the tested fields allowed the detectable limits of the QA to be determined. The monthly QA has consistently matched the baseline within a 3% dose limit on film. Analyzing the synchronicity film with a gamma test using a 5%/0.5mm tolerance showed a monthly pass rate of over 99%. The DMLC test has been identical for the entire course of VMAT QA. Furthermore, intentional changes in the MLC speed were noticed on the synchronicity test in the form of a smaller gamma pass rate as the MLC error was increased. There is a monthly agreement with all films testing individual parameters, ion chamber DMLC readings, as well as agreement with the synchronicity test. This collection of data has lead to the conclusion that only the synchronicity test and DMLC readings need to be performed on a monthly basis. If those tests fail, then individual parameters need to be tested to determine the singular cause of the error. Having a single test used as a red flag increases the efficiency of the monthly QA and is being implemented with an EPID to eliminate the use of film.

SU-E-T-598: Clinical Experience of Configuration, Commission and Implementation for SmartArc with MOSAIQ R&V System.

Clinical experience for configuration, commission and implementation of SmartArc with MOSAIQ R&V system. SmartArc is Pinnacle's solution for VMAT. On July 2011 we updated to Pinnacle 9.0 and purchased SmartArc. A standalone Eclipse workstation has been used 3 years for VMAT planning. Our clinical setting: Mosaiq 2.2; Varian Trilogy driven by 4DiTC and Varian 21ex driven by sequencer. Some key physics parameters have been studied: machine dose rate; MLC leaf speed; Leaf motion per gantry rotation. Tabletop was created by user to improve the dose accuracy for planning. In-house sandwich phantom was used with MapCheck for planner dose verification. A PTW 0.6cc ion chamber was included for absolute dose comparison. A copy of current machine data with default highest dose rate is recommended. It is due to after 10th iteration of optimization, the default dose rate will kick in. 2.5cm/s is the constraint for Varian Millennium 120 MLC; a buffer zone of 10% is suggested to reduce the MLC error on treatment. 2.25cm/s is used in our configuration. This results in MLC interlock if not configured correct. Maximum leaf motion per gantry rotation of 0.46cm/degree has to be checked for planning with Mosaiq R&V. Otherwise, undeliverable plan will show up sometimes on 4DiTC.Tabletop was exported as a DICOM structure from Eclipse to Pinnacle; we created a ROI template based on the matched tabletop.QA using in-house phantom for different sites were tested. Results for both planner dose and absolute chamber measurement are satisfactory. Special attentions need to be paid for dose rate, MLC leaf speed, leaf motion per gantry rotation when configuring SmartArc. Varian 21ex is supported but is slow for clinical delivery. Users need to create your own tabletop to improve planning accuracy. Conventional commission procedures for RapidArc also apply for SmartArc.

SU‐E‐T‐209: Testing of Prototype Delta4AT Diode Array for In‐Vivo Radiotherapy Dosimetry

Purpose: The Delta, 4AT, (ScandiDos, Sweden,) is a head‐mounted diode‐array detector that is intended to be used during a patientˈs treatment. The device currently only exists in prototype form. By introducing MU and MLC errors into radiotherapy plans, the ability of the hardware and software to detect errors was measured. The closest separation of the diodes, (when projected to the isocentre plane,) is 8mm in the prototype, but it is intended to reduce this to 2mm in the final device. Methods: The Delta, 4AT&lt;/sup&gt; is calibrated for each patient by taking a simultaneous measurement with the existing Delta, 4PT&lt;/sup&gt; phantom. The Delta, 4AT&lt;/sup&gt; then reports subsequent measurements of the same plan as predicted doses on the two perpendicular diodes arrays within the cylindrical Delta, 4PT&lt;/sup&gt; phantom. To test the accuracy of the Delta, 4AT, the Delta, 4PT&lt;/sup&gt; was left on the couch for the subsequent beam deliveries and the predicted doses in the Delta, 4PT&lt;/sup&gt; were compared to the actual doses. The Delta, 4AT&lt;/sup&gt; was calibrated with known good conformal and RapidArcTM plans, (Varian Medical Systems.). MU errors of −1.5%, −3%, +1.5% and +3% were used and MLC errors of +1.5 mm to the X1 bank, +3mm to the X1 bank and +1.5mm and +3mm to an individual MLC. Results: For a RapidArc plan and MU errors, the Delta, 4AT&lt;/sup&gt; measurements were within gamma &lt;1 (0.5%/0.5mm) of the Delta, 4PT&lt;/sup&gt; for &gt;99% of diodes, (detectors above 20% included). For a RapidArc plan and smaller MLC errors (+1.5mm), the Delta, 4AT&lt;/sup&gt; measurements were within gamma &lt; 1 for &gt; 95% of diodes at 3%/3mm. For a conformal plan and a +3mm X1 bank error the Delta, 4AT&lt;/sup&gt; measurements were within gamma &lt;1 of the Delta, 4PT&lt;/sup&gt; for &gt;95% of diodes at 2%/2mm. Conclusions: The prototype Delta4, AT&lt;/sup&gt; has been shown to produce accurate measurements and will allow verification throughout a patientˈs treatment.Thomas Matzen is employed by ScandiDos, the manufacturer of the Delta, 4AT. The meeting registration fees and travel costs of the presenting author may or may not be partly paid by ScandiDos.

SU‐E‐T‐267: A Comprehensive QA Method of MLC Accuracies for Dynamic IMRT

Purpose: The purpose of this study was to investigate the usefulness of our original quality assurance (QA) program and patientsˈ treatment of MLC data analysis for dynamic IMRT as routine practice. Method and Materials: As to the QA program, we have made three MLC sweep test patterns by changing MLC travel speed of 0.5, 1.0 and 2.0 cm/sec, respectively. All three fields were irradiated by three gantry angles of 0, 90 and 270 degree as daily QA. Each MLC logs of the QA program and patientsˈ treatment were recorded during Aug 2009 to Nov 2010. Actual MLC position errors (Bank A, Bank B) and gap errors compared with planned position were analyzed by a program (Matlab, Cybernet Inc.). Systematic and random errors of each MLC and gap were determined by everyday and every month. And then, the temporal data were also evaluated. Results: The accuracy of MLC position was depended upon its travel speed and gantry angle. Our data showed that the MLC errors in the QA program got worse as twice if the gantry angle was 270 deg compared to other angles. From patientsˈ treatment data analysis, we could monitor the accuracy of MLC during the whole treatment courses. Conclusion: In this study, we have found that our original QA program and patientsˈ treatment of MLC data analysis could be useful for clinical routine MLC QA for dynamic IMRT to identify MLC position accuracies by time‐line.

SU‐E‐T‐411: Validation of Plan Dose Perturbation Software for Use in Patient Specific IMRT Quality Assurance

Purpose: Patient‐specific IMRT quality assurance typically compares measured and calculated dose distributions in‐phantom. The clinical importance of disagreement between measured and calculated dose is often difficult to interpret. Recently, software has been developed (3DVH Sun Nuclear, Florida USA) that calculates the “delivered” dose distributions in patients by perturbing the calculated dose using errors detected in planar dose measurements. The aim of this work was to validate the accuracy of 3DVH versus commercial treatment planning software (Varian Eclipse 8.9). Methods: 17 DMLC prostate IMRT treatment plans were modified by Matlab script by adding randomly distributed positional errors of 1 or 3 mm in the planned MLC positions. These modified plans were delivered to a 2D diode array (MapCheck2, Sun Nuclear). Measured doses were compared to the dose planes from the original or recalculated MLC modified plans and 3DVH calculated a “delivered” dose in the patient. DVH comparisons were made based on mean dose and D99 for the PTV, and mean and maximum dose for the bladder and rectal for each patient and MLC error. Results: Paired Data showed good agreement for planning system and 3DVH dose volume histograms for both targets and organs at risk for both modified and unmodified plans, regardless of MLC error induced. Structure mean dose values as measured for PTV, Bladder and Rectum agreed to within 1% and dose maximum data showing a discrepancy of 2%. Conclusions: 3DVH is a potentially useful tool to calculate delivered dose distributions during patient specific IMRT quality assurance. The 3DVH data allows the user to analyze the difference between planned and delivered dose to given structures. By introducing random delivery errors, we showed that the 3DVH software was able to detect small fluence changes. This demonstrates the accuracy of 3DVH software against a commercial treatment planning system.

WE-D-AUD B-03: Assessment of An In-House Independent Phantom Dose Calculation Algorithm for IMRT QA

Purpose: To test the RTP-Filter, an in-house independent phantom dose calculation system for IMRT QA. Method and Materials: The dose from all IMRT beams delivered in treatment position was calculated in coronal planes where actual film densitometry for 100 cases was performed in solid water phantoms. A ray trace through the phantom determines the depth and location of points relative to an MLC shape, whose leaf pairs define integration elements for TPR, head scatter and phantom scatter. A virtual scatter source determines the head scatter. Factors for the off-axis ratio, penumbra, and out-of-field scatter and leakage are applied. The calculations were compared with the pinnacle planned dose to create 3%-3mm and 5%-5mm decision maps. The same analysis was performed for the original film measurements. Eight physicists scored 200 cases as pass, conditional, or fail, resulting in 1600 scores. Cases exceeding tolerances but that might be clinically acceptable were conditional. Results: Excellent agreement between pinnacle convolution calculations and the RTP-Filter was obtained, with 89.4% passing, 10.2% conditional, and 0.4% failing. The respective values for the film-to-pinnacle comparisons were 67.1%, 26.3%, and 6.6%. Both the film and the RTP-Filter had the best plan agreement for prostate cases and the least agreement for brain cases. While the trends for error identification were the same (e.g., presence of many small fields), the RTP-Filter represented more closely the historical decisions that were made, with errors in film densitometry and MLC mis-calibration identified as the causes for out-of-tolerance deviations. The good results of the RTP-Filter are most likely due to careful modeling of the out-of-field and penumbra components. Conclusion: The RTP-Filter provides fast and robust IMRT QA, decoupling the plan and machine QA. The rate of plan disagreement is lower than that for film due to the absence of film and MLC errors.

SU‐GG‐T‐151: On the Sensitivity of Patient‐Specific IMRT QA to MLC Position Errors

Purpose: to assess the ability of patient‐specific IMRT QA procedure using planar dose comparison in detecting MLC positioning errors. Method and Materials: Eight IMRT plans (total of 53 fields) previously used for patient treatment with 6 MV photon beams from Varian Trilogy were randomly selected. Treatment plans were generated with a commercial planning system (Pinnacle3). Random errors (±2 mm) and systematic errors (±1 mm and ±2 mm) of MLC positioning were introduced into each plan. Planar dose distribution, calculated for each field with and without MLC positioning errors, was compared to measured dose distribution using a 2‐dimensional diode array (MapCHECK™). The agreement was analyzed using absolute DTA comparison as well as γ analysis with 2%/2mm and 3%/3mm criteria. The decrease in average passing rates was used to assess the sensitivity of patient‐specific IMRT QA procedure to MLC positioning errors. Results: Decrease in average passing rates with absolute DTA comparison and γ analysis showed similar trends with the introduction of MLC positioning errors. With 3%/3mm, only slight decrease (∼2%) was observed with random errors and ±1 mm systematic errors. Though ∼7% drop was noticed with ±2 mm systematic errors, the average passing rates were all above the error detection threshold (90%). When using 2%/2mm, average passing rates with systematic errors dropped below the threshold, with ∼5% and ∼15% decrease for ±1 mm and ±2 mm errors respectively. The average passing rates with random errors were still above the threshold. Conclusion: Our study indicates that IMRT QA procedure using either DTA criteria or γ index is not sensitive to MLC positioning errors up to 2 mm when using 3%/3mm criteria. When using 2%/2mm, only systematic errors of ±2 mm could be detected. Patient‐specific IMRT QA should be combined with a periodic MLC positioning check in order to guarantee the accuracy of IMRT delivery.

SU‐GG‐T‐147: Automatic Detection of MLC Errors in VMAT Delivery

Purpose: The purpose of this study was a) to characterize the effects of deviations in MLC leaf positions and velocity on Intensity Modulated Arc Treatment (IMAT) delivery, b) to detect MLC leaf pre‐collision events. Method and materials: MLC position errors were measured performing IMAT delivery sequences produced by a planning system. The MLC delivery files were converted into a single arc using “in‐house” software. Each test plan was delivered with three different beam‐on times. The dose rates for the plans were adjusted accordingly. MLC leaf positions were recorded using the MLC controller at a sampling rate of 20 sec−1. Results: In the test sequences the MLC leaves exhibited a number of positional errors greater than 0.3 cm when moved from or to the “under‐jaw” position. The MLC leaf position errors arose from the discrepancy between the planned and the maximum velocity the leaf was able to attain. Deviations in the leaves positions may cause the leaves to get closer than 0.5 cm to each other, which was considered as a potential collision. Performing the IMAT delivery we were able to detect leaves which had the interleaf distance of 0.15 cm. A possible cause of the detected pre‐collision events was the degraded performance of the leaf motor. Conclusion: Leaf position errors caused by the rapid leaf movements from or towards the “under jaw” position tend to induce errors in the segments where the leaf is expected to assume a certain position different from the “under jaw”. An algorithm has been developed to reveal segments that are prone to leaf position errors caused by “slow” leaves. A method to detect segments where there is a high risk of leaf collisions has been developed and applied to the performed delivery sequences. An analysis of relative positional errors has been performed.