K Langen

Benchmarking of five commercial deformable image registration algorithms for head and neck patients

Benchmarking is a process in which standardized tests are used to assess system performance. The data produced in the process are important for comparative purposes, particularly when considering the implementation and quality assurance of DIR algorithms. In this work, five commercial DIR algorithms (MIM, Velocity, RayStation, Pinnacle, and Eclipse) were benchmarked using a set of 10 virtual phantoms. The phantoms were previously developed based on CT data collected from real head and neck patients. Each phantom includes a start of treatment CT dataset, an end of treatment CT dataset, and the ground‐truth deformation vector field (DVF) which links them together. These virtual phantoms were imported into the commercial systems and registered through a deformable process. The resulting DVFs were compared to the ground‐truth DVF to determine the target registration error (TRE) at every voxel within the image set. Real treatment plans were also recalculated on each end of treatment CT dataset and the dose transferred according to both the ground‐truth and test DVFs. Dosimetric changes were assessed, and TRE was correlated with changes in the DVH of individual structures. In the first part of the study, results show mean TRE on the order of 0.5 mm to 3 mm for all phantoms and ROIs. In certain instances, however, misregistrations were encountered which produced mean and max errors up to 6.8 mm and 22 mm, respectively. In the second part of the study, dosimetric error was found to be strongly correlated with TRE in the brainstem, but weakly correlated with TRE in the spinal cord. Several interesting cases were assessed which highlight the interplay between the direction and magnitude of TRE and the dose distribution, including the slope of dosimetric gradients and the distance to critical structures. This information can be used to help clinicians better implement and test their algorithms, and also understand the strengths and weaknesses of a dose adaptive approach. PACS number(s): 87.57.nj, 87.55.dk, 87.55.QrBenchmarking is a process in which standardized tests are used to assess system performance. The data produced in the process are important for comparative purposes, particularly when considering the implementation and quality assurance of DIR algorithms. In this work, five commercial DIR algorithms (MIM, Velocity, RayStation, Pinnacle, and Eclipse) were benchmarked using a set of 10 virtual phantoms. The phantoms were previously developed based on CT data collected from real head and neck patients. Each phantom includes a start of treatment CT dataset, an end of treatment CT dataset, and the ground-truth deformation vector field (DVF) which links them together. These virtual phantoms were imported into the commercial systems and registered through a deformable process. The resulting DVFs were compared to the ground-truth DVF to determine the target registration error (TRE) at every voxel within the image set. Real treatment plans were also recalculated on each end of treatment CT dataset and the dose transferred according to both the ground-truth and test DVFs. Dosimetric changes were assessed, and TRE was correlated with changes in the DVH of individual structures. In the first part of the study, results show mean TRE on the order of 0.5 mm to 3 mm for all phantoms and ROIs. In certain instances, however, misregistrations were encountered which produced mean and max errors up to 6.8 mm and 22 mm, respectively. In the second part of the study, dosimetric error was found to be strongly correlated with TRE in the brainstem, but weakly correlated with TRE in the spinal cord. Several interesting cases were assessed which highlight the interplay between the direction and magnitude of TRE and the dose distribution, including the slope of dosimetric gradients and the distance to critical structures. This information can be used to help clinicians better implement and test their algorithms, and also understand the strengths and weaknesses of a dose adaptive approach. PACS number(s): 87.57.nj, 87.55.dk, 87.55.Qr.

Efficient interplay effect mitigation for proton pencil beam scanning by spot-adapted layered repainting evenly spread out over the full breathing cycle

PURPOSE To develop and implement a practical repainting method for efficient interplay effect mitigation in proton pencil beam scanning (PBS). METHODS AND MATERIALS A new flexible repainting scheme with spot-adapted numbers of repainting evenly spread out over the whole breathing cycle (assumed to be 4 seconds) was developed. Twelve fields from 5 thoracic and upper abdominal PBS plans were delivered 3 times using the new repainting scheme to an ion chamber array on a motion stage. One time was static and 2 used 4-second, 3-cm peak-to-peak sinusoidal motion with delivery started at maximum inhalation and maximum exhalation. For comparison, all dose measurements were repeated with no repainting and with 8 repaintings. For each motion experiment, the 3%/3-mm gamma pass rate was calculated using the motion-convolved static dose as the reference. Simulations were first validated with the experiments and then used to extend the study to 0- to 5-cm motion magnitude, 2- to 6-second motion periods, patient-measured liver tumor motion, and 1- to 6-fraction treatments. The effect of the proposed method was evaluated for the 5 clinical cases using 4-dimensional (4D) dose reconstruction in the planning 4D computed tomography scan. The target homogeneity index, HI = (D2 - D98)/Dmean, of a single-fraction delivery is reported, where D2 and D98 is the dose delivered to 2% and 98% of the target, respectively, and Dmean is the mean dose. RESULTS The gamma pass rates were 59.6% ± 9.7% with no repainting, 76.5% ± 10.8% with 8 repaintings, and 92.4% ± 3.8% with the new repainting scheme. Simulations reproduced the experimental gamma pass rates with a 1.3% root-mean-square error and demonstrated largely improved gamma pass rates with the new repainting scheme for all investigated motion scenarios. One- and two-fraction deliveries with the new repainting scheme had gamma pass rates similar to those of 3-4 and 6-fraction deliveries with 8 repaintings. The mean HI for the 5 clinical cases was 14.2% with no repainting, 13.7% with 8 repaintings, 12.0% with the new repainting scheme, and 11.6% for the 4D dose without interplay effects. CONCLUSIONS A novel repainting strategy for efficient interplay effect mitigation was proposed, implemented, and shown to outperform conventional repainting in experiments, simulations, and dose reconstructions. This strategy could allow for safe and more optimal clinical delivery of thoracic and abdominal proton PBS and better facilitate hypofractionated and stereotactic treatments.

Comparison of multi-institutional Varian ProBeam pencil beam scanning proton beam commissioning data

Purpose Commissioning beam data for proton spot scanning beams are compared for the first two Varian ProBeam sites in the United States, at the Maryland Proton Treatment Center (MPTC) and Scripps Proton Therapy Center (SPTC). In addition, the extent to which beams can be matched between gantry rooms at MPTC is investigated. Method Beam data for the two sites were acquired with independent dosimetry systems and compared. Integrated depth dose curves (IDDs) were acquired with Bragg peak ion chambers in a 3D water tank for pencil beams at both sites. Spot profiles were acquired at different distances from the isocenter at a gantry angle of 0° as well as a function of gantry angles. Absolute dose calibration was compared between SPTC and the gantries at MPTC. Dosimetric verification of test plans, output as a function of gantry angle, monitor unit (MU) linearity, end effects, dose rate dependence, and plan reproducibility were compared for different gantries at MPTC. Results The IDDs for the two sites were similar, except in the plateau region, where the SPTC data were on average 4.5% higher for lower energies. This increase in the plateau region decreased as energy increased, with no marked difference for energies higher than 180 MeV. Range in water coincided for all energies within 0.5 mm. The sigmas of the spot profiles in air were within 10% agreement at isocenter. This difference increased as detector distance from the isocenter increased. Absolute doses for the gantries measured at both sites were within 1% agreement. Test plans, output as function of gantry angle, MU linearity, end effects, dose rate dependence, and plan reproducibility were all within tolerances given by TG142. Conclusion Beam data for the two sites and between different gantry rooms were well matched.

SU‐E‐J‐201: What is the Importance of Dose Recalculation for Adaptive Radiotherapy Dose Assessment?

PURPOSE To quantify the effect of using the planned dose distributions in lieu of performing dose recalculations on daily in-room images for adaptive radiotherapy (ART) dose assessment of head and neck cancer patients. METHODS 16 patients with cancers of the head and neck were treated using the TomoTherapy Hi-Art II (Accuray Inc., Sunnyvale, CA). Images of all patients were acquired prior to each treatment using the megavoltage CT (MVCT) capability of the TomoTherapy unit. Overall, images from 528 fractions were evaluated. For every image set, the delivered dose was estimated by both recalculating the dose distribution using the acquired MVCT and also by simply overlaying the planned distribution on the new images. ART dose assessment was performed using deformable image registration (DIR) to deform contours from the treatment plan to the images acquired during each fraction and to accumulate the estimated dose delivered during each fraction back to the reference treatment plan. The same DIR maps were applied to both dose estimation methods. Dosimetric endpoints were then compared between the DVHs computed using the recalculated or planned dose distributions. RESULTS The mean PTV D95% and D05% endpoints were 0.6±0.5% and 1.4±0.8% lower using the planned dose distributions compared to the recalculated distributions, respectively, across all patients. The mean parotid D50% was 2.4±1.5% greater using the planned distributions compared to the recalculated distributions. The parotid D50% from the planned distributions was also highly correlated with the parotid D50% from the recalculated distributions for each patient (mean r2=0.97±0.09). CONCLUSIONS For this treatment modality and site, dosimetric differences observed between overlaying the planned dose distributions and recalculating distributions on daily images were typically well within ±5%. This indicates that the dose distribution itself is robust against anatomic variations. This study was funded, in part, by a grant from Accuray Inc.

Free Breathing versus Breath-Hold Scanning Beam Proton Therapy and Cardiac Sparing in Breast Cancer

Purpose To assess dose errors caused by the interplay effects of free-breathing (FB) motion and to assess the value of breath-hold (BH) in terms of cardiac dose reduction for scanning beam proton therapy (SBPT). Materials and Methods Three patients with left-sided breast cancer previously treated with photon therapy were included in this dosimetric study: 2 following breast-conserving surgery with 2 hypothetical target volumes (whole breast alone and whole breast plus regional nodes, including supraclavicular, axillary, and internal mammary lymph nodes); and 1 postmastectomy, with the target volume including the chest wall plus regional nodes. SBPT plans were generated with various beam angles that ranged between 2 tangential directions. For treatment with FB, nominal dose and dose with interplay effects considered were calculated based on FB 4-dimensional computed tomography scans. SBPT plans on the BH computed tomography were also calculated for one of the patients, who was selected to be treated with photon therapy with BH. Results Dosimetric differences between nominal and interplay dose were small (average target mean dose, -0.06 Gy; range, -0.23 to 0.06 Gy; average heart mean dose, 0.001 Gy; range, -0.12 to 0.05 Gy). The largest dose deviations occurred in plans calculated with tangential beam arrangements; the smallest was noted with the en face beam. The average value of the mean heart dose with FB was <1 Gy. For the selected patient, the mean heart doses were 0.5 and 0.2 Gy for FB and BH, respectively. Conclusion Dose deviations caused by the interplay effects of respiratory motion during FB do not have a significant impact in SBPT with en face beam arrangement. BH does not significantly reduce cardiac dose. SBPT delivery is feasible with FB and can provide optimal target coverage and maximal sparing of the cardiopulmonary system, which can translate into improved clinical outcomes and a decrease in treatment-related morbidity in left-sided breast cancer patients or those who require internal mammary node coverage.

Concepts of PTV and Robustness in Passively Scattered and Pencil Beam Scanning Proton Therapy

Concepts of planning target volume and plan robustness in proton therapy are described. Implementation of these concepts into treatment planning is described. Proton plan sensitivity and interfractional and intrafractional anatomical variation are also discussed.

Pencil beam scanning versus passively scattered proton therapy for unresectable pancreatic cancer

Background With an increasing number of proton centers capable of delivering pencil beam scanning (PBS), understanding the dosimetric differences in PBS compared to passively scattered proton therapy (PSPT) for pancreatic cancer is of interest. Methods Optimized PBS plans were retrospectively generated for 11 patients with locally advanced pancreatic cancer previously treated with PSPT to 59.4 Gy on a prospective trial. The primary tumor was targeted without elective nodal coverage. The same treatment couch, target coverage and normal tissue dose objectives were used for all plans. A Wilcoxon t-test was performed to compare various dosimetric points between the two plans for each patient. Results All target volume coverage goals were met in all PBS and passive scattering (PS) plans, except for the planning target volume (PTV) coverage goal (V100% >95%) which was not met in one PS plan (range, 81.8-98.9%). PBS was associated with a lower median relative dose (102.4% vs. 103.8%) to 10% of the PTV (P=0.001). PBS plans had a lower median duodenal V59.4 Gy (37.4% vs. 40.4%; P=0.014), lower small bowel median V59.4 Gy (0.11% vs. 0.37%; P=0.012), lower stomach median V59.4 Gy (0.01% vs. 0.1%; P=0.023), and lower median dose to 0.1 cc of the spinal cord {35.0 vs. 38.7 Gy [relative biological effectiveness (RBE)]; P=0.001}. Liver dose was higher in PBS plans for median V5 Gy (24.1% vs. 20.2%; P=0.032), V20 Gy (3.2% vs. 2.8%; P=0.010), and V25 Gy (2.6% vs. 2.2%; P=0.019). There was no difference in kidney dose between PBS and PS plans. Conclusions Proton therapy for locally advanced pancreatic cancer using PBS was not clearly associated with clinically meaningful reductions in normal tissue dose compared to PS. Some statistically significant improvements in PTV coverage were achieved using PBS. PBS may offer improved conformality for the treatment of irregular targets, and further evaluation of PBS and PS incorporating elective nodal irradiation should be considered.

SU‐F‐T‐220: Validation of Hounsfield Unit‐To‐Stopping Power Ratio Calibration Used for Dose Calculation in Proton Radiotherapy

PURPOSE To validate the stoichiometric calibration of the Hounsfield Unit (HU) to Stopping Power Ratio (SPR) calibration used to commission a commercial treatment planning system (TPS) for proton radiotherapy dose calculation. METHODS AND MATERIALS The water equivalent thickness (WET) of several individual pig tissues (lung, fat, muscle, liver, intestine, rib, femur), mixed tissue samples (muscle/rib, ice/femur, rib/air cavity/muscle), and an intact pig head were measured with a multi-layer ionization chamber (MLIC). A CT scan of each sample was obtained and imported into a commercial TPS. The WET calculated by the TPS for each tissue sample was compared to the measured WET value to determine the accuracy of the HU-to-SPR calibration curve used by the TPS to calculate dose. RESULTS The WET values calculated by the TPS showed good agreement (< 2.0%) with the measured values for bone and all soft tissues except fat (3.1% difference). For the mixed tissue samples and the intact pig head measurements, the difference in the TPS and measured WET values all agreed to within 3.5%. In addition, SPR values were calculated from the measured WET of each tissue, and compared to SPR values of reference tissues from ICRU 46 used to generate the HU-to-SPR calibration for the TPS. CONCLUSION For clinical scenarios where the beam passes through multiple tissue types and its path is dominated by soft tissues, we believe using an uncertainty of 3.5% of the planned beam range is acceptable to account for uncertainties in the TPS WET determination.

SU-F-T-188: A Robust Treatment Planning Technique for Proton Pencil Beam Scanning Cranial Spinal Irradiation

PURPOSE To propose a proton pencil beam scanning (PBS) cranial spinal irradiation (CSI) treatment planning technique robust against patient roll, isocenter offset and proton range uncertainty. METHOD Proton PBS plans were created (Eclipse V11) for three previously treated CSI patients to 36 Gy (1.8 Gy/fractions). The target volume was separated into three regions: brain, upper spine and lower spine. One posterior-anterior (PA) beam was used for each spine region, and two posterior-oblique beams (15° apart from PA direction, denoted as 2PO_15) for the brain region. For comparison, another plan using one PA beam for the brain target (denoted as 1PA) was created. Using the same optimization objectives, 98% CTV was optimized to receive the prescription dose. To evaluate plan robustness against patient roll, the gantry angle was increased by 3° and dose was recalculated without changing the proton spot weights. On the re-calculated plan, doses were then calculated using 12 scenarios that are combinations of isocenter shift (±3mm in X, Y, and Z directions) and proton range variation (±3.5%). The worst-case-scenario (WCS) brain CTV dosimetric metrics were compared to the nominal plan. RESULTS For both beam arrangements, the brain field(s) and upper-spine field overlap in the T2-T5 region depending on patient anatomy. The maximum monitor unit per spot were 48.7%, 47.2%, and 40.0% higher for 1PA plans than 2PO_15 plans for the three patients. The 2PO_15 plans have better dose conformity. At the same level of CTV coverage, the 2PO_15 plans have lower maximum dose and higher minimum dose to the CTV. The 2PO_15 plans also showed lower WCS maximum dose to CTV, while the WCS minimum dose to CTV were comparable between the two techniques. CONCLUSION Our method of using two posterior-oblique beams for brain target provides improved dose conformity and homogeneity, and plan robustness including patient roll.

SU-F-T-151: Measurement Evaluation of Skin Dose in Scanning Proton Beam Therapy for Breast Cancer

PURPOSE To measure the skin dose and compare it with the calculated dose from a treatment planning system (TPS) for breast cancer treatment using scanning proton beam therapy (SPBT). METHODS A single en-face-beam SPBT plan was generated by a commercial TPS for two breast cancer patients. The treatment volumes were the entire breasts (218 cc and 1500 cc) prescribed to 50.4 Gy (RBE) in 28 fractions. A range shifter of 5 cm water equivalent thickness was used. The organ at risk (skin) was defined to be 5 mm thick from the surface. The skin doses were measured in water with an ADCL calibrated parallel plate (PP) chamber. The measured data were compared with the values calculated in the TPS. Skin dose calculations can be subject to uncertainties created by the definition of the external contour and the limitations of the correction based algorithms, such as proton convolution superposition. Hence, the external contours were expanded by 0, 3 mm and 1 cm to include additional pixels for dose calculation. In addition, to examine the effects of the cloth gown on the skin dose, the skin dose measurements were conducted with and without gown. RESULTS On average the measured skin dose was 4% higher than the calculated values. At deeper depths, the measured and calculated doses were in better agreement (< 2%). Large discrepancy occur for the dose calculated without external expansion due to volume averaging. The addition of the gown only increased the measured skin dose by 0.4%. CONCLUSION The implemented TPS underestimated the skin dose for breast treatments. Superficial dose calculation without external expansion would result in large errors for SPBT for breast cancer.