Gamma Metric Research Articles

Evaluating the Hounsfield unit assignment and dose differences between CT-based standard and deep learning-based synthetic CT images for MRI-only radiation therapy of the head and neck.

Magnetic resonance image only (MRI-only) simulation for head and neck (H&N) radiotherapy (RT) could allow for single-image modality planning with excellent soft tissue contrast. In the MRI-only simulation workflow, synthetic computed tomography (sCT) is generated from MRI to provide electron density information for dose calculation. Bone/air regions produce little MRI signal which could lead to electron density misclassification in sCT. Establishing the dosimetric impact of this error could inform quality assurance (QA) procedures using MRI-only RT planning or compensatory methods for accurate dosimetric calculation. The aim of this study was to investigate if Hounsfield unit (HU) voxel misassignments from sCT images result in dosimetric errors in clinical treatment plans. Fourteen H&N cancer patients undergoing same-day CT and 3T MRI simulation were retrospectively identified. MRI was deformed to the CT using multimodal deformable image registration. sCTs were generated from T1w DIXON MRIs using a commercially available deep learning-based generator (MRIplanner, Spectronic Medical AB, Helsingborg, Sweden). Tissue voxel assignment was quantified by creating a CT-derived HU threshold contour. CT/sCT HU differences for anatomical/target contours and tissue classification regions including air (<250 HU), adipose tissue (-250 HU to -51 HU), soft tissue (-50 HU to 199 HU), spongy (200 HU to 499 HU) and cortical bone(>500 HU) were quantified. t-test was used to determine if sCT/CT HU differences were significant. The frequency of structures that had a HU difference>80 HU (the CT window-width setting for intra-cranial structures) was computed to establish structure classification accuracy. Clinical intensity modulated radiation therapy (IMRT) treatment plans created on CT were retrospectively recalculated on sCT images and compared using the gamma metric. The mean ratio of sCT HUs relative to CT for air, adipose tissue, soft tissue, spongy and cortical bone were 1.7±0.3, 1.1±0.1, 1.0±0.1, 0.9±0.1 and 0.8±0.1 (value of 1 indicates perfect agreement). T-tests (significance set at t=0.05) identified differences in HU values for air, spongy and cortical bone in sCT images compared to CT. The structures with sCT/CT HU differences>80 HU of note were the left and right (L/R) cochlea and mandible (>79% of the tested cohort), the oral cavity (for 57% of the tested cohort), the epiglottis (for 43% of the tested cohort) and the L/R TM joints (occurring>29% of the cohort). In the case of the cochlea and TM joints, these structures contain dense bone/air interfaces. In the case of the oral cavity and mandible, these structures suffer the additional challenge of being positionally altered in CT versus MRI simulation (due to a non-MR safe immobilizing bite block requiring absence of bite block in MR). Finally, the epiglottis HU assignment suffers from its small size and unstable positionality. Plans recalculated on sCT yielded global/local gamma pass rates of 95.5%±2% (3mm, 3%) and 92.7%±2.1% (2mm, 2%). The largest mean differences in D95, Dmean , D50 dose volume histogram (DVH) metrics for organ-at-risk (OAR) and planning tumor volumes (PTVs) were 2.3%±3.0% and 0.7%±1.9% respectively. In this cohort, HU differences of CT and sCT were observed but did not translate into a reduction in gamma pass rates or differences in average PTV/OAR dose metrics greater than 3%. For sites such as the H&N where there are many tissue interfaces we did not observe large scale dose deviations but further studies using larger retrospective cohorts are merited to establish the variation in sCT dosimetric accuracy which could help to inform QA limits on clinical sCT usage.

Feasibility of Using Pseudo-CT for Dosimetry, Radiomics, and Efficacy Assessment in IMRT/VMAT of Brain Tumors: A Multi-Omics Analysis

Pseudo-CT generated by convolutional neural networks (CNN) and planning MRI has facilitated the promotion of MRI-Only. The technology not only reduces the time and money spent on CT scans, but also eliminates the cumbersome CT-MR registration. The feasibility in Stereotactic Brain Radiotherapy has been analyzed in previous studies by our team. However, when the prescribed requirements are not met, IMRT/VMAT are still selected. The study aims to evaluate the feasibility of pseudo-CT in IMRT/VMAT for brain cancer via the following 5 aspects: (1) image difference, (2) dose accuracy, (3) radiomics feature, (4) efficacy assessment, and (5) correlation analysis. Brain tumor patients who had received radiotherapy at our institution and had planning MRI and CT were included in the study. Redesign of IMRT and VMAT radiotherapy plans according to 3 × 15Gy for each patient. Hounsfield unit (HU) values for PTV and OARs were used to assess image differences. And dose accuracy analysis contained a 2D dose volume histogram (DVH) metrics (Dmax, Dmean, D2%, D50%, D98%, HI, CI) and 3D gamma metrics (criteria: 1-3%/2mm, 1%/1mm, 10% threshold). Then 107 original image features of PTV and OARs were extracted for radiometry analysis. And tumor control probability (TCP) of PTV (Poisson model) and normal tissue complication probability (NTCP) of OARs (Lyman-Kutcher-Burman model) were used for the variance analysis of efficacy assessment. Wilcox-test was used for significance of differences test (0.05), and spearman correlation analysis was used to explore the key features of the dose bias. A total of 42 patients were included, with 42 planning CTs and pseudo-CTs (mDixon-T1), and 38 pseudo-CTs (mDixon-T1-CE). The median volume of PTV was 4.1 cc (range 0.5-27.3), with no significant differences in HU, DVH, 3D gamma, and NTCP/TCP metrics. The median local gamma passing rates (1%/1mm) between planning CTs and pseudo-CTs (mDixon-T1) were 93.1% (range 65.5%-99.7%, IMRT) and 93.3% (range 63.9%-99.6%, VMAT). And more than 85% original radiomics features have significant difference. Further, the feature HU-Min was found to be more correlated with dose metrics in the correlation analysis. We speculate that it may be caused by the smoothing of the low frequency signal before outputting image. And since Shape_MeshVolume, Shape_VoxelVolume and PTV volume difference are highly correlated with dose deviation, it indicates that dose deviation affected by CT-MR registration. This study has the potential to provide guidance for the clinical application of pseudo-CT in the MRI-Only workflow with IMRT/VMAT for brain tumors. These quantitative results strongly indicate pseudo-CT can be used as a substitute for the initial CT in IMRT/VMAT for small brain lesions (size <5 cm, numbers <5), but not for radiomics analysis. Additionally, the impact of inter-image differences on dose accuracy is less significant compared to the deviation caused by image registration.

Quantifying the Dosimetric Impact of Tissue Hounsfield Unit Assignment in Deep Learning-based Synthetic CT Images For MRI-Only Radiation Therapy of The Head and Neck

MRI-only simulation for head and neck (HN) radiotherapy (RT) could allow for single-image modality planning with excellent soft tissue contrast. In the MRI-only simulation workflow, synthetic CTs (sCTs) are generated from MRI to provide electron density information for dose calculation. Bone/air regions produce little MRI signal which could lead to electron density misclassification in sCTs. Establishing the dosimetric impact of this error could inform quality assurance (QA) procedures using MRI-only RT planning. In this study we quantify differences in Hounsfield Unit (HU) values between paired CT/ sCTs of HN cancer patients and investigate the dosimetric impact on clinical treatment plans. Fourteen patients with head and neck cancer undergoing same-day CT and 3T MRI simulation were retrospectively identified. MRIs were deformed to the CT using multimodal deformable image registration. SCTs were generated from T1w DIXON MRI using a commercially available deep learning-based generator (MRIplanner, Spectronics). Tissue voxel assignment was quantified by creating a CT-derived HU threshold contour. CT/sCT HU differences for anatomical/target contours and tissue classification regions including air (<-250HU), adipose tissue (-250HU to -51HU), soft tissue (-50HU to 199HU), spongy (200HU to 499HU) and cortical bone (> 500HU) were quantified. T-test was used to determine if sCT/CT HU differences were significant. The frequency of structures that had a HU difference >70HU (the CT window-width setting for intra-cranial structures) was computed to establish structure classification accuracy. Clinical IMRT treatment plans created on CTs were retrospectively recalculated on sCT images using compared using the gamma metric. The mean ratio of sCT HUs relative to CT for air, adipose tissue, soft tissue, spongy and cortical bone were 1.7±0.3, 1.1±0.1, 1.0±0.1, 0.9±0.1 and 0.8±0.1 (value of 1 indicates perfect agreement). T-tests (significance set at t = 0.05) identified differences in HU values for air, spongy and cortical bone in sCT images compared to CT. The structures with sCT/CT HU differences > 70HU were the L/R cochlea and mandible, occurring in >78% of the tested cohort. These structures contain dense bone/air interfaces. Plans recalculated on sCTs yielded global/local gamma pass rates of 98.7%±1.2% (3mm,3%) and 95.5%±2.5% (1mm,1%). Mean differences in D95, D50, D10 and D2 dose volume histogram (DVH) metrics for organ-at-risk (OAR) and planning tumor volumes (PTV) were 0.8%±1.5% and 2.1% ± 1.2% respectively. In this cohort, HU differences in sCTs were observed but did not translate into a reduction in gamma pass rates and OAR/PTV DVH metrics. The acquisition of additional training data such as ultrashort echo time MRI could improve bone/air contrast and reduce bone/air sCT misclassifications. Further studies will establish the variation in sCT dosimetric accuracy using a larger retrospective cohort to inform QA limits on clinical sCT usage.

Utilizing the Neuronal Behavior of Spiking Neurons to Recognize Music Signals Based on Time Coding Features

This paper presents a Spiking Neural Network(SNN) architecture to distinguish two musical instruments: piano and violin. The acoustic characteristics of music such as frequency and time convey a lot of information that help humans in distinguishing music instruments within few seconds. SNNs are neural networks that work effectively with temporal data. In this study, 2-layer SNN temporal based architecture is implemented for instrument (piano and violin) recognition. Further, this research investigates the behaviour of spiking neurons for piano and violin samples through different spike based statistics. Additionally, a Gamma metric that utilises spike time information and Root Mean Square Error (RMSE) from the membrane potential are used for classification and recognition. SNN achieved an overall classification accuracy of 92.38% and 93.19%, indicating the potential of SNNs in this inherently temporal recognition and classification domain. On the other hand, we implemented rate-coding techniques using machine learning (ML) techniques. Through this research, we demonstrated that SNN are more effective than conventional ML methods for capturing important the acoustic characteristics of music such as frequency and time. Overall, this research showed the potential capability of temporal coding over rate coding techniques while processing spatial and temporal data.

Comparative performance analysis of 2D and 3D gamma metrics for patient specific QA in VMAT using Octavius 4D with 2D-Array 1500.

Gamma pass percentage (GPP) is the predominant metric used for Patient Specific Quality Assurance (PSQA) in radiation therapy. The dimensionality of the measurement geometry in PSQA has evolved from 2D planar to 3D planar, and presently to state-of-the-art 3D volumetric geometry. We aim to critically examine the performance of the three-dimensional gammas vis-à-vis the older gamma metrics of lower dimensionality to determine their mutual fungibility in PSQA, using clinically approved Volumetric Arc Therapy (VMAT) plans. Gamma pass percentages derived from PSQA for VMAT plans using Octavius 4D phantom with 2D-Array 1500 and its proprietary software were recorded. 2D planar, 3D planar, and 3D volumetric gamma pass percentages were retrospectively extracted for multiple treatment plans at three sites, using three acceptance limits, and for two modes of normalization. The differences in mean pass percentages, and the pairwise correlation between geometries were calculated within limits of statistical significance. A significant increase in mean pass rates was observed from 2D planar to 3D planar geometries. The difference was less pronounced from 3D planar to 3D volumetric. 2D planar v/s 3D planar showed a significant degree of correlation among themselves, which was not seen against most of the 3D volumetric pass rates. The mean gamma pass rates show conclusive evidence of the benefits of shifting from 2D planar to higher dimensions measurement geometries, but the benefits of using 3D volumetric compared to 3D planar is not always unequivocal. The correlations show mixed results regarding the interdependence of pass percentages at different geometries.

Human visual cortical gamma reflects natural image structure

Many studies have reported visual cortical gamma-band activity related to stimulus processing and cognition. Most respective studies used artificial stimuli, and the few studies that used natural stimuli disagree. Electrocorticographic (ECoG) recordings from awake macaque areas V1 and V4 found gamma to be abundant during free viewing of natural images. In contrast, a study using ECoG recordings from V1 of human patients reported that many natural images induce no gamma and concluded that it is not necessary for seeing. To reconcile these apparently disparate findings, we reanalyzed those same human ECoG data recorded during presentation of natural images. We find that the strength of gamma is positively correlated with different image-computable metrics of image structure. This holds independently of the precise metric used to quantify gamma. In fact, an average of previously used gamma metrics reflects image structure most robustly. Gamma was sufficiently diagnostic of image structure to differentiate between any possible pair of images with >70% accuracy. Thus, while gamma might be weak for some natural images, the graded strength of gamma reflects the graded degree of image structure, and thereby conveys functionally relevant stimulus properties.

P249] Influence of gamma index comparison on differences in dose calculation algorithm deposition method

Purpose Phantomless patient specific QA saves time and is able to detect errors by comparing the TPS dose distribution with a secondary independent dose calculation. TG 218 recommends that gamma analysis should be performed on metrics such as, mean gamma, γ > 1 and γ > 1.5, per structure. The aim of this study is to investigate how gamma index comparison is affected by differences on dose calculation algorithms when treating bone metastases. Methods 10 patients treated for bone metastases in different sites were chosen and 3D and VMAT plans were created using Elekta Monaco TPS in a Versa linac and 10 MV photons. Both 3D and VMAT dose distributions were calculated with Monaco XVMC Monte Carlo (MC) calculation algorithm using dose to medium in medium MC(Dm,m) and dose to water in medium MC(Dw,m) dose deposition modes. All plans were also calculated and evaluated using IBA Compass incorporating a Collapsed Cone convolution superposition algorithm calculating dose to water in medium CC(Dw,m). 3D dose distribution evaluation was performed in terms of gamma metrics such as mean gamma values and % PTV with γ > 1 and γ > 1.5 using 3%, 3 mm criteria for the following combinations: (a) MC(Dw,m) vs CC(Dw,m) and (b) MC(Dm,m) vs CC(Dw,m). Results For the 3D plans, average mean gamma values for MC(Dw,m) vs CC(Dw,m) and MC(Dm,m) vs CC(Dw,m) were found 0.37 ± 0.1 and 0.52 ± 0.12 respectively. The average % PTV points with γ > 1 were found 6.34 ± 5.7% and 12.6 ± 7.5% and the average % PTV points with γ > 1.5 1.45 ± 1.87% and 2.21 ± 2.28%, respectively. Corresponding values for the VMAT plans were 0.46 ± 0.12 and 0.68 ± 0.09 (mean gamma values), 9.48 ± 7.26% and 21.5 ± 7.3% (% PTV points with γ > 1) and 2.37 ± 2.8% and 7.95 ± 7.4% (% PTV points with γ > 1.5). Conclusions Gamma index comparison revealed that both MC and CC calculation algorithms are in fair agreement in dose calculation determination providing that the same dose deposition mode (i.e. Dw,m or Dm,m) is used. The use of different dose deposition mode, results in increased differences in dose calculation determination in structures/areas where materials different than water, such as bone, is included. The observed differences are also increased as the complexity of the plan increases (e.g. VMAT vs 3D).

Translucent poly(vinyl alcohol) cryogel dosimeters for simultaneous dose buildup and monitoring during chest wall radiation therapy.

Chest wall radiation therapy treatment delivery was monitored using a 5 mm thick radiochromic poly(vinyl alcohol) cryogel that also provided buildup material. The cryogels were used to detect positioning errors and measure the impact of shifts for a chest wall treatment that was delivered to a RANDO phantom. The phantom was shifted by ±2,±3, and ±5 mm from the planned position in the anterior/posterior (A/P) direction; these shifts represent setup errors and the uncertainty associated with lung filling during breath‐hold. The two‐dimensional absolute dose distributions measured in the cryogel at the planned position were compared with the distributions at all shifts from this position using gamma analysis (3%/3 mm, 10% threshold). For shifts of ±2,±3, and ±5 mm the passing rates ranged from 94.3% to 95.6%, 74.0% to 78.8%, and 17.5% to 22.5%, respectively. These results are consistent with the same gamma analysis performed on dose planes calculated in the middle of the cryogel and on the phantom surface using our treatment planning system, which ranged from 94.3% to 95.0%, 76.8% to 77.9%, and 23.5% to 24.3%, respectively. The Pinnacle dose planes were then scaled empirically and compared to the cryogel measurements. Using the same gamma metric, the pass rates ranged from 97.0% to 98.4%. The results of this study suggest that cryogels may be used as both a buildup material and to evaluate errors in chest wall treatment positioning during deep‐inspiration breath‐hold delivery. The cryogels are sensitive to A/P chest wall shifts of less than 3 mm, which potentially allows for the detection of clinically relevant errors.PACS number(s): 87.55.km, 87.57.uq

EPID-based dosimetry using a commercial software: Gamma analysis and DVH metrics for VMAT pre-treatment and in vivo QA

Introduction 3D EPID-based dosimetry can play a key role in the chain of verification procedures in VMAT treatments. Anatomical-based metrics could improve evaluation through meaningful clinical dosimetric parameters both for pre-treatment and in vivo QA. Purpose To evaluate the use of a commercial software to reconstruct dose from EPID measurements for pre-treatment verification and transit dosimetry using 3D patient-based metrics. Materials and methods Dosimetry Check is used for pre-treatment and in vivo 3D evaluations applied to ten VMAT treatments for various anatomical sites (prostate, head & neck, spine and abdomen), planned by Eclipse and delivered by Clinac 2100CD, equipped with a S1000 EPID. Furthermore, for each case, a verification plan is delivered to an homogeneous elliptical shaped solid water phantom to compare in vivo-phantom dose reconstruction and dose measurements by a micro ionization chamber CC01. Results Whole dose grid 3D gamma evaluation is performed with criteria 3%-3 mm and 5%-3 mm. Gamma metrics is applied to specific ROI of interest: PTV, CTV and OARs. DVHs are compared in terms of D2, D98, D50 and OARs constraints values used during planning. Differences between pre-treatment and in vivo reconstructed dose are mostly within 3% (5% in sites with small inhomogeneities). Isodose levels reconstructed from Dosimetry Check in phantom are within 3% compared to absolute dose measured by micro ionization chamber. Conclusion Patient-based metrics can be used to evaluate results in terms of target coverage and OAR constraints, improving QA evaluation for VMAT treatments. Further studies are needed to evaluate action levels for ROI-based gamma evaluation and critical DVH goals.

SU‐F‐T‐289: MLC Fluence Sonogram Based Delivery Quality Assurance for Bilateral Breast Irradiation

Purpose:Performing DQA for Bilateral (B‐L) breast tomotherapy is a challenging task due to the limitation of any commercially available detector array or film. Aim of this study is to perform DQA for B‐L breast tomotherapy plan using MLC fluence sinogram.Methods:Treatment plan was generated on Tomotherapy system for B‐L breast tumour. B‐L breast targets were given 50.4 Gy prescribed over 28 fractions. Plan is generated with 6 MV photon beam &amp; pitch was set to 0.3. As the width of the total target is 39 cm (left &amp; right) length is 20 cm. DQA plan delivered without any phantom on the mega voltage computed tomography (MCVT) detector system. The pulses recorded by MVCT system were exported to the delivery analysis software (Tomotherapy Inc.) for reconstruction. The detector signals are reconstructed to a sonogram and converted to MLC fluence sonogram. The MLC fluence sinogram compared with the planned fluence sinogram. Also point dose measured with cheese phantom and ionization chamber to verify the absolute dose componentResults:Planned fluence sinogram and reconstructed MLC fluence sinogram were compared using Gamma metric. MLC positional difference and intensity of the beamlet were used as parameters to evaluate gamma. 3 mm positional difference and 3% beamlet intensity difference were used set for gamma calculation. A total of 26784 non‐zero beamlets were included in the analysis out of which 161 beamlets had gamma more than 1. The gamma passing rate found to be 99.4%. Point dose measurements were within 1.3% of the calculated dose.Conclusion:MLC fluence sinogram based delivery quality assurance performed for bilateral breast irradiation. This would be a suitable alternate for large volume targets like bilateral breast, Total body irradiation etc. However conventional method of DQA should be used to validate this method periodically.

Contextual Atlas Regression Forests: Multiple-Atlas-Based Automated Dose Prediction in Radiation Therapy.

Radiation therapy is an integral part of cancer treatment, but to date it remains highly manual. Plans are created through optimization of dose volume objectives that specify intent to minimize, maximize, or achieve a prescribed dose level to clinical targets and organs. Optimization is NP-hard, requiring highly iterative and manual initialization procedures. We present a proof-of-concept for a method to automatically infer the radiation dose directly from the patient's treatment planning image based on a database of previous patients with corresponding clinical treatment plans. Our method uses regression forests augmented with density estimation over the most informative features to learn an automatic atlas-selection metric that is tailored to dose prediction. We validate our approach on 276 patients from 3 clinical treatment plan sites (whole breast, breast cavity, and prostate), with an overall dose prediction accuracies of 78.68%, 64.76%, 86.83% under the Gamma metric.

Ir-192 HDR Source Parameter Validation Using Multichannel Gafchromic Film

Currently, Ir-192 HDR treatment planning system commissioning is only based on a single absolute measurement of source activity supplemented by validation of TG-43 parameters via dose point spot checks without independent verification that the planned distribution corresponds to the actual delivered dose. Reports in the literature indicate a need for more thorough tests to evaluate end-to-end accuracy of the treatment planning system. Gafchromic film has the potential for widespread clinical use to verify multiple source parameters due to its convenience for application without requiring a processor, excellent spatial resolution imperative in brachytherapy, and reported accuracy in external beam quality assurance. The purpose on this work is to present a methodology to validate the planned dose distribution for a single HDR dwell position as part of initial treatment planning commissioning and each source exchange to ensure that there are no unexpected changes in the source itself. The procedure presented independently calculates the one-dimensional dose perpendicular to the source, creates a robust calibration curve, and then compares a planned two-dimensional dose distribution with measured film dosimetry. A sheet of Gafchromic EBT3 film positioned in contact with a delivery catheter surrounded by water-equivalent scatter material is irradiated with an Ir-192 Nucletron HDR source. Dwell time is selected to be slightly below the film saturation density at the normalization point 5.0 mm from the source center. The film is digitized with an Epson 10000 XL color scanner. Film analysis is performed with MatLab imaging toolbox. Transmission values for the green channel, previously determined to have the greatest range and contrast, was converted into optical density (OD) as a function of perpendicular distance from the center of the calculated line source center. A corresponding theoretical dose curve was calculated using the TG43 formalism. The calculated dose in 0.1 mm increments was fitted to the measured OD in the 5-25 mm range to a sixth order polynomial. A comparable single source plan was created using the Oncentra treatment planning system. The 3D dose matrix was exported to a Varian Eclipse work station for convenient extraction of the 2D coronal dose plane with predetermined origin and scale at the level of the source for MatLab analysis. The Oncentra and measured film dose planes were registered at their corresponding origins as determined by lateral and vertical symmetry. The 2D dose matrices were individually scaled to 0.1 mm per pixel. The OD calibration curve consisting of 200 points was statistically valid, robust, and in agreement with Oncentra planned dose within +/- 1.5 % in the clinical range. A sixth order polynomial was necessary to accurately fit the film response throughout the usable film density range. Agreement between film and planned dose matrices was calculated at each corresponding position in the 2D dose plane as the absolute difference between the plan and measured doses. Areas near the source (5 mm lateral and 10 mm vertical to the source center) were excluded. Over 97% of the points had a dose agreement difference of less than 3% at exact corresponding points, which is an improvement over previously-published results using a gamma metric of 3% AND 3 mm. Figure below shows the agreement difference distribution. For clarity, only one side of the symmetrical image is displayed. The area within the white lines is the excluded region near the source. An accurate and robust calibration curve using the green channel of Gafchromic film can be easily created for use in validating HDR treatment planning systems. A single dwell position treatment plan verified with film dosimetry can serve to document the accuracy of all the theoretical model parameters. The methodology presented will be useful in implementing more comprehensive quality assurance on treatment planning systems and may even be extended in future work to verify patient-specific dose distributions.

SU‐E‐J‐56: Dosimetric Impact of Patient Roll Variability in Peripheral‐Lung SBRT

Purpose:Positional variation about the patient longitudinal axis may introduce deleterious effects to clinical treatment plans, the effects of which are exacerbated by hypo‐fractionated treatment regimens. Our goal is to evaluate the dosimetric impact of rotational setup error in stereotactic body radiation therapy (SBRT) for peripherally‐located lung lesions.Methods:Six patients who had been treated for peripheral lung lesions using varying SBRT techniques were analyzed. Each patient's planning CT was processed using custom in‐house software to simulate 1, 3, and 5 degree roll in both directions about their longitudinal axis. The original plan was fused and re‐calculated. Agreement between planned and error‐simulated dose distributions was quantified using 3%, 3mm criteria for the 3‐dimensional gamma metric, and changes in target coverage and adjacent normal anatomy were also quantified.Results:Increased rotational setup error resulted in reduced target coverage, with changes in equivalent uniform dose ranging from −1.24% to −6.11% for simulated roll greater than or equal to 3 degrees, increasing with magnitude of error and skewed for roll directed toward the involved side. Changes to normal tissue metrics are linearly related to simulated roll, with rib D(1cc) changes ranging from −4.23% to 3.27% relative to the original treatment plan. Error‐simulated dose distributions demonstrate agreement by the gamma metric in the planned target of 72.56%, 81.02%, and 99.52% for 5, 3, and 1 degree simulated roll scenarios, respectively. Corresponding agreement for the regional dose in excess of 50% of the maximum were 65.71%,70.77%, and 97.88%.Conclusion:This study suggests that the ability to detect and compensate for rotational positioning variations beyond a threshold level may improve care for patients with peripherally‐located lesions beyond that which may be accomplished using only translational axes. Positional error of one degree of roll had little influence, though three or more degrees of error introduced significant discrepancy.

SU‐E‐J‐87: Improvements to the Computational Environment for Radiotherapy Research

Purpose: To present new tools in CERR to analyze image registration, as well as other software updates and additions. Method: Two new features were added to CERR to evaluate the accuracy of image registration: (1) gamma‐based metric for comparing base and registered scans: All scan quantitative scan‐comparison methods rely on using image intensities. This means, these methods cannot quantify spatial misalignment between base and registered images. The gamma metric (similar to the ones used for dose comparisons) was modified for image‐comparisons. (2) a color‐coded mirror grid, based on mutual information or mean‐square‐error metrics: Color‐coding blocks within the grid help users focus on the well vs. poorly registered areas. In addition to these features, other noteworthy additions to CERR include (1) the ability to use colormaps for scans. This is an improvement over grayscale‐only scan display and is particularly useful for images produced in nuclear medicine. (2) the ability to change scan‐windowing by dragging the mouse along a view, similar to PACS viewers. Interactive windowing is particularly useful while delineating structures. (3) The ability to import Image‐annotations. The DICOM modality PR is used to store graphical and text annotations for images. These are widely used in diagnostic radiology and could be extracted from PACS. CERR can now imports these objects and store them under the planC{indexS.GSPS} element. Results: A 3‐d gamma metric was implemented in CERR for comparing image registration. It was evaluated on prostate cancer registration with criteria of 4mm and 50HU. The fused images were compared against the Gamma metric map. It was observed that the regions that matched up poorly had gamma value above 1. Conclusion: A novel gamma based approach to compare image registration allows for specification of acceptance criteria. Features added to CERR increase its utility in ART, registration, segmentation and collaboration.

SU-E-T-133: The Accuracy of Dose Calculations for the Brainlab Imaging Couch Top with the Eclipse Treatment Planning System

Purpose: To evaluate the accuracy of the Brainlab imaging couch top model within the Eclipse treatment planning system for dose calculations. Methods: A 25cm3 polystyrene phantom was irradiated at 6MV with an ion chamber (IC) approximately 12cm deep. A second setup was constructed for film collection of depth dose data to the surface of the phantom. Irradiations were performed with and without the couch in the beam path and compared to the calculated dose using Eclipse version 10 and AAA (Varian Medical Systems, Palo Alto, CA). Dose and water equivalent depth (WED) were calculated using the standard Brainlab couch model (−1000HU interior, −300HU surface) and a modified model (−890HU interior, +700HU surface) taken from the iPlan treatment planning system. A comparison was performed using triple channel dosimetry in FilmQA Pro (Ashland, Wayne, NJ) with a gamma metric using 2%/2mm as a tolerance. Results: The mean HU values on CBCT were −925HU in the interior and −425HU on the surface of the couch. WED through the couch at the couch-phantom surface was 0.47cm without the adjusted HU values and 1.4cm with the modified values. This agreed better with the Brainlab published WED of 1.2cm. IC agreement to the treatment plan was 1.58% for the AP field and −1.78% for the PA field with the default HU and 1.16% with the adjusted HU values. Film depth dose measurements showed a decrease in gamma from 99.8% to 94.4% points passing with the default values. Dose at the couch - phantom interface was approximately 20% different on film for the default values and 1% for the adjusted HU values. Conclusion: The default couch HU values were found to be similar to what was observed on CT. However, for our configuration they proved to be insufficient in accurately calculating dose, especially dose to the phantom surface.

TH-A-137-08: On the Validity of Target Density Overrides for RapidArc Lung SBRT Treatment Planning

Purpose: It is common practice to calculate lung SBRT plans on free breathing CT scans. Recently, literature has suggested using time average scans or ITV density overridden scans to more accurately calculate dose in lung. We evaluate each of these techniques. Methods: A free breathing (FB) and 4D-CT were acquired for a Quasar Motion Phantom with lung insert and 3 cm spherical target (Modus Medical Devices Inc, Canada). Target motion was set to ± 1cm parallel to CT table travel at 16 cycles per minute. An ITV was defined on the 4D-CT with a PTV defined as a 5 mm expansion. RapidArc plans giving 1000 cGy to 96–97% of the PTV in two arcs were created on four image sets: the time average, FB, FB with the ITV set to tissue density, and FB with the PTV set to tissue density. Plans were delivered on a Varian TrueBeam STx (Varian Medical Systems, Palo Alto, CA) and measured on EBT3 film placed in the phantom target. Reference dose distributions that accounted for film motion were created for each plan by shifting the dose in 2 mm steps over the range of motion and convolving these shifted doses. The film was analyzed in with a gamma metric of 1mm/1%. Results: The passing gamma percentages were 77.9%, 83.6%, 89.1%, and 92.7% for the FB, time average, ITV, and PTV plans, respectively. Profiles along the direction of travel showed hot spots of 5-10% above the predicted dose just outside the target for the FB and average plans. The ITV plan showed similar hot spots ∼2% above predicted dose. The PTV profile was in good agreement with predicted dose. Conclusion: Though this study is a greatly simplified case, it shows that planning to a tissue density target in lung may produce the most accurate dose predictions.

Evaluation of the accuracy of 3DVH software estimates of dose to virtual ion chamber and film in composite IMRT QA

A novel patient-specific intensity modulated radiation therapy (IMRT) QA system, 3DVH software and mapcheck 2, purports to be able to use diode array-measured beam doses and the patient's DICOM RT plan, structure set, and dose files to predict the delivered 3D dose distribution in the patient for comparison to the treatment planning system (TPS) calculated doses. In this study, the composite dose to an ion chamber and film in phantom predicted by the 3DVH and mapcheck 2 system is compared to the actual measured chamber and film doses. If validated in this context, then 3DVH can be used to perform an equivalent dose analysis as that obtained with film dosimetry and ion chamber-based composite IMRT QA. This is important for those losing their ability to perform film dosimetry for true composite IMRT QA and provides a measure of confidence in the accuracy of 3DVH 3D dose calculations which may replace phantom-based IMRT QA. The dosimetric results from 15 consecutive patient-specific IMRT QA tests performed by composite field irradiation of ion chamber and EDR2 film in a solid water phantom were compared to the predicted doses for those virtual detectors based on the calculated 3D dose by the 3DVH software using mapcheck 2 measured doses of each beam within each plan. For each of the 15 cases, immediately after performing the ion chamber plus film measurements, the mapcheck 2 was used to measure the dose for each beam of the plan. The dose to the volume of the virtual ion chamber and the dose distribution in the plane of the virtual film calculated by the 3DVH software was extracted. The ratio of the measured to 3DVH or eclipse-predicted ion chamber doses was calculated. The same plane in the phantom measured using film and calculated with eclipse was exported from 3DVH and the 2D gamma metric was used to compare the relationship between the film doses and the eclipse or 3DVH predicted planar doses. Also, the 3D gamma value was calculated in the 3DVH software which compares the eclipse dose to the 3DVH predicted dose distribution. For the 2D and 3D gamma metrics, 2% dose and 2 mm distance to agreement (DTA) were used. In addition, a simple dose difference was performed using either a 2% or 3% dose difference tolerance. The mean ratio ± standard deviation of the measured vs 3DVH or vs eclipse-predicted dose to the ion chamber was 1.013 ± 0.015 and 1.003 ± 0.012, respectively. For 3DVH vs eclipse, the mean percentage of pixels failing the 3D gamma metric was 1.2% ± 1.4% while the failure rate for the 2D gamma metric was 1.1% ± 0.9%. When either 3DVH or eclipse was compared to EDR2 film, the gamma failure rate was 2.3% ± 2.0% and 1.6% ± 1.7%, respectively. Mean dose difference failures were 9%-27% ± 5%-15% for 2 or 3% dose difference tolerances, depending on the combination of systems tested. No statistically significant differences were found for any of the planar dosimetric comparisons. 3DVH + mapcheck 2 predicts the same absolute dose, the percent of pixels failing the gamma metric, and the percent of pixels failing 2% or 3% dose difference tolerance tests as one would have obtained had one made measurements in solid water phantom using an ion chamber and coronal film instead of a diode array. This is also a necessary although not sufficient condition for validation of the accuracy of 3DVH predictions of the 3D dose using beam-by-beam measurements.

SU‐D‐BRA‐05: Evaluation of the Accuracy of 3DVH+Mapcheck2 Predictions of Dose to Virtual Ion Chamber and Film in Composite IMRT QA

Purpose: Two common methods of performing patient‐specific IMRT QA are the composite dose (ion chamber and film) and beam‐by‐beam dose measurement (diode array). Neither method by itself gives a complete enough measure of the dose distribution. DVH+Mapcheck2 was evaluated for its ability to provide both types of doses with just the array measurement. Methods: The dosimetric results from 15 patient‐specific IMRT QA tests performed by composite irradiation of ion chamber and EDR film (in coronal plane) in phantom were compared to the predicted doses for the same locations based on Mapcheck 2 measured doses of each beam within each plan and interpretation by 3DVH software. The ratio of the measured to predicted ion chamber doses was calculated and the Gamma metric was used to compare the relationship between the film doses to the Eclipse or 3DVH predicted planar doses. Also, the 3D gamma value was calculated in the 3DVH software which compares the Eclipse to 3DVH predicted dose distributions. For the 2D and 3D Gamma metrics, 2% dose and 2mm DTA were used. In addition, a simple dose difference was performed using either a 2% or 3% dose difference tolerance. Results: The mean ratio(SD) of the chamber to 3DVH predicted dose was 1.013(.015) while the ratio of the measured to Eclipse calculated dose to the chamber was 1.003(.012). The mean percentage of pixels failing the 3D gamma=1 metric was 1.2%(1.4%) while the failure rate for the 2D Gamma metric (3DVH to Eclipse) was 1.1%(1%). The Gamma failure rate for the film vs. Eclipse was 1.6%(2%). Mean dose differences, highly sensitive measures of dosimetric difference when not paired with DTA, were not significantly different between 3DVH+Mapcheck2 or film vs. Eclipse. Conclusions: 3DVH+Mapcheck2 accurately predicts the doses one would have measured had one instead used an ion chamber and coronal film in plastic phantom.Sun Nuclear has provided support for this project

SU-FF-T-522: Incorporation of Setup Uncertainties Through Gaussian Convolution of the Dose Distribution

Purpose: The purpose of this work was to evaluate the simulation of setup uncertainty by convolve the dose distribution calculated by the treatment planning system with a Gaussian kernel. Materials and Methods: Setup uncertainty was explicitly simulated by targeting many beams at isocenters that were displaced in such a way that the original isocenter was surrounded symmetrically in R3 space, imitating a patient displacement relative to the position in the original treatment plan. The weighting of the displaced beams was made proportional to a Gaussian probability distribution for several choices of σ. The dose distribution created by multiple beams was compared to the dose distribution created by a Gaussian convolution of the original beam setup using the gamma metric. A commercial radiation treatment planning system (Pinnacle3, Philips Medical Systems, Milpitas CA) was programmed to collect 2,197 beam dose grids for each treatment beam in each of 3 evaluated treatment plans (totaling 32,955 beam dose grids) in order to collect a large sample of displaced isocenters. Subsequent analysis was performed using custom software in MATLAB (The MathWorks, Inc., Natick, MA). Results: For all choices of σ, the gamma values associated with voxels of the plan's regions of interest are almost entirely below 0.4 to 0.5, with only several voxels from many thousands having values from 1.0 to 1.4. Conclusions: The application of a Gaussian convolution is an accurate method of approximating the final dose as it would be delivered by multiple beams randomly displaced from their intended isocenter in the patient, implying that this application is suitable for simulating the effect of setup uncertainty on the treatment plan. This work has been supported in part by a Sponsored Research Agreement with Philips Healthcare.

SU‐FF‐T‐195: Effect of Respiratory Gating On the Dose Distributions of Dynamic Wedge and IMRT Treatment Fields

Purpose: To determine the magnitude of error introduced into treatment delivery due to timing effects between respiratory‐gated delivery and dynamic beam delivery (enhanced dynamic wedge (EDW) and step‐and‐shoot intensity modulated radiation therapy (IMRT)). Method and Materials: EDW and IMRT fields were delivered on a linear accelerator (Trilogy, Varian Medical Systems). Gating of the beam was achieved using a commercial respiratory monitoring system (RPM, Varian Medical Systems), and respiratory motion was simulated using a commercial respiratory motion phantom (RPM phantom, Varian Medical Systems). All fields were delivered with no gating, a 1‐mm gating window, and a 2‐mm gating window.Film and ion chamber measurements were made for both EDW and IMRT fields in a water‐equivalent IMRT QA phantom, which was stationary during all measurements. Film measurements used EDR‐2 film, while ion chamber measurements were made with a 0.04 cc volume pin‐point chamber. Leakage current was subtracted from the ion chamber readings.Wedged fields of 15 and 60 degrees were delivered at a dose rate of 600 MU/min. Results were evaluated for the full IMRT plan and for one of the individual beams, at dose rates of 400 and 600 MU/min. Results: All ion chamber measurements were analyzed as percent difference between the gated (1‐mm or 2‐mm gating window) and non‐gated delivery. All differences were less than 1%. There were no significant differences between 1‐mm and 2‐mm gated delivery. Film results, as analyzed from isodose curves and the gamma metric, showed no substantial differences between gated and non‐gated beam delivery. Conclusion: The use of gated treatment delivery of EDW and step‐and‐shoot IMRT fields does not introduce clinically significant differences into the resulting dose distributions, as quantified by film and ion chamber measurements.