4DCT Image Sets Research Articles

Assessment of Daily Target Motion to Optimize ITV Margins in Online Adaptive Plan for Liver and Pancreatic SBRT on MR-LINAC

Accurate measurement of daily target motion on online plan adaptation is essential for re-optimizing the internal target volume (ITV) margins. In this study, by analyzing 2D cine MR images acquired in a 1.5T MR-Linac, we assessed daily target motion and its inter-fractional variations, which is then used to customize ITV margins for MR-based online adaptive plan of liver and pancreatic stereotactic body radiotherapy (SBRT). Twenty-threepatients, enrolled in this study, underwent MR-guided online adaptive SBRT on 1.5 T MR-Linac: liver (n = 12 without compression belt (CB), n = 2 with CB) and pancreas (n = 9 with CB). Target motion in a 2D coronal and sagittal MR images (> 200 cine images to cover multiple breathing cycles), as well as a set of 4D-CT images during the simulation, was analyzed to determine patient specific ITV margins for a reference plan. On every fraction of online adaptive SBRT, the 2D MR images were exported to commercially available software along with a planning MR images: a T1-wighted MR image set for pancreas and or a navigated T2-wighted MR image set for liver. Daily target motion was then measured by overlaying the orthogonal cine image set on the T1- or T2-wighted image set. The ITV margin was adjusted when required prior to online plan optimization RESULTS: Thisstudy successfully assessed the daily target motion and inter-fractional variations in 23 liver or pancreatic cancer patients. We found a discrepancy of daily target motion of orthogonal 2D cine MR images from that of 4D-CT, required to adjust the ITV margin for online plan adaptation. In liver SBRT without CB, 15 out of 45 fractions from 12 patients (33.3 %) went through adjusted ITV margin. The tumor motion was large up to a 2.2 cm along the inferior direction in a patient without CB and thus the patient was treated on a conventional Linac with the active breathing control (ABC) unit for breath-hold. On the other hand, no adjustments were necessary for 2 liver patients with CB. In 9 patients with pancreatic cancers, we found the inter-fractional ITV variations in 3 out of 45 fractions (6.7 %) adjusted the ITV margin. The daily target motion in the pancreatic patients with CB was more stable and relatively smaller than that of the liver patients. Theorthogonal 2D cine MR images are evidently essential for assessing daily target motion to evaluate the inter-fractional variation and achieve daily ITV-customized liver and pancreatic SBRT in the 1.5 T MR-Linac. The customized ITV approach will improve the treatment outcome of the MR-guided online adaptive therapy.

Dosimetric Assessment for Real-Time Anatomical Motion during MR-Guided Radiation Therapy by MR-LINAC

<h3>Purpose/Objective(s)</h3> To develop a method to assess the dosimetric impact of real-time anatomical motion during MR-guided LINAC radiotherapy, and to validate adaptive treatment plans. <h3>Materials/Methods</h3> In MR-guided adaptive RT, the treatment plan is adapted to account for changes in the patient's anatomy before each fraction. This adaptive plan is presumably effective if the target motion throughout the beam-on duration is within the PTV margin. However, the amount of dosimetric impact from the internal motion during dose delivery is undetermined. It is important to precisely assess this dosimetric impact so that the current adaptive plan can be validated or be possibly improved by better margin definition. To reconstruct the 3D anatomical motion, three steps were conducted: (1) the real-time two-dimensional anatomical motion during dose delivery was captured by 2D monitoring MR images (mMR) continuously. The mMR images were time stamped and grouped into ten phases in a respiration cycle, which will then be synchronized to each segment (or control point) individually according to the log file. (2) A set of pre-acquired 4DCT images were separated into ten phases of CT volumes. All phases of CT volume were then registered to the 3D planning MR images (pMR) of the patient in the treatment position. (3) The registered CT volumes were correlated to the processed mMR images in a phase-to-phase basis and correlated to the segments individually. An in-house Monte Carlo method developed was used to simulate the dose distribution for each segment individually in the related CT volumes. The accuracy of the in-house MC tool is confirmed by benchmarking to the commissioning data in water phantom with difference <1% and standard uncertainty < 0.8 %. The delivered dose was then estimated by accumulating the dose from all segments in the plan. On the basis of this information, dose-volume histograms and 3D dose distributions were calculated to assess the discrepancy from the planned dose caused by anatomical motion. In this project, three prostate, pancreas, and liver cancer patients were retrospectively studied. The hypothesis is that the margin defined during adaptive planning is dosimetrically effective after taking into account the internal motions during dose delivery. <h3>Results</h3> The characteristic dose indices such as the minimum dose to the gross tumor volume (D2cc) and the maximum dose to the surrounding OARs are slightly different (both < 2%) for individual fractions but this effect was reduced after multiple fractions and no significant difference observed (<1%) in DVH. <h3>Conclusion</h3> The margin definition of adaptive planning was dosimetrically valid among three cases. The same setup and methodology of this study can be employed to test the patients with greater respiratory motion (such as lung cancers) or those treated with stereotactic body radiation therapy, which would benefit more from this assessment because the dose variation will be less faded over multiple fractions.

Modeling Respiratory Signals by Deformable Image Registration on 4DCT Lung Images.

The lung organ of human anatomy captured by a medical device reveals inhalation and exhalation information for treatment and monitoring. Given a large number of slices covering an area of the lung, we have a set of three-dimensional lung data. And then, by combining additionally with breath-hold measurements, we have a dataset of multigroup CT images (called 4DCT image set) that could show the lung motion and deformation over time. Up to now, it has still been a challenging problem to model a respiratory signal representing patients' breathing motion as well as simulating inhalation and exhalation process from 4DCT lung images because of its complexity. In this paper, we propose a promising hybrid approach incorporating the local binary pattern (LBP) histogram with entropy comparison to register the lung images. The segmentation process of the left and right lung is completely overcome by the minimum variance quantization and within class variance techniques which help the registration stage. The experiments are conducted on the 4DCT deformable image registration (DIR) public database giving us the overall evaluation on each stage: segmentation, registration, and modeling, to validate the effectiveness of the approach.

Prediction of Large Change in Lung Tumor Motion During Treatment Course: An Ensemble Learning Method With Enhanced Positive Predictive Values

During lung cancer radiotherapy, tumor motion magnitude may change, resulting in either inadequate margins or non-ideal dose conformality (margins are unnecessarily too big). To predict large motion change using only patient-specific clinical information and the simulation 4DCT images, we proposed an ensemble learning model (ELM) with robust feature selection (FS). For this specific study, we minimized the false negative prediction so that we will not miss patients who might have a subsequent large change in tumor motion. Thus, we employed positive predictive values (PPV) as the evaluation metric instead of commonly used accuracy scores.Seventeen features were collected including both clinical factors and imaging features. The motion change is defined by motion difference in SI direction between the simulation 4DCT and mid-course repeat 4DCT with the clinical threshold set as 3 mm. The dataset consisted of 140 patients, where 31 showed large tumor motion change. Four hundred seventy-two sets of 4DCT images were analyzed in this study. The ELM adopts 5 conventional predictive models (logistic regression (LR), decision tree (DT), support vector machine (SVM), random forest (RF), and XGBoost (XGB)) as base predictors. Then, the results are employed as inputs for a two-layer neural network to obtain a final prediction. To enhance the feature sensitivity and to reduce data noise, we developed a robust FS strategy with sequential backward selection. Each criterion score is an average from 5-fold cross-valuation, and the target feature of each round is based on 1000 repeated experiments. The peak of the criterion score after sequential removal of the least sensitive features was used to select the optimal feature set.By employing PPV as the criterion function, all the base predictors ranked initial SI tumor motion as the most important feature. ROC curve with selected sensitive features consistently outperformed the one with all features. For ELM, the AUCs (mean std. ± dev.) were 0.63 ± 0.04 when all features were included and increased to 0.72 ± 0.02 for PPV-based selected features only. The prediction comparisons were performed for all base predictors and ELM with 100 repetitions. The PPV results (Table 1) demonstrated the proposed ELM obtained superior performance than each base predictor.We proposed an ELM to identify patients with potential large tumor motion change using only information prior to the first treatment. The ELM showed superior performance with PPV-driven metrics. We also developed a feature selection strategy for each base predictor, and the consistent superiority was demonstrated by ROC results. In future work, additional information from on-treatment evaluation CTs may further improve the prediction.

Assessing the Interplay Effect Based on a Precise Machine-Specific Delivery Sequence and Time for Cyclotron Accelerator Proton Therapy System

Proton Pencil Beam Scanning (PBS) delivery sequence is critical in the evaluation of the motion interplay effect. We proposed an experimental approach to build a precise machine-specific model for standard delivery, volumetric repainting delivery, and layer repainting delivery based on a cyclotron accelerator system and assessed the interplay effect quantitatively.Test fields and clinical treatment plans were used to derive each beam delivery parameters experimentally. The machine delivery log files were retrospectively analyzed to model beam delivery parameters, such as energy layer switching time (ELST), spot switching time, and spot drill time, for standard, volumetric and layer repainting delivery techniques. To quantitatively evaluate the interplay effect, a series of digital thoracic 4DCT image sets were used through 4D dynamic dose accumulation method. A small tumor in radius 5mm (labeled by R5) and a large tumor in radius 2cm (R20) phantom were created to simulate the mobile target with different sizes. Target motion was simulated based on the periodic respiratory motion with an amplitude 5 mm. The breathing cycle is 4s. Two-field Single Field Uniform Dose (SFUD) plans were generated using PA and lateral field. Different delivery techniques such as standard delivery, volumetric repainting delivery (n = 2,3,4) and layer repainting delivery (n = 2,3,5,25) were simulated based on the machine-specific delivery sequence model and a simplified model published by West German Proton Therapy Essen (WPE). D99 (Dose received by 99% of target volume) of the target is used to estimate the delivery accuracy.The results showed that the WPE model's spot delivery sequence deviated from the log file significantly compared to the machine-specific model, which results the difference in the interplay effect evaluation. More specifically, for a lung treatment plan with target size (65 mm³) and layer repainting 25 times, the difference is about 21.01% due to a coarse spot scanning speed modeling and an oversimplified assumption of a constant ELST from WPE model. Such a difference also resulted in a very different interplay effects estimation between the two models even though both institutions used the same proton system from IBA and calculated using the same 4DCT imaging set.A precise machine-specific delivery sequence model is highly recommended to ensure an accurate estimation of mobile target treatment's interplay effect for each institution.

Lung tumor segmentation in 4D CT images using motion convolutional neural networks.

Manual delineation on all breathing phases of lung cancer 4D CT image datasets can be challenging, exhaustive, and prone to subjective errors because of both the large number of images in the datasets and variations in the spatial location of tumors secondary to respiratory motion. The purpose of this work is to present a new deep learning-based framework for fast and accurate segmentation of lung tumors on 4D CT image sets. The proposed DL framework leverages motion region convolutional neural network (R-CNN). Through integration of global and local motion estimation network architectures, the network can learn both major and minor changes caused by tumor motion. Our network design first extracts tumor motion information by feeding 4D CT images with consecutive phases into an integrated backbone network architecture, locating volume-of-interest (VOIs) via a regional proposal network and removing irrelevant information via a regional convolutional neural network. Extracted motion information is then advanced into the subsequent global and local motion head network architecture to predict corresponding deformation vector fields (DVFs) and further adjust tumor VOIs. Binary masks of tumors are then segmented within adjusted VOIs via a mask head. A self-attention strategy is incorporated in the mask head network to remove any noisy features that might impact segmentation performance. We performed two sets of experiments. In the first experiment, a five-fold cross-validation on 20 4D CT datasets, each consisting of 10 breathing phases (i.e., 200 3D image volumes in total). The network performance was also evaluated on an additional unseen 200 3D images volumes from 20 hold-out 4D CT datasets. In the second experiment, we trained another model with 40 patients' 4D CT datasets from experiment 1 and evaluated on additional unseen nine patients' 4D CT datasets. The Dice similarity coefficient (DSC), center of mass distance (CMD), 95th percentile Hausdorff distance (HD95 ), mean surface distance (MSD), and volume difference (VD) between the manual and segmented tumor contour were computed to evaluate tumor detection and segmentation accuracy. The performance of our method was quantitatively evaluated against four different methods (VoxelMorph, U-Net, network without global and local networks, and network without attention gate strategy) across all evaluation metrics through a paired t-test. The proposed fully automated DL method yielded good overall agreement with the ground truth for contoured tumor volume and segmentation accuracy. Our model yielded significantly better values of evaluation metrics (p<0.05) than all four competing methods in both experiments. On hold-out datasets of experiment 1 and 2, our method yielded DSC of 0.86 and 0.90 compared to 0.82 and 0.87, 0.75 and 0.83, 081 and 0.89, and 0.81 and 0.89 yielded by VoxelMorph, U-Net, network without global and local networks, and networks without attention gate strategy. Tumor VD between ground truth and our method was the smallest with the value of 0.50 compared to 0.99, 1.01, 0.92, and 0.93 for between ground truth and VoxelMorph, U-Net, network without global and local networks, and networks without attention gate strategy, respectively. Our proposed DL framework of tumor segmentation on lung cancer 4D CT datasets demonstrates a significant promise for fully automated delineation. The promising results of this work provide impetus for its integration into the 4D CT treatment planning workflow to improve the accuracy and efficiency of lung radiotherapy.

Four-dimensional dose calculations for dynamic tumour tracking with a gimbal-mounted linear accelerator.

PurposeIn this study we present a novel method for re‐calculating a treatment plan on different respiratory phases by accurately modeling the panning and tilting beam motion during DTT (the “rotation method”). This method is used to re‐calculate the dose distribution of a plan on multiple breathing phases to accurately assess the dosimetry.MethodssIMRT plans were optimized on a breath hold computed tomography (CT) image taken at exhale (BHexhale) for 10 previous liver stereotactic ablative radiotherapy patients. Our method was used to re‐calculate the plan on the inhale (0%) and exhale (50%) phases of the four‐dimensional CT (4DCT) image set. The dose distributions were deformed to the BHexhale CT and summed together with proper weighting calculated from the patient’s breathing trace. Subsequently, the plan was re‐calculated on all ten phases using our method and the dose distributions were deformed to the BHexhale CT and accumulated together. The maximum dose for certain organs at risk (OARs) was compared between calculating on two phases and all ten phases.ResultsIn total, 26 OARs were examined from 10 patients. When the dose was calculated on the inhale and exhale phases six OARs exceeded their dose limit, and when all 10 phases were used five OARs exceeded their limit.ConclusionDynamic tumor tracking plans optimized for a single respiratory phase leave an OAR vulnerable to exceeding its dose constraint during other respiratory phases. The rotation method accurately models the beam’s geometry. Using deformable image registration to accumulate dose from all 10 breathing phases provides the most accurate results, however it is a time consuming procedure. Accumulating the dose from two extreme breathing phases (exhale and inhale) and weighting them properly provides accurate results while requiring less time. This approach should be used to confirm the safety of a DTT treatment plan prior to delivery.

Commissioning a four-dimensional Computed Tomography Simulator for minimum target size due to motion in the Anterior-Posterior direction: a procedure and treatment planning recommendations.

The purpose of this work is to develop a procedure for commissioning four‐dimensional computed tomography (4DCT) algorithms for minimum target reconstruction size, to quantify the effect of anterior–posterior (AP) motion artifacts on known object reconstruction for periodic and irregular breathing patterns, and to provide treatment planning recommendations for target sizes below a minimum threshold. A mechanical platform enabled AP motion of a rod and lung phantom during 4DCT acquisition. Static, artifact‐free scans of the phantoms were first acquired. AP sinusoidal and patient breathing motion was applied to obtain 4DCT images. 4DCT reconstruction artifacts were assessed by measuring the apparent width and angle of the rod. Comparison of known tumor diameters and volumes between the static image parameters with the 4DCT image sets was used to quantify the extent of AP reconstruction artifact and contour deformation. Examination of the rod width, under sinusoidal motion, found it was best represented during the inhale and exhale phases for all periods and ranges of motion. From the gradient phases, the apparent width of the rod decreased with increasing amplitude and decreasing period. The rod angle appeared larger on the reconstructed images due to the presence of motion artifact. The apparent diameters of the spherical tumors on the gradient phases were larger/equivalent than the true values in the AP/LR direction, respectively, while the exhale phase consistently displayed the spheres at the approximately correct diameter. The Eclipse calculated diameter matched closely with the true diameter on the exhale phase and was found to be larger on the inhale, MIP, and Avg scans. The procedure detailed here may be used during the acceptance and commissioning period of a computed tomography simulator or retroactively when implementing a SBRT program to determine the minimum target size that can be reliably reconstructed.

Impact of planned dose reporting methods on Gamma pass rates for IROC lung and liver motion phantoms treated with pencil beam scanning protons

PurposeThe purpose of this study is to evaluate the impact of two methods of reporting planned dose distributions on the Gamma analysis pass rates for comparison with measured 2D film dose and simulated delivered 3D dose for proton pencil beam scanning treatment of the Imaging and Radiation Oncology Core (IROC) proton lung and liver mobile phantoms.Methods and materialsFour-dimensional (4D) computed-tomography (CT) image sets were acquired for IROC proton lung and liver mobile phantoms, which include dosimetry inserts that contains targets, thermoluminescent dosimeters and EBT2 films for plan dose verification. 4DCT measured fixed motion magnitudes were 1.3 and 1.0 cm for the lung and liver phantoms, respectively. To study the effects of motion magnitude on the Gamma analysis pass rate, three motion magnitudes for each phantom were simulated by creating virtual 4DCT image sets with motion magnitudes scaled from the scanned phantom motion by 50, 100, and 200%. The internal target volumes were contoured on the maximum intensity projection CTs of the 4DCTs for the lung phantom and on the minimum intensity projection CTs of the 4DCTs for the liver phantom. Treatment plans were optimized on the average intensity projection (AVE) CTs of the 4DCTs using the RayStation treatment planning system. Plan doses were calculated on the AVE CTs, which was defined as the planned AVE dose (method one). Plan doses were also calculated on all 10 phase CTs of the 4DCTs and were registered using target alignment to and equal-weight-summed on the 50% phase (T50) CT, which was defined as the planned 4D dose (method two). The planned AVE doses and 4D doses for phantom treatment were reported to IROC, and the 2D-2D Gamma analysis pass rates for measured film dose relative to the planned AVE and 4D doses were compared. To evaluate motion interplay effects, simulated delivered doses were calculated for each plan by sorting spots into corresponding respiratory phases using spot delivery time recorded in the log files by the beam delivery system to calculate each phase dose and accumulate dose to the T50 CTs. Ten random beam starting phases were used for each beam to obtain the range of the simulated delivered dose distributions. 3D-3D Gamma analyses were performed to compare the planned 4D/AVE doses with simulated delivered doses.ResultsThe planned 4D dose matched better with the measured 2D film dose and simulated delivered 3D dose than the planned AVE dose. Using planned 4D dose as institution reported planned dose to IROC improved IROC film dose 2D-2D Gamma analysis pass rate from 92 to 96% on average for three films for the lung phantom (7% 5 mm), and from 92 to 94% in the sagittal plane for the liver phantom (7% 4 mm), respectively, compared with using the planned AVE dose. The 3D-3D Gamma analysis (3% 3 mm) pass rate showed that the simulated delivered doses for lung and liver phantoms using 10 random beam starting phases for each delivered beam matched the planned 4D dose significantly better than the planned AVE dose for phantom motions larger than 1 cm (p ≤ 0.04).ConclusionsIt is recommended to use the planned 4D dose as the institution reported planned dose to IROC to compare with the measured film dose for proton mobile phantoms to improve film Gamma analysis pass rate in the IROC credentialing process.

Technical Note: Deriving ventilation imaging from 4DCT by deep convolutional neural network.

Ventilation images can be derived from four-dimensional computed tomography (4DCT) by analyzing the change in HU values and deformable vector fields between different respiration phases of computed tomography (CT). As deformable image registration (DIR) is involved, accuracy of 4DCT-derived ventilation image is sensitive to the choice of DIR algorithms. To overcome the uncertainty associated with DIR, we develop a method based on deep convolutional neural network (CNN) to derive ventilation images directly from the 4DCT without explicit image registration. A total of 82 sets of 4DCT and ventilation images from patients with lung cancer were used in this study. In the proposed CNN architecture, the CT two-channel input data consist of CT at the end of exhale and the end of inhale phases. The first convolutional layer has 32 different kernels of size 5×5×5, followed by another eight convolutional layers each of which is equipped with an activation layer (ReLU). The loss function is the mean-squared-error (MSE) to measure the intensity difference between the predicted and reference ventilation images. The predicted images were comparable to the label images of the test data. The similarity index, correlation coefficient, and Gamma index passing rate averaged over the tenfold cross validation were 0.880±0.035, 0.874±0.024, and 0.806±0.014, respectively. The results demonstrate that deep CNN can generate ventilation imaging from 4DCT without explicit deformable image registration, reducing the associated uncertainty.

Four-Dimensional Dose Reconstruction for Lung Cancer in Volumetric Modulated Arc Therapy

The purpose of this study is to evaluate actual dose distribution for the cases with different respiratory motion using four-dimensional (4D) dose reconstruction system. In the thorax region, respiratory motion of tumor may lead changes in dose distribution due to interplay effects caused by motions of both multi-leaf collimators (MLCs) and tumors. It is important to assess actual dose distribution in each treatment. The authors have developed a 4D dose reconstruction system based on Linac log data and respiratory wave form during VMAT treatment. The 4D dose reconstruction was performed by reproducing the actual irradiation conditions on the treatment planning system using 4D-CT image sets and various field groups that were divided into multiple fields DICOM RT-Plan files created using log data according to each tumor position based on respiratory wave form and EPID cine images. The authors reported in the previous ASTRO annual meeting about results of the phantom experiment that this system could reconstruct the actual dose distribution within the error of 2%. In this study, the case of one lung cancer patient was evaluated. The patient received the 2-arc VMAT irradiation under free breathing (2Gy / fr). The reconstructed dose distributions with different respiratory motion were compared with each other, as well as the planned dose distribution. The patient was treated using Clinac21EX linear accelerator equipped with aS1000 EPID (Varian) and real time position management system (Varian) for acquisition the respiratory wave form. Dose calculation was performed using Eclipse 10.0.28 (Varian) on 4D-CT images (10 phases), and deformable registration was used to accumulate dose distributions of each respiratory phase. Dose distribution of each respiratory phase varied between different treatment sessions, i.e. different respiratory motion, while the end-of-exhale phase contained the highest point dose, and the phase included 30 to 40% of the total dose. The dose distributions and DVHs after accumulation of all respiratory phases were quite similar through different treatment sessions. Any effects of changes in respiratory conditions could not be observed. Additionally the reconstructed dose distributions and planned dose distribution showed similar shape, and the reconstructed CTV center doses were agreed with the planned dose within 0.5%. The change of the dose distribution was quite small due to small tumor motion. On the other hand, it was confirmed that the irradiation of the treatment plan has been correctly performed using (calculated on) this dose reconstruction system. Using the developed system, 4D dose distribution in the clinical case was successfully reconstructed, and the distribution difference between different respiratory motions was evaluated.

Automatic assessment of average diaphragm motion trajectory from 4DCT images through machine learning

To automatically estimate average diaphragm motion trajectory (ADMT) based on four-dimensional computed tomography (4DCT), facilitating clinical assessment of respiratory motion and motion variation and retrospective motion study. We have developed an effective motion extraction approach and a machine-learning-based algorithm to estimate the ADMT. Eleven patients with 22 sets of 4DCT images (4DCT1 at simulation and 4DCT2 at treatment) were studied. After automatically segmenting the lungs, the differential volume-per-slice (dVPS) curves of the left and right lungs were calculated as a function of slice number for each phase with respective to the full-exhalation. After 5-slice moving average was performed, the discrete cosine transform (DCT) was applied to analyze the dVPS curves in frequency domain. The dimensionality of the spectrum data was reduced by using several lowest frequency coefficients (ƒv) to account for most of the spectrum energy Multiple linear regression (MLR) method was then applied to determine the weights of these frequencies by fitting the ground truth—the measured ADMT, which are represented by three pivot points of the diaphragm on each side. The ‘leave-one-out’ cross validation method was employed to analyze the statistical performance of the prediction results in three image sets: 4DCT1, 4DCT2, and 4DCT1 + 4DCT2. Seven lowest frequencies in DCT domain were found to be sufficient to approximate the patient dVPS curves (R = 91%–96% in MLR fitting). The mean error in the predicted ADMT using leave-one-out method was 0.3 ± 1.9 mm for the left-side diaphragm and 0.0 ± 1.4 mm for the right-side diaphragm. The prediction error is lower in 4DCT2 than 4DCT1, and is the lowest in 4DCT1 and 4DCT2 combined. This frequency-analysis-based machine learning technique was employed to predict the ADMT automatically with an acceptable error (0.2 ± 1.6 mm). This volumetric approach is not affected by the presence of the lung tumors, providing an automatic robust tool to evaluate diaphragm motion.

An Efficiency and Feasibility Trial Comparing Novel Contouring and Dosimetric Techniques in 4DCT Respiratory Gating Planning for SBRT in NSCLC

The gold standard for tumor delineation in free-breathing SBRT for NSCLC at our facility is a 10-phase binning method to produce each patient’s internal tumor volumes (ITV). We attempted to assess the reproducibility of tumor volume and location in three dimensions when contouring lung tumors with a single ITV (ITVc) on a 4DCT image set showing the extent of tumor displacement with active, cinematic respiratory motion. This novel technique would, in theory, provide a ten-fold decrease in the required total number of tumor volumes to be contoured. Therefore, we also evaluated for any variances in efficiency in regards to contouring time and workflow between these two techniques. Our endpoints were reproducibility of tumor volumes and locations, as well as noninferiority in workflow efficiency. 25 patients (5 patients with tumors in each of the 5 lobes) were selected from patients treated at RPCI. Each patient underwent a 4DCT simulation that provided image sets of ten sequential phases of a complete respiratory cycle. The phases each represented approximately 10% of a complete inhalation and exhalation. The cinematic image set is a compilation of the ten phase-based scans on which contours can be drawn while actively viewing or scrolling through the respiratory cycle. The contrast was set to 240 HU and -160 HU, which allowed for the clearest image on all the phases. Each tumor was contoured as ten separate respiratory-phase specific volumes that were then binned together as an ITV (ITV10) and was then contoured as a single ITV (ITVc) using the cinematic image set. Both techniques were then compared with respect to their volumes as well as to their center of mass and lobe/location. Time spent to complete each method was also recorded and contrasted. Only 8 of the 25 evaluated cases ended up with an acceptable (<5%) difference from the standard 10-phases ITV [range 1.3%-24.4%; median 9.9%]. However, these are very small volumes making even a small change result in large change in percentage. No correlation, either on the geographical location of the tumor or on the tumor volume, could be identified that explains why excessive deviation occurred in certain patients and not in others. The mean difference between the centers of mass was relatively small for tumors in all lobes except the left lower lobe (LLL). The LLL tumor experienced more deviation between the centers of mass, but the difference was still small (0.8 mm). While the ITVc method of tumor volume delineation offers substantial gain in workflow efficiency, variance in volumes compared to the current standard are too significant to report noninferiority. A larger sample size is needed to assess the accuracy of this method. This study is being repeated with additional researchers and clinicians providing contours, as results can rely heavily on user-specific techniques and skill.

TU-EF-304-01: BEST IN PHYSICS (THERAPY): Plan Robustness and Optimality Improvement with 4-Dimensional Robust Optimization for Lung Cancer Patients

Purpose: To develop the first 4D robust optimization (RO) method accounting for respiratory motion and evaluate its potential to improve plan robustness and optimality compared to 3D RO and PTV-based optimization. Methods: A set of 4D CT images are used to track respiratory motion and deformation of tumors and organs. For each of 10 respiration phases, dose distributions for nine different uncertainty scenarios including the nominal one, those incorporating ±5mm setup uncertainties long x, y and z directions and ±3.5% range uncertainties are calculated. All 90 dose distributions are simultaneously optimized to achieve full dose coverage of 10 CTVs and sparing of normal structures. ITV-based 3D RO and PTV-based optimization based on the average CT are also carried for the same patient using same dose volume constrains. After optimization, 4D robustness evaluation was performed for all resulting plans. The CTV coverage and the sparing of normal tissue in 10 phases are evaluated and compared among the three methods. The widths of DVH bands represent the robustness of dose distributions in the structures. Results: For one patient studied so far, the worst case CTV coverage by the prescription dose among all 90 scenarios is: 99% for 4D RO; 88.9% for 3D RO, and 85.2% for PTV based optimization. 4D RO also results in best robustness with the narrowest DVH’ bandwidths for the CTV. 4D and 3D RO have similar organ sparing while PTV based optimization results in worst organ sparing. Conclusion: 4D robust optimization which accounts for anatomy motion and deformation in the optimization process, significantly improves plan robustness and achieves higher quality treatment plans for lung cancer patients. The method is being evaluated for multiple patients with different tumor and motion characteristics. Acknowledge the grant U19, the grant number is 2U19CA021239-35

TU-E-213-03: Tools for Ensuring Quality and Safety

Purpose: To develop the first 4D robust optimization (RO) method accounting for respiratory motion and evaluate its potential to improve plan robustness and optimality compared to 3D RO and PTV-based optimization. Methods: A set of 4D CT images are used to track respiratory motion and deformation of tumors and organs. For each of 10 respiration phases, dose distributions for nine different uncertainty scenarios including the nominal one, those incorporating ±5mm setup uncertainties long x, y and z directions and ±3.5% range uncertainties are calculated. All 90 dose distributions are simultaneously optimized to achieve full dose coverage of 10 CTVs and sparing of normal structures. ITV-based 3D RO and PTV-based optimization based on the average CT are also carried for the same patient using same dose volume constrains. After optimization, 4D robustness evaluation was performed for all resulting plans. The CTV coverage and the sparing of normal tissue in 10 phases are evaluated and compared among the three methods. The widths of DVH bands represent the robustness of dose distributions in the structures. Results: For one patient studied so far, the worst case CTV coverage by the prescription dose among all 90 scenarios is: 99% for 4D RO; 88.9% for 3D RO, and 85.2% for PTV based optimization. 4D RO also results in best robustness with the narrowest DVH’ bandwidths for the CTV. 4D and 3D RO have similar organ sparing while PTV based optimization results in worst organ sparing. Conclusion: 4D robust optimization which accounts for anatomy motion and deformation in the optimization process, significantly improves plan robustness and achieves higher quality treatment plans for lung cancer patients. The method is being evaluated for multiple patients with different tumor and motion characteristics.

TU-E-213-01: Remote Supervision

Purpose: To develop the first 4D robust optimization (RO) method accounting for respiratory motion and evaluate its potential to improve plan robustness and optimality compared to 3D RO and PTV-based optimization. Methods: A set of 4D CT images are used to track respiratory motion and deformation of tumors and organs. For each of 10 respiration phases, dose distributions for nine different uncertainty scenarios including the nominal one, those incorporating ±5mm setup uncertainties long x, y and z directions and ±3.5% range uncertainties are calculated. All 90 dose distributions are simultaneously optimized to achieve full dose coverage of 10 CTVs and sparing of normal structures. ITV-based 3D RO and PTV-based optimization based on the average CT are also carried for the same patient using same dose volume constrains. After optimization, 4D robustness evaluation was performed for all resulting plans. The CTV coverage and the sparing of normal tissue in 10 phases are evaluated and compared among the three methods. The widths of DVH bands represent the robustness of dose distributions in the structures. Results: For one patient studied so far, the worst case CTV coverage by the prescription dose among all 90 scenarios is: 99% for 4D RO; 88.9% for 3D RO, and 85.2% for PTV based optimization. 4D RO also results in best robustness with the narrowest DVH’ bandwidths for the CTV. 4D and 3D RO have similar organ sparing while PTV based optimization results in worst organ sparing. Conclusion: 4D robust optimization which accounts for anatomy motion and deformation in the optimization process, significantly improves plan robustness and achieves higher quality treatment plans for lung cancer patients. The method is being evaluated for multiple patients with different tumor and motion characteristics.

TU-E-213-04: Return On Investment

Purpose: To develop the first 4D robust optimization (RO) method accounting for respiratory motion and evaluate its potential to improve plan robustness and optimality compared to 3D RO and PTV-based optimization. Methods: A set of 4D CT images are used to track respiratory motion and deformation of tumors and organs. For each of 10 respiration phases, dose distributions for nine different uncertainty scenarios including the nominal one, those incorporating ±5mm setup uncertainties long x, y and z directions and ±3.5% range uncertainties are calculated. All 90 dose distributions are simultaneously optimized to achieve full dose coverage of 10 CTVs and sparing of normal structures. ITV-based 3D RO and PTV-based optimization based on the average CT are also carried for the same patient using same dose volume constrains. After optimization, 4D robustness evaluation was performed for all resulting plans. The CTV coverage and the sparing of normal tissue in 10 phases are evaluated and compared among the three methods. The widths of DVH bands represent the robustness of dose distributions in the structures. Results: For one patient studied so far, the worst case CTV coverage by the prescription dose among all 90 scenarios is: 99% for 4D RO; 88.9% for 3D RO, and 85.2% for PTV based optimization. 4D RO also results in best robustness with the narrowest DVH’ bandwidths for the CTV. 4D and 3D RO have similar organ sparing while PTV based optimization results in worst organ sparing. Conclusion: 4D robust optimization which accounts for anatomy motion and deformation in the optimization process, significantly improves plan robustness and achieves higher quality treatment plans for lung cancer patients. The method is being evaluated for multiple patients with different tumor and motion characteristics.

TU-E-213-00: Management Skills for Medical Physicists

Purpose: To develop the first 4D robust optimization (RO) method accounting for respiratory motion and evaluate its potential to improve plan robustness and optimality compared to 3D RO and PTV-based optimization. Methods: A set of 4D CT images are used to track respiratory motion and deformation of tumors and organs. For each of 10 respiration phases, dose distributions for nine different uncertainty scenarios including the nominal one, those incorporating ±5mm setup uncertainties long x, y and z directions and ±3.5% range uncertainties are calculated. All 90 dose distributions are simultaneously optimized to achieve full dose coverage of 10 CTVs and sparing of normal structures. ITV-based 3D RO and PTV-based optimization based on the average CT are also carried for the same patient using same dose volume constrains. After optimization, 4D robustness evaluation was performed for all resulting plans. The CTV coverage and the sparing of normal tissue in 10 phases are evaluated and compared among the three methods. The widths of DVH bands represent the robustness of dose distributions in the structures. Results: For one patient studied so far, the worst case CTV coverage by the prescription dose among all 90 scenarios is: 99% for 4D RO; 88.9% for 3D RO, and 85.2% for PTV based optimization. 4D RO also results in best robustness with the narrowest DVH’ bandwidths for the CTV. 4D and 3D RO have similar organ sparing while PTV based optimization results in worst organ sparing. Conclusion: 4D robust optimization which accounts for anatomy motion and deformation in the optimization process, significantly improves plan robustness and achieves higher quality treatment plans for lung cancer patients. The method is being evaluated for multiple patients with different tumor and motion characteristics.

TU-E-213-02: Credentialing

Purpose: To develop the first 4D robust optimization (RO) method accounting for respiratory motion and evaluate its potential to improve plan robustness and optimality compared to 3D RO and PTV-based optimization. Methods: A set of 4D CT images are used to track respiratory motion and deformation of tumors and organs. For each of 10 respiration phases, dose distributions for nine different uncertainty scenarios including the nominal one, those incorporating ±5mm setup uncertainties long x, y and z directions and ±3.5% range uncertainties are calculated. All 90 dose distributions are simultaneously optimized to achieve full dose coverage of 10 CTVs and sparing of normal structures. ITV-based 3D RO and PTV-based optimization based on the average CT are also carried for the same patient using same dose volume constrains. After optimization, 4D robustness evaluation was performed for all resulting plans. The CTV coverage and the sparing of normal tissue in 10 phases are evaluated and compared among the three methods. The widths of DVH bands represent the robustness of dose distributions in the structures. Results: For one patient studied so far, the worst case CTV coverage by the prescription dose among all 90 scenarios is: 99% for 4D RO; 88.9% for 3D RO, and 85.2% for PTV based optimization. 4D RO also results in best robustness with the narrowest DVH’ bandwidths for the CTV. 4D and 3D RO have similar organ sparing while PTV based optimization results in worst organ sparing. Conclusion: 4D robust optimization which accounts for anatomy motion and deformation in the optimization process, significantly improves plan robustness and achieves higher quality treatment plans for lung cancer patients. The method is being evaluated for multiple patients with different tumor and motion characteristics.

SU‐E‐T‐651: Quantification of Dosimetric Accuracy of Respiratory Gated Stereotactic Body Radiation Therapy

Purpose:To quantify the dosimetric accuracy of respiratory gated stereotactic body radiation therapy delivery using dynamic thorax phantom.Methods:Three patients with mobile target (2 lung, 1liver) were chosen. Retrospective 4DCT image sets were acquired for using Varian RPM system. An in‐house MATLAB program was designed for MIP, MinIP and AvgIP generation. ITV was contoured on MIP image set for lung patients and on MinIP for liver patient. Dynamic IMRT plans were generated on selected phase bin image set in Eclipse (v10.0) planning system. CIRS dynamic thorax phantom was used to perform the dosimetric quality assurance. Patient breathing pattern file from RPM system was converted to phantom compatible file by an in‐house MATLAB program. This respiratory pattern fed to the CIRS dynamic thorax phantom. 4DCT image set was acquired for this phantom using patient breathing pattern. Verification plans were generated using patient gating window and delivered on the phantom. Measurements were carried out using with ion chamber and EBT2 film. Exposed films were analyzed and evaluated in FilmQA software.Results:The stability of gated output in comparison with un‐gated output was within 0.5%. The Ion chamber measured and TPS calculated dose compared for all the patients. The difference observed was 0.45%, −0.52% and −0.54 for Patient 1, Patient2 and Patient 3 respectively.Gamma value evaluated from EBT film shows pass rates from 92.41% to 99.93% for 3% dose difference and 3mm distance to agreement criteria.Conclusion:Dosimetric accuracy of respiratory gated SBRT delivery for lung and liver was dosimetrically acceptable. The Ion chamber measured dose was within 0.203±0.5659% of the expected dose. Gamma pass rates were within 96.63±3.84% of the expected dose.