Predicting Fluid Dynamics in Physical-informed Mesh-reduced Space
Explore the source record for details and available documents.
Engineering topics
Publications and source records attributed to Hu, Y.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Purpose: To develop a biological modeling strategy which incorporates the response observed on the mid-treatment PET/CT into a dose escalation design for adaptive radiotherapy of non-small-cell lung cancer. Method: FDG-PET/CT was acquired midway through standard fractionated treatment and registered to pre-treatment planning PET/CT to evaluate radiation response of lung cancer. Each mid-treatment PET voxel was assigned the median SUV inside a concentric 1cm-diameter sphere to account for registration and imaging uncertainties. For each voxel, the planned radiation dose, pre- and mid-treatment SUVs were used to parameterize the linear-quadratic model, which was then utilized to predict the SUV distribution after the full prescribed dose. Voxels with predicted post-treatment SUV≥2 were identified as the resistant target (response arm). An adaptive simultaneous integrated boost was designed to escalate dose to the resistant target as high as possible, while keeping prescription dose to the original target and lung toxicity intact. In contrast, an adaptive target volume was delineated based only on the intensity of mid-treatment PET/CT (intensity arm), and a similar adaptive boost plan was optimized. The dose escalation capability of the two approaches was compared. Result: Images of three patients were used in this planning study. For one patient, SUV prediction indicated complete response and no necessary dose escalation. For the other two, resistant targets defined in the response arm were multifocal, and on average accounted for 25% of the pre-treatment target, compared to 67% in the intensity arm. The smaller response arm targets led to a 6Gy higher mean target dose in the adaptive escalation design. Conclusion: This pilot study suggests that adaptive dose escalation to a biologically resistant target predicted from a pre- and mid-treatment PET/CT may be more effective than escalation based on the mid-treatment PET/CT alone. More plans and ultimately clinical protocols are needed to validate this approach. MSKCC has a research agreement with Varian Medical System.
Purpose: The purpose of this study was to validate the use of a cascaded linear system model for MV cone-beam CT (CBCT) using a multi-layer (MLI) electronic portal imaging device (EPID) and provide experimental insight into image formation. A validated 3D model provides insight into salient factors affecting reconstructed image quality, allowing potential for optimizing detector design for CBCT applications. Methods: A cascaded linear system model was developed to investigate the potential improvement in reconstructed image quality for MV CBCT using an MLI EPID. Inputs to the three-dimensional (3D) model include projection space MTF and NPS. Experimental validation was performed on a prototype MLI detector installed on the portal imaging arm of a Varian TrueBeam radiotherapy system. CBCT scans of up to 898 projections over 360 degrees were acquired at exposures of 16 and 64 MU. Image volumes were reconstructed using a Feldkamp-type (FDK) filtered backprojection (FBP) algorithm. Flat field images and scans of a Catphan model 604 phantom were acquired. The effect of 2×2 and 4×4 detector binning was also examined. Results: Using projection flat fields as an input, examination of the modeled and measured NPS in the axial plane exhibits good agreement. Binning projection images was shown to improve axial slice SDNR by a factor of approximately 1.4. This improvement is largely driven by a decrease in image noise of roughly 20%. However, this effect is accompanied by a subsequent loss in image resolution. Conclusion: The measured axial NPS shows good agreement with the theoretical calculation using a linear system model. Binning of projection images improves SNR of large objects on the Catphan phantom by decreasing noise. Specific imaging tasks will dictate the implementation image binning to two-dimensional projection images. The project was partially supported by a grant from Varian Medical Systems, Inc. and grant No. R01CA188446-01 from the National Cancer Institute.
Purpose: The purpose of this study was to quantify the improvement in tumor tracking, with and without fiducial markers, afforded by employing a multi-layer (MLI) electronic portal imaging device (EPID) over the current state-of-the-art, single-layer, digital megavolt imager (DMI) architecture. Methods: An ideal observer signal-to-noise ratio (d’) approach was used to quantify the ability of an MLI EPID and a current, state-of-the-art DMI EPID to track lung tumors from the treatment beam’s-eye-view. Using each detector modulation transfer function (MTF) and noise power spectrum (NPS) as inputs, a detection task was employed with object functions describing simple three-dimensional Cartesian shapes (spheres and cylinders). Marker-less tumor tracking algorithms often use texture discrimination to differentiate benign and malignant tissue. The performance of such algorithms is simulated by employing a discrimination task for the ideal observer, which measures the ability of a system to differentiate two image quantities. These were defined as the measured textures for benign and malignant lung tissue. Results: The NNPS of the MLI ∼25% of that of the DMI at the expense of decreased MTF at intermediate frequencies (0.25≤<f≤<0.5 cycles/mm). Lung tissue textures were found to follow a power-law trend in the frequency domain. It was found that for both fiducial marker tracking and lung tissue texture differentiation, the signal power of each task is retained at low frequencies. Thus, little difference may be observed in signal power when DMI and MLI are compared. In all cases, improvements in tracking were greater than a factor of 2. Conclusion: MLI performance in tumor tracking is greatly improved by the additional imager layers. This implies that further improvements in tracking may be gained through increasing the thickness of each MLI layer. For tracking, the MLI performance is limited by noise response. Losses in MTF result in negligible differences in d’. The project was partially supported by a grant from Varian Medical Systems, Inc. and grant No. R01CA188446-01 from the National Cancer Institute.
Purpose: Beams-eye-view imaging applications such as real-time soft-tissue motion estimation and MV-CBCT are hindered by the inherently low image contrast of electronic portal imaging devices (EPID) currently in clinical use. We investigate a cost effective scintillating glass that provides substantially increased detective quantum efficiency (DQE) and contrast to noise ratio (CNR). Methods: A pixelated scintillator prototype was built from LKH-5 glass. The array is 12mm thick; 42.4×42.4cm2 wide and features 1.51mm pixel pitch with 20µm separation (glue+septa). The LKH-5 array was mounted on the active matrix flat panel imager (AMPFI) of an AS-1200 (Varian) with the GdO2S2:Tb removed. A second AS-1200 was utilized as reference detector. The prototype EPID was characterized in terms of CNR, modulation transfer function (MTF) and DQE. Additionally, the visibility of various fiducial markers typically used in the clinic as well as a realistic 3D-printed lung tumor model was assessed. All items were placed in a 12cm thick solid water phantom. CNR is estimated using a Las Vegas contrast phantom, presampled MTF is estimated using a slanted slit technique and the DQE is calculated from measured normalized noise power spectra (NPS) and the MTF. Results: DQE(0) for the LKH-5 prototype increased by a factor of 8× to about 10%, compared to the AS-1200 equipped with its standard GdO2S2:Tb scintillator. CNR increased by a factor of 5.3×. Due to the pixel size the MTF50 decreased by about 55% to 0.23lp/mm. The visibility of all fiducial markers as well as the tumor model were however markedly improved in comparison to an acquisition with the same parameters using the GdO2S2:Tb scintillator. Conclusion: LKH-5 scintillating glasses allow the cost effective construction of thick pixelated scintillators for portal imaging which can yield a substantial increase in DQE and CNR. Soft tissue and fiducial marker visibility was found to be markedly improved. The project was supported in part by NIH grant R01CA188446-01 and a grant from Varian Medical Systems.
Purpose: To develop and validate a Monte Carlo (MC) model of a novel multi-layer imager (MLI) for megavolt (MV) energy beams. The MC model will enable performance optimization of the MLI design for clinical applications including patient setup verification, tumor tracking and MVCBCT. Methods: The MLI is composed of four layers of converter, scintillator and light detector, each layer similar to the current clinical AS1200 detector (Varian Medical Systems, Inc). The MLI model was constructed using the Geant4 Application for Tomographic Emission (GATE v7.1) and includes interactions for x-ray photons, charged particles and optical photons. Computational experiments were performed to assess Modulation Transfer Function (MTF), Detective Quantum Efficiency (DQE) and Noise Power Spectrum normalized by photon fluence and average detector signal (qNNPS). Results were compared with experimental measurements. The current work incorporates, in one model, the complete chain of events occurring in the imager; i.e. starting from x-ray interaction to charged particle transport and energy deposition to subsequent generation, interactions and detection of optical photons. Results: There is good agreement between measured and simulated MTF, qNNPS and DQE values. Normalized root mean squared error (NRMSE) between measured and simulated values over all four layers was 2.18%, 2.43% and 6.05% for MTF, qNNPS and DQE respectively. The relative difference between simulated and measured values for qNNPS(0) was 1.68% and 1.57% for DQE(0). Current results were obtained using a 6MV Varian Truebeam™ spectrum. Conclusion: A comprehensive Monte Carlo model of the MLI prototype was developed to allow optimization of detector components. The model was assessed in terms of imaging performance using standard metrics (i.e. MTF, qNNPS, DQE). Good agreement was found between simulated and measured values. The model will be used to assess alternative detector constructions to facilitate advanced clinical imaging applications including MV-CBCT and tumor tracking. The project was supported, partially, by a grant from Varian Medical Systems, Inc., and Award No. R01CA188446-01 from the National Cancer Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Purpose: The aim of the current work is to investigate the feasibility of megavoltage spectral imaging using a multiple layered detector for enhancement of low contrast detectability through material segmentation and discrimination (such as bone, markers and metal implants). Potentially the technique can be applied to improve detection and reduce dose in Megavoltage Cone Beam Computed Tomography (MV-CBCT). Methods: Experiments were performed with a prototype multi-layer imager (MLI) which has higher detective efficiency and lower noise characteristics than conventional Electronic Portal Imaging Devices (EPIDs). Images of a solid water phantom were acquired at 2.5 MV, 6MV and 6MV without flattening filter (FFF). The following materials were placed within a stack of solid water: aluminum, copper and gold. Material separation was assessed based on Contrast-to-Noise Ratio (CNR) of the weighted image, formed by a weighted subtraction of the images from two layers of the MLI. A range of weighting factors were investigated for material separation. Results: CNR can be minimized for each material by appropriate selection of the subtraction weighting factor. This is equivalent to a selective subtraction of specific materials from the image. Using multiple layers simultaneously also decreases the dose requirement and removes any registration errors. The minimum CNR for aluminum, copper and gold at the weighted image formed with 2.5MV was obtained at weighting factors equal to 0.92, 0.76 and 0.64 respectively. The corresponding values at 6MVFFF were 0.99, 0.92 and 0.78 respectively. Conclusion: In the current work, an MV spectral imaging feasibility study was attempted using a novel multi-layer prototype EPID imager. Initial results suggest that material separation based on spectral differences between different layers is possible. This spectral imaging technique has potential advantages in MV-CBCT for real-time target tracking, patient set-up imaging and adaptive radiotherapy. The project was supported, partially, by a grant from Varian Medical Systems, Inc., and Award No. R01CA188446-01 from the National Cancer Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Purpose: The clinical commissioning of IMRT subject to a magnetic field is challenging. The purpose of this work is to develop a GPU-accelerated Monte Carlo dose calculation platform based on PENELOPE and then use the platform to validate a vendor-provided MRIdian head model toward quality assurance of clinical IMRT treatment plans subject to a 0.35 T magnetic field. Methods: We first translated PENELOPE from FORTRAN to C++ and validated that the translation produced equivalent results. Then we adapted the C++ code to CUDA in a workflow optimized for GPU architecture. We expanded upon the original code to include voxelized transport boosted by Woodcock tracking, faster electron/positron propagation in a magnetic field, and several features that make gPENELOPE highly user-friendly. Moreover, we incorporated the vendor-provided MRIdian head model into the code. We performed a set of experimental measurements on MRIdian to examine the accuracy of both the head model and gPENELOPE, and then applied gPENELOPE toward independent validation of patient doses calculated by MRIdian’s KMC. Results: We achieve an average acceleration factor of 152 compared to the original single-thread FORTRAN implementation with the original accuracy preserved. For 16 treatment plans including stomach (4), lung (2), liver (3), adrenal gland (2), pancreas (2), spleen (1), mediastinum (1) and breast (1), the MRIdian dose calculation engine agrees with gPENELOPE with a mean gamma passing rate of 99.1% ± 0.6% (2%/2 mm). Conclusions: We developed a Monte Carlo simulation platform based on a GPU-accelerated version of PENELOPE. We validated that both the vendor provided head model and fast Monte Carlo engine used by the MRIdian system are accurate in modeling radiation transport in a patient using 2%/2 mm gamma criteria. Future applications of this platform will include dose validation and accumulation, IMRT optimization, and dosimetry system modeling for next generation MR-IGRT systems.
Purpose: To investigate how spot size and spacing affect plan quality, especially, plan robustness and the impact of interplay effect, of robustly-optimized intensity-modulated proton therapy (IMPT) plans for lung cancer. Methods: Two robustly-optimized IMPT plans were created for 10 lung cancer patients: (1) one for a proton beam with in-air energy dependent large spot size at isocenter (σ: 5–15 mm) and spacing (1.53σ); (2) the other for a proton beam with small spot size (σ: 2–6 mm) and spacing (5 mm). Both plans were generated on the average CTs with internal-gross-tumor-volume density overridden to irradiate internal target volume (ITV). The root-mean-square-dose volume histograms (RVH) measured the sensitivity of the dose to uncertainties, and the areas under RVH curves were used to evaluate plan robustness. Dose evaluation software was developed to model time-dependent spot delivery to incorporate interplay effect with randomized starting phases of each field per fraction. Patient anatomy voxels were mapped from phase to phase via deformable image registration to score doses. Dose-volume-histogram indices including ITV coverage, homogeneity, and organs-at-risk (OAR) sparing were compared using Student-t test. Results: Compared to large spots, small spots resulted in significantly better OAR sparing with comparable ITV coverage and homogeneity in the nominal plan. Plan robustness was comparable for ITV and most OARs. With interplay effect considered, significantly better OAR sparing with comparable ITV coverage and homogeneity is observed using smaller spots. Conclusion: Robust optimization with smaller spots significantly improves OAR sparing with comparable plan robustness and similar impact of interplay effect compare to larger spots. Small spot size requires the use of larger number of spots, which gives optimizer more freedom to render a plan more robust. The ratio between spot size and spacing was found to be more relevant to determine plan robustness and the impact of interplay effect than spot size alone. This research was supported by the National Cancer Institute Career Developmental Award K25CA168984, by the Fraternal Order of Eagles Cancer Research Fund Career Development Award, by The Lawrence W. and Marilyn W. Matteson Fund for Cancer Research, by Mayo Arizona State University Seed Grant, and by The Kemper Marley Foundation.
Purpose: For proton beam therapy, small fiducial markers are preferred for patient alignment due to less interference with the proton beam. Visualizing small fiducial markers can be challenging in MRI. This study intends to investigate MRI imaging protocols for better visualization of small fiducial markers. Methods: Two carbon and two coil-shaped gold markers were placed into a gel phantom. Both carbon markers had a diameter of 1mm and a length of 3mm. Both gold markers had a length of 5mm. One gold marker had a diameter of 0.5mm and the other had a diameter of 0.75mm. T1 VIBE, T2 SPACE, TrueFISP and susceptibility weighted (SW) images were acquired. To improve marker contrast, high spatial resolution was used to reduce partial volume effect. Slice thickness was 1.5mm for all four sequences and in-plane resolution was 0.6mm for TrueFISP, 0.7mm for T1 VIBE, and 0.8mm for T2 SPACE and SW. For comparison purpose, a 3D T1 VIBE image set at 3mm slice thickness and 1.2mm in-plane resolution was also acquired. Results: All markers were visible in all high-resolution image sets. In each image set, marker-induced signal void was the smallest (in diameter) for carbon markers, followed by the 0.5mm gold marker and the largest for the 0.75mm gold marker. The SW images had the largest marker-induced signal void. However, those might be confused by susceptibility-gradient-induced signal voids. T1 VIBE had good visualization of markers with nicely defined edges. T2 SPACE had reasonable visualization of markers but edges were slightly blurred. TrueFISP had good visualization of markers only if they were not masked by banding artifacts. As a comparison, all markers were hardly visible in the standard resolution T1 VIBE images. Conclusion: 3D high-resolution T1 VIBE and SW have great potential in providing good visualization of small fiducial markers for proton beam therapy.
Purpose: To build a fast, accurate and easily-deployable research platform for Monte-Carlo dose calculations. We port the dose calculation engine PENELOPE to C++, and accelerate calculations using GPU acceleration. Simulations of a Co-60 beam model provided by ViewRay demonstrate the capabilities of the platform. Methods: We built software that incorporates a beam model interface, CT-phantom model, GPU-accelerated PENELOPE engine, and GUI front-end. We rewrote the PENELOPE kernel in C++ (from Fortran) and accelerated the code on a GPU. We seamlessly integrated a Co-60 beam model (obtained from ViewRay) into our platform. Simulations of various field sizes and SSDs using a homogeneous water phantom generated PDDs, dose profiles, and output factors that were compared to experiment data. Results: With GPU acceleration using a dated graphics card (Nvidia Tesla C2050), a highly accurate simulation – including 100*100*100 grid, 3×3×3 mm3 voxels, <1% uncertainty, and 4.2×4.2 cm2 field size – runs 24 times faster (20 minutes versus 8 hours) than when parallelizing on 8 threads across a new CPU (Intel i7-4770). Simulated PDDs, profiles and output ratios for the commercial system agree well with experiment data measured using radiographic film or ionization chamber. Based on our analysis, this beam model is precise enough for general applications. Conclusions: Using a beam model for a Co-60 system provided by ViewRay, we evaluate a dose calculation platform that we developed. Comparison to measurements demonstrates the promise of our software for use as a research platform for dose calculations, with applications including quality assurance and treatment plan verification.
Purpose: Lack of plan robustness may contribute to local failure in volumetric-modulated arc therapy (VMAT) to treat head and neck (H&N) cancer. Thus we compared plan robustness of VMAT with intensity-modulated radiation therapy (IMRT). Methods: VMAT and IMRT plans were created for 9 H&N cancer patients. For each plan, six new perturbed dose distributions were computed — one each for ± 3mm setup deviations along the S-I, A-P and L-R directions. We used three robustness quantification tools: (1) worst-case analysis (WCA); (2) dose-volume histograms (DVHs) band (DVHB); and (3) root-mean-square-dose deviation (RMSD) volume histogram (DDVH). DDVH represents the relative volume (y) on the vertical axis and the RMSD (x) on the horizontal axis. Similar to DVH, this means that y% of the volume of the indicated structure has the RMSD at least x Gy[RBE].The width from the first two methods at different target DVH indices (such as D95 and D5) and the area under the DDVH curves (AUC) for the target were used to indicate plan robustness. In these robustness quantification tools, the smaller the value, the more robust the plan is. Plan robustness evaluation metrics were compared using Wilcoxon test. Results: DVHB showed the width at D95 from IMRT to be larger than from VMAT (unit Gy) [1.59 vs 1.18 (p=0.49)], while the width at D5 from IMRT was found to be slightly larger than from VMAT [0.59 vs 0.54 (p=0.84)]. WCA showed similar results [D95: 3.28 vs 3.00 (p=0.56); D5: 1.68 vs 1.95 (p=0.23)]. DDVH showed the AUC from IMRT to be slightly smaller than from VMAT [1.13 vs 1.15 (p=0.43)]. Conclusion: VMAT plan robustness is comparable to IMRT plan robustness. The plan robustness conclusions from WCA and DVHB are DVH parameter dependent. On the other hand DDVH captures the overall effect of uncertainties on the dose to a volume of interest. NIH/NCI K25CA168984; Eagles Cancer Research Career Development; The Lawrence W. and Marilyn W. Matteson Fund for Cancer Research Mayo ASU Seed Grant; The Kemper Marley Foundation.
Purpose: This study was to investigate size-specific dose estimates (SSDE) for routine adult abdominal CT examinations in Taiwan. Methods: A national survey was conducted in Taiwan in 2014 to investigate SSDEs for routine adult abdominal CT examinations. The hospitals involved in this study provided CT images of their typical patients. The CT image in the level of the middle liver was selected to record the corresponding tube current, slice mAs or effective mAs. The image was also used to estimate the dimensions of patient as measuring the lengths in the anterior to posterior (AP) and lateral (LAT) directions. The effective diameter was then calculated from AP and LAT, and used to look up conversion factors in the AAPM 204 report. The volume CTDI (CTDIvol) for each CT unit was measured on sites using a 32-cm cylindrical standard dose phantom and a calibrated pencil-type ionization chamber. Individual patient’s SSDEs were then calculated from the corresponding SSDE conversion factor and the CTDIvol. Results: The study cohort included 111 CT units. The ratio of turning on automatic tube current modulation (ATCM) or not is 88:23. Effective diameters are 258.7±25.1 mm (167–366 mm). 99.3% of typical patients selected by each hospital have smaller effective diameter than the 32-cm dosimetry phantom. Adult abdomenal SSDE is 17.5 ± 8.8 mGy (1.9-58 mGy). The SSDE seems to decrease as the effective diameter increases as the ATCM turns off, and independent with the effective diameter as the ATCM turns on. Conclusion: The SSDE for typical patients in Taiwan was investigated. We continue to complete this investigation in 2015 to include more valid data to establish SSDE reference level in Taiwan. This study was financially supported by the Atomic Energy Council in Taiwan.
Purpose: MR based 3D gel has been proposed for radiation therapy dosimetry. However, access to MR scanner has been one of the limiting factors for its wide acceptance. Recent commercialization of an on-board MR-IGRT device (ViewRay) may render the availability issue less of a concern. This work reports our attempts to simulate MR based dose measurement accuracy on ViewRay using three different gels. Methods: A spherical BANG gel dosimeter was purchased from MGS Research. Cylindrical MAGIC gel and Fricke gel were fabricated in-house according to published recipes. After irradiation, BANG and MAGIC were imaged using a dual-echo spin echo sequence for T2 measurement on a Philips 1.5T MR scanner, while Fricke gel was imaged using multiple spin echo sequences. Difference between MR measured and TPS calculated dose was defined as noise. The noise power spectrum was calculated and then simulated for the 0.35 T magnetic field associated with ViewRay. The estimated noise was then added to TG-119 test cases to simulate measured dose distributions. Simulated measurements were evaluated against TPS calculated doses using gamma analysis. Results: Given same gel, sequence and coil setup, with a FOV of 180×90×90 mm3, resolution of 3×3×3 mm3, and scanning time of 30 minutes, the simulated measured dose distribution using BANG would have a gamma passing rate greater than 90% (3%/3mm and absolute). With a FOV 180×90×90 mm3, resolution of 4×4×5 mm3, and scanning time of 45 minutes, the simulated measuremened dose distribution would have a gamma passing rate greater than 97%. MAGIC exhibited similar performance while Fricke gel was inferior due to much higher noise. Conclusions: The simulation results demonstrated that it may be feasible to use MAGIC and BANG gels for 3D dose verification using ViewRay low-field on-board MRI scanner.
Purpose: To evaluate the feasibility of using the weighted hybrid iterative spiral k-space encoded estimation (WHISKEE) technique to improve spatial resolution of tracking images for onboard MR image guided radiation therapy (MR-IGRT). Methods: MR tracking images of abdomen and pelvis had been acquired from healthy volunteers using the ViewRay onboard MRIGRT system (ViewRay Inc. Oakwood Village, OH) at a spatial resolution of 2.0mm*2.0mm*5.0mm. The tracking MR images were acquired using the TrueFISP sequence. The temporal resolution had to be traded off to 2 frames per second (FPS) to achieve the 2.0mm in-plane spatial resolution. All MR images were imported into the MATLAB software. K-space data were synthesized through the Fourier Transform of the MR images. A mask was created to selected k-space points that corresponded to the under-sampled spiral k-space trajectory with an acceleration (or undersampling) factor of 3. The mask was applied to the fully sampled k-space data to synthesize the undersampled k-space data. The WHISKEE method was applied to the synthesized undersampled k-space data to reconstructed tracking MR images at 6 FPS. As a comparison, the undersampled k-space data were also reconstructed using the zero-padding technique. The reconstructed images were compared to the original image. The relatively reconstruction error was evaluated using the percentage of the norm of the differential image over the norm of the original image. Results: Compared to the zero-padding technique, the WHISKEE method was able to reconstruct MR images with better image quality. It significantly reduced the relative reconstruction error from 39.5% to 3.1% for the pelvis image and from 41.5% to 4.6% for the abdomen image at an acceleration factor of 3. Conclusion: We demonstrated that it was possible to use the WHISKEE method to expedite MR image acquisition for onboard MR-IGRT systems to achieve good spatial and temporal resolutions simultaneously. Y. Hu and O. green receive travel reimbursement from ViewRay. S. Mutic has consulting and research agreements with ViewRay. Q. Zeng, R. Nana, J.L. Patrick, S. Shvartsman and J.F. Dempsey are ViewRay employees.