Volume 36, Issue 6, June 2009
Index of content:
- Joint Imaging/Therapy Moderated Poster Session: Exhibit Hall ‐ Area 3
- Moderated Poster ‐ Future Imaging Strategies in Radiotherapy
SU‐DD‐A3‐01: Swallowing‐Induced Target Motion and Dosimetric Impact for Head and Neck Cancer Radiotherapy as Assessed by Dynamic MRI36(2009); http://dx.doi.org/10.1118/1.3181081View Description Hide Description
Purpose: Deglutition (i.e., swallowing) results in dynamic changes in the geometry of aerodigestive tract anatomy, which may affect the doses delivered to targets and critical structures in head and neck cancerradiotherapy. The purpose of this study was to determine (1) whether MRI could be used to capture the motion of swallowing, (2) the geometrical variation induced by deglutition, and (3) the dosimetric impact of deglutition on head and neck cancerradiotherapy.Method and Materials: Five patients with aerodigestive tract cancer were imaged in treatment position on a 3.0T Siemens Verio scanner. Two time series of single‐slice sagittal images were acquired using a 2D FLASH sequence (6.667 frames/sec with and 3.257 frames/sec without parallel imaging). Swallowing frequency and duration were determined from time series data and motion was assessed using contours of gross tumor volumes (GTV) drawn during one swallowing event. Using each patient's IMRT plan and swallowing frequency and duration results, dosimetric impact was assessed by worst‐case scenario of the GTV moving completely out of the treatment field during deglutition. Results: Deglutition was non‐periodic, with frequency and duration ranging from 2.89–24.18 mHz and 3.86–6.10 sec. Dynamic changes in GTV position, size, and shape resulted from deglutition, ranging from 0–28.36mm and 0–60% change in area of GTV projection. Deglutition resulted in a 1.4–14.3% reduction in dose delivered to GTV. Conclusion:MRI is able to capture the motion of swallowing with or without the use of parallel imaging. Deglutition is non‐periodic, and results in dynamic changes in GTV position, size, and shape. The overall impact of deglutition on dose delivery for head and neck cancer, although generally insignificant, can be non‐negligible for patients with extremely high swallowing frequency and long duration.
Conflict of Interest: Partially supported by Siemens Healthcare and the MCW Cancer Center Meinerz Fellowship.
SU‐DD‐A3‐02: Quantification of Heterogeneous Tumor Blood Volume as Early Predictors for Treatment Outcome in Advanced Head‐And‐Neck Cancer36(2009); http://dx.doi.org/10.1118/1.3181082View Description Hide Description
Purpose: To evaluate the predicting values of metrics that quantify poorly perfused tumor volumes in patients with advanced head‐and‐neck cancer (HNC) treated with chemo‐radiation therapy (chemo‐RT) for outcomes. Materials and Method: Dynamic contrast‐enhanced (DCE) MRI scans were acquired for thirteen patients treated with concomitant chemo‐RT for advanced HNC, before therapy and 2 weeks after start of therapy. Blood volume (BV) maps were computed from the DCE MRI. Fuzzy C‐means clusteringanalysis (FCM), a segmentation method for partitioning a dataset into multiple clusters by minimizing the intracluster distance among members, was utilized to analyze heterogeneous BV in primary tumor volumes. The pre‐treatment tumor BV (TBV) values from all patients were used as training data and partitioned into 2, 3, or 4 clusters using FCM; and the resulting prototype vectors were adopted to partition each individual TBV before and during therapy. The volumes of the clusters with the lowest BV values were tested for their associations with treatment outcome. In addition, the cut‐off BV values for the lowest 1 cc, 2 cc, or 5 cc primary tumor sub‐volumes were calculated at both time points; and their predicting values for treatment outcome were evaluated. Results: Four of the 13 patients had local failure. Eight metrics differentiate local failure vs. local control significantly: pre‐treatment cut‐off TBV for the lowest 5 cc (p=0.03); average cut‐off TBVs of the lowest 2 cc (p=0.01) or 5 cc (p=0.04) over pre‐ and during‐ treatment; the volumes of the lowest TBV cluster at pre‐treatment (p=0.02), during‐treatment (p⩽0.03), as well as the average value over pre‐ and during‐ treatment (p⩽0.03). Conclusion: Our results suggest that poorly perfused sub‐volumes of the tumor before and/or early after start of treatment differentiate local failure vs. local control. The metrics discovered here will be candidates for future larger studies.
Supported by NIHP01CA59827.
36(2009); http://dx.doi.org/10.1118/1.3181083View Description Hide Description
Purpose: Prompt gamma rays emitted from proton‐nucleus interactions in tissue present a promising non‐invasive, in situ means of monitoring proton beam based radiotherapy. This study investigates the fluence and energy distribution of prompt gamma rays emitted during proton irradiation of phantoms. This information was used to develop a correlation between the measured and calculated gamma emission and the proton beam range, which would allow treatments to more effectively exploit the sharp distal falloff in the dose distributions of protons.Method & Materials: Measurements were performed at the Francis H. Burr Proton Therapy Center using a 5nA pencil proton beam, a homogeneous cylindrical Lucite phantom, and a lead collimated sodium iodide (NaI) detector arranged orthogonal to the beam. Emulating this experiment, a model of the phantom with a monoenergetic proton beam and an annular array of ideal photon tallies arranged orthogonal to the beam was developed using the Monte Carlo code MCNPX 2.6.0. Results: Experimental and computational results indicated a correlation between gamma emission and the proton depth‐dose profile. Several peaks were evident in the calculated energy spectrum and the 4.44MeV emission from was the most intense line having any apparent correlation with the depth dose profile. Arbitrary energy binning of 4–5MeV and 4–8MeV was performed on the Monte Carlo data; this binned data yielded a distinct emission peak 1cm proximal to the Bragg peak. In all cases in the Lucite phantom the position of the Bragg peak's 80% distal falloff corresponded with the position of the 4–8MeV binned 50% distal falloff. Conclusions: The results are promising and indicate the feasibility of prompt gamma emissiondetection as a means of characterizing the proton beam range in situ. This study has established the measurement and computational tools necessary to pursue the development of this technique.
36(2009); http://dx.doi.org/10.1118/1.3181084View Description Hide Description
Purpose: A microCT‐based small‐animal‐radiotherapy system with a variable‐aperture collimator has been constructed, and a Monte Carlomodel of this system has been developed. A series of Monte Carlo(MC) simulations were performed in order to calibrate the model against measured beam commissioning data. Method and Materials: The EGSnrc codes were used for MC simulation. The simulations included different source models, different apertures, and included or excluded the effect of the plastic cylindrical shell enclosing the CT bore. We considered sources including a circular x‐ray beam with spectra generated by Boone‐Fewell formula, and a simulated x‐ray tube with parameters fit to measured beam profile. The beam width was set to 2, 5, 10, and 20 mm. The depth dose in water and the beam profiles simulated with and without the CT bore were compared, and the calibration factor for the conversion from MCdose to the measured dose rate in water was computed. Results: The simulation showed that a model with 2.5 mm Al filtration matches the measured depth‐dose curves, with the mean error on all measurement points for 2.5 mm Al was 2.5%. The CT bore introduces a 6% attenuation but does not significantly affect the beam energy spectrum or depth‐dose characteristics. The low energy part of the spectrum from the x‐ray‐tube simulation varied substantially from that predicted by the Boone‐Fewell model. The conversion factor was found to be 7.7415×1018 particles/min. Conclusion:Models and parameters in the Monte Carlo planning system have been determined from the depth‐dose measurements and simulations. The observed effect of the CT bore suggested that the effect can be attributed to additional attenuation, which significantly simplifies the software complexity of the Monte Carlo planning system.
SU‐DD‐A3‐05: Radiosensitization of Endothelial Cells Model Using Gold Nanoparticles for Microbeam Radiotherapy36(2009); http://dx.doi.org/10.1118/1.3181085View Description Hide Description
Purpose: Microbeam radiotherapy(MRT) is a technique that use array parallel thin (<100μm) slices of synchrotron generated x‐ray beam. In this study, we investigated the radio‐sensitizing effects of goldnanoparticles (AuNps) on endothelial cells culture model in combination with irradiation of MRT.Method and Materials: Bovine aortic endothelial cells (BAECs) were cultured as a confluent monolayer on a 2 well chamber slides with 0mM, 0.5 mM and 1.0mM of AuNps. The cells were irradiated with 10 Gy of synchrotron generated x‐ray beam of median energy 150 kVp. Each microbeam is approximately 30 microns wide with a spacing of 200 microns between adjacent microbeams. Gafchromic films were attached to the cells culture slide to verify the dose received by the cells. The experiments were performed on the BL28B2 beamline at the SPring8 Synchrotron Japan. The cells were then fixed at 6, 12 and 24 hours after irradiation and were imaged using phase contrast microscope. Cells viability assays using tryphan blue exclusion method were performed after 24 hours of irradiation. Results: The observations under phase contrast microscopy show all the cells were dead at the area of irradiation. Samples with AuNps clearly showed the path of microbeam which is visible as a straight line compared to the samples without AuNps. There are some neighbouring cells start to migrate to the irradiated area, filling the gap for both samples. The cells viability results showed the dose enhancement effects where only 52% cells survived for 0.5 mMol AuNps and only 18% survive for 1mMol AuNps when irradiated with microbeam. These results were expressed as percentage relative to the control samples. Conclusion: The results demonstrate that the AuNps are effective radiosensitizer that will increase the therapeutic efficacy of MRT.
36(2009); http://dx.doi.org/10.1118/1.3181086View Description Hide Description
Introduction: An electronic portal imaging device(EPID) in cine mode can be used for validating respiratory gating and stereotactic body radiation therapy(SBRT) by tracking implanted fiducials. Manual tracking methods are time and labor intensive, limiting the utility of the validation. We have developed a fully automated algorithm to quickly and accurately extract the markers in EPIDimages and reconstruct their 3D positions. Materials and Methods: The markers were detected and recognized using an image processing algorithm based on the Laplacian of Gaussian (LoG) filter. To reduce false marker detection, a marker registration technique was applied using image intensity as well as the geometric spatial transformations between the reference marker positions produced from the projection of 3D CTimages and the estimated marker positions. An average marker position in 3D was reconstructed by backprojecting, towards the source, the position of each marker on the 2D image.Results: From phantom studies, spatial accuracies of <1 mm were achieved in both 2D and 3D marker locations. Using only the LoG algorithm, the marker detection success rate was 88.8%. However, adding the registration technique which utilizes prior CT information, the success rate was increased to 100%. In addition, we have examined the cases of 5 patients being treated under an SBRT protocol for hepatic metastases. The intrafractional tumor motion (3.1–11.3 mm) in the SI direction was measured using the 2D images. The interfractional patient setup errors (0.1–12.7 mm) in the SI, AP, and LR directions were obtained from the marker locations reconstructed in 3D and compared to the reference planning CTimage.Conclusions: The measured intrafractional tumor motion and the interfractional daily patient setup error can be used for off‐line retrospective verification of SBRT.
This work was partially supported by a grant from Varian Medical Systems, Inc.
- Moderated Poster ‐ Target Localization
SU‐EE‐A3‐01: Comparison of Six Manual and Automatic 4D CT Registration Methods for Lung SBRT Image Guidance36(2009); http://dx.doi.org/10.1118/1.3181105View Description Hide Description
Purpose: 4D CT is now widely used in radiation therapy for generating accurate ITV's or for facilitation of gated treatmentdelivery. It is, as yet, relatively unexplored for use as a potential tool for image guidance of SBRT of lung. Use of 4D imaging for image guidance in SBRT requires that we determine an acceptable method for fusing 2 4D data sets.Method and Materials: We compare five different manual and/or automatic 4D registration methods and present the comparison results for phantom and patient data: Individual phase‐to‐phase registration; Average of phase to phase; MIP‐to‐MIP; Average Intensity Projection; Auto‐segmentation with centroid; Manual GE SimMD methods. Phantom studies were performed using the CIRS Dynamic Thorax phantom (CIRS, Norfolk, VA). A 2‐cm target was set into motion with a periodic sinusoidal 3D motion inside the anthropomorphic phantom. The baseline motion envelope of the target for simulation day movement was 4.2 mm L/R, 6.0 mm A/P, and 14.0 mm I/S. The target was then re‐programmed to move with the same motion envelope, but shifted away from the simulation motion envelope centroid, for three scenarios: a) Axial (L/R and A/P) only shift; b) Longitudinal only shift; and c) combination of axial (L/R & A/P) and longitudinal shift. Results: Individual phase‐to‐phase fusions were within 2.4 mm for all phases, with some phases performing better than others. The average and AIP methods were seen to perform very similarly, and this seems reasonable when we consider that both are, essentially, averaging methods. The MIP‐to‐MIP and Manual fusion methods were least accurate, but still likely to be considered clinically acceptable. Conclusion: Multiple methods have been explored for registration of 2 4D data sets. The averaging methods were seen to perform best, but all performed within what would reasonably be considered clinically acceptable limits. The centroid method performed particularly well.
SU‐EE‐A3‐02: Imaging Guided Frameless Stereotactic RadioSurgery Using CBCT 6D Image Registration and 6D Couch On Novalis Tx™ System36(2009); http://dx.doi.org/10.1118/1.3181106View Description Hide Description
Purpose: To investigate localization accuracy for image‐guided frameless stereotactic‐radiosurgery (IG‐SRS) using CBCT 6D image registration and 6D‐couch on Novalis Tx system. Method and Materials: Novalis Tx system is a powerful radiosurgery system featuring BrainLAB 6D‐couch and Varian CBCT. In this work, we performed phantom and patient studies to investigate IG‐SRS with CBCT 6D guidance. In phantom studies, a head phantom was positioned on couch, scanned with CBCT, and matched with planning CT using 6D image registration. Matching results were used to test accuracy of CBCT 6D/6D match for correcting translational and rotational setup errors. In patient studies, 10 intracranial SRS cases were randomly selected. Patients were initially positioned with BrainLAB mask system with target‐positioning‐overlays (TAPO). Patients were then scanned with CBCT, and positions were adjusted based on CBCT. Discrepancies between CBCT 3D/3D match and 6D/6D match were quantitatively analyzed. Results: Phantom experiments showed that translational discrepancies in root‐mean‐square (RMS) were 0.39, 0.23, and 0.36mm along lateral, longitudinal, and vertical directions. Rotational discrepancies in RMS were 0.38°, 0.19°, and 0.22° in pitch, roll, yaw, respectively. In retrospective patient studies, setup displacements measured in RMS between BrainLAB TAPO and CBCT 6D/6D match were upto 2.5mm translationally and 1.0° rotationally. If positioning was corrected using CBCT 3D/3D match, translational discrepancies in RMS were <0.5mm except 0.71mm in vertical, and rotational displacements in RMS were 0.67°, 0.88° and 0.22° in pitch, roll, yaw, as compared with CBCT 6D/6D match. CBCT guided 6D correction was performed in two cases to achieve necessary positioning accuracy. After 6D adjustments, translational and rotational displacements of upto 4mm and 0.9° were measured in CBCT 6D/6D match were fully corrected using 6D‐couch. Conclusion:CBCT 6D/6D match offers unprecedented view to verify patient positioning in both translations and rotations. Combined with 6D‐couch, translational and rotational setup errors can be effectively minimized.
36(2009); http://dx.doi.org/10.1118/1.3181107View Description Hide Description
Purpose: Few research groups are involved in integrating magnetic resonance imaging(MRI) with a linear accelerator in order to obtain real‐time MRIimages during the treatment beam‐on. Therefore, the radio frequency (RF) coils used in MRI will be irradiated during this real‐time image guided radiotherapy. The radiation effects on RF coils include instantaneous induced currents and long term radiation damage to components. These effects are potential obstacles facing linac‐MRI integration. This work measures and characterizes the instantaneous effect of pulsed radiation on MRI coils. Method and Materials: A CAT Solenoid coil was placed inside an RF cage, to remove RF noise, and connected to a current amplifier. The amplifier output was sent through RF filters to the RF cage exterior and then to an oscilloscope. The cage was placed in the pulsed beam of a linac and the current induced by pulsed radiation was measured by the oscilloscope; the waveforms were then transferred to a PC and power spectral density was calculated. Results: The RF cage was very effective in eliminating the extraneous RF noise from the radiation induced signal. The induced signal is only present when the radiation beam is on and incident upon the coil. The power spectrum of the induced current indicates that most of the power is contained below 1 MHz, but there is signal is present at the coil resonance frequency (8.5 MHz). The induced current a) occurs mainly in the copper winding, b) is reduced if buildup material is placed on the coil and c) increases linearly with the dose rate. Conclusion: Radiation induced current is present in MRI coils, but its impact on imaging is yet to be determined. This, along with magnetic field impact on induced current will be examined using the functioning 0.2 Tesla linac‐MR prototype.
SU‐EE‐A3‐04: In Vivo Detection of the Anterior Rectal Wall Using a Scintillation Fluid‐Filled Rectal Balloon for Prostate Cancer Radiotherapy36(2009); http://dx.doi.org/10.1118/1.3181108View Description Hide Description
Purpose: The prostate and the anterior rectum interface is perhaps the most important landmark but the most difficult soft tissue region for localization in image‐guided procedures. The goal of this study is to utilize the detection of rectal dose as a surrogate for anterior rectal wall (ARW) localization for daily prostate patient set‐up without using an imaging method. Methods: In this study, we designed a sweeping MLC beam to scan the anterior‐posterior (AP) prostate‐rectum interface from the lateral directions. The dose response from the inside of rectum will be measured using liquid scintillator and associated fiber optics. To validate our design, a 3mm wide lateral MLC segment was designed to move across the prostate‐rectum interface, and the mean rectal dose was calculated for each position in 8 prostate cancer patients with endorectal balloon immobilization and daily CTimage sets. Then the dose‐position response curve was measured daily and fitted to the reference relationship to determine the shift which best reproduces the reference dose‐position relationship relative to the isocenter. To validate this method, the predicted alignments using the dose‐position relationship were compared with the manual AP alignment of the ARW using our in‐house CT registration software. Feasibility of the sliding segment technique was tested in an IMRT phantom fitted with a cylindrical liquid scintillator filled tube, optic fibers, and photodiode for current generation and detection. Results: The mean agreement of the dose‐calculated alignments and the manual image registration was 0.76 millimeters. The primary source of uncertainty was the variation in the insertion of the endorectal balloon. Measured signal increased monotonically with beam position within the rectum and rectal dose.Conclusion: In vivo detection of the ARW is feasible without an on‐board x‐ray imaging technique to achieve better alignment to anatomically important soft tissue structures.
36(2009); http://dx.doi.org/10.1118/1.3181109View Description Hide Description
Purpose: To keep safety margins in lung stereotactic body radiation therapy(SBRT) small and provide retrospective calculation of the delivereddose, the tumor motion should be monitored. We propose a tumor tracking algorithm that can estimate the tumor location from portal images taken during the treatment without the help of fiducial markers. Method and Materials: An algorithm based on a normalized mutual information technique was developed for tumor tracking. First a tumor template and a search region are identified on a DRR set reconstructed from a 4DCT acquired prior to the treatment. The set consists of 10 images relating to 10 equally sized breathing phase bins. The template is then used to track the tumor over the sequence of portal images. To estimate the tracking precision a dynamic thorax phantom was employed. Results: The phantom study showed a sub millimeter tracking accuracy in the superior‐ inferior direction for anterior‐posterior and lateral fields. In a preliminary retrospective patient study the algorithm was able to track the tumor motion throughout the whole imagesequence. Manual verification yielded a tracking magnitude error of xy = (3.4 ± 0.8) mm. Furthermore the algorithm's robustness was tested with portal imagesequences from two other patients with different tumor motion amplitude and contrast. The accuaracy was estimated by comparison with manual tracking and yielded xy = (1.7 ± 1.9) mm and xy = (1.2 ± 0.8) mm, respectively. Conclusion: The algorithm has shown great potential for markerless lungtumor tracking. First test results showed that it can perform tumor tracking on portal images and DRRs even if the tracking template was defined in the other modality respectively.
Conflict of Interest: Varian Medical Systems, Inc.
36(2009); http://dx.doi.org/10.1118/1.3181110View Description Hide Description
Purpose: To determine geometric uncertainties in GTV and ITV delineations using RPM sorted 4DCT images for radiotherapytreatment planning.Methods and Materials: Three targets have been embedded within a cork block (to simulate lungtissue) and attached to a programmable motion controller to simulate lung/tumor motion. We input several breathing patterns including ideal sine waves with varying amplitudes, and realistic patient breathing patterns. The cine slices were reconstructed using RPM signal. The three targets (of known dimensions) were contoured and their dimensions were measured in the direction of motion in the 4DCT images. We measured GTV dimensions reconstructed in the phase‐binned images (10 phases) and ITV dimensions reconstructed in the CT MIP and CT average images.Results: ITV measured in CT MIP images is superior to all other techniques used in our study, with errors < 2 mm in all cases except one, and on average 1.1 mm. The reconstruction is accurate even in cases of irregular breathing and varying amplitudes, with dimensions along the direction of motion within 9% of the expected values. For each set of phase‐binned CTimages, the GTV is not consistently greater or less than the actual target dimension. The errors for individual targets were maximized at large input amplitudes with the largest uncertainties at 15 mm. At phase 0 and 50 %, the measured errors were minimum. Conclusions: We have performed a comprehensive evaluation of the limitations of 4DCT image reconstruction and its use in treatment planning for target delineation. Specifically, we observed that a general trend of increased GTV delineation uncertainty is seen in phase‐binned CTimages with increasing breathing amplitude. However, CT MIP images allowed consistent ITV delineation with minimal uncertainties for all input amplitudes. This study may be useful in guiding clinical judgments in 4DCT and gating procedures.
- Moderated Poster ‐ Image Guided Radiation Therapy
MO‐EE‐A3‐01: Comparison of Prostate Rotation and Calypso Beam Rotation for Prostate Margin Evaluation36(2009); http://dx.doi.org/10.1118/1.3182259View Description Hide Description
Purpose: The Beacon rotation angles reported by the Calypso system have not been utilized clinically due to the difficulties in rotating the patient. The actual prostate rotation angle is affected by Beacons' migration, prostate's shrinkage and deformation due to rectal/bladder filling. This work investigates the actual prostate rotation angles between the planning CT and the treatment CBCT and to compare those reported by the Calypso system. Method and Materials: The Calypso system reports centroid shifts and rotation angles of the 3 implanted Beacons, relative to their locations on the planning CT.CBCT scans were obtained for 9 treatment fractions of 5 patients. The same sets of the Beacons in the planning CT and post‐treatment CBCT were segmented according to their high intensities and aligned according to their centroid positions. An iterative closest point (ICP) method was developed to find the best matching iteratively between two sets of Beacons after rotation and translation. Results: The maximum inter‐Beacon distance varied from 1.1 to 4.7 mm with inter‐Beacon distances from 14.4 to 40.4 mm. The mean Beacon rotation angle reported by the Calypso system in each plane was 4.7±3.8°, which were likely caused by the inter‐Beacon distance change. After the best 3D matching of the Beacons, our method reported a mean rotation angle of 1.5±1.7° for all the fractions. If the projected inter‐Beacon distance was >9mm there was good agreement between our method and the Calypso system. As the inter‐Beacon distance decreased, the uncertainty in the reported rotation angle increased. Conclusion: The actual rotation angles for the prostate were smaller than those for the Beacons reported by the Calypso system. It is recommend to avoid small (<9mm) projected inter‐Beacon distances to reduce the uncertainty in the reported rotation angle.
MO‐EE‐A3‐02: Inferring Nodal Volume and Primary Tumor Positions From Multiple Anatomical Surrogates Using 4D CT in Stage III Lung Cancer36(2009); http://dx.doi.org/10.1118/1.3182260View Description Hide Description
Purpose: To investigate the feasibility of modeling primary tumor and nodal volume positions from anatomical surrogates in order to reduce the contouring burden of planning from 4D CTs in Stage III lungcancer.Method and Materials: To localize their centroid positions at each respiratory phase, we contoured nodal volumes and primary tumors in 16 Stage III lungcancer planning 10‐equal‐phase 4D CTs. We also localized a series of anatomical respiratory surrogates (carina, xyphoid, nipples, midsternal external marker) in each image. To explore the feasibility of our proposed method, we 1) characterized the correlations between target and surrogate 3D motion, 2) applied Ordinary Least‐Squares (OLS) and Partial Least‐Squares (PLS) to a random subset (3–8) of images to predict the target positions in the remaining images, 3) determined the best set of three respiratory phase bins to contour, and 4) used them to create 3‐phase models using either all anatomical surrogates or carina alone. Results: The surrogate best‐correlated to target motion was most often the carina but varied widely. Depending on the number of phases used to build the models, mean errors ranged from 1.0mm to 1.4mm and from 0.8mm to 1.0mm for OLS and PLS, respectively. When the 0%, 40%, and 70% respiration phases were used, 3‐phase models had mean(±standard deviation) errors of 0.8±0.5mm and 2.4±9.0mm for models based on all surrogates and carina alone, respectively. For target coordinates with motion>5mm, the mean 3‐phase error was 1.3mm. Conclusion: Using only three contoured respiratory phases to train the models, the mean model error was on the order of CT resolution. Inferential modeling of the primary tumor and nodal volume positions may have the potential to decrease the time required to process 4D CT scans, thereby improving therapy by allowing for incorporation of patient‐specific margins in the planning process.
MO‐EE‐A3‐03: Ordinary Least Squares and Partial Least Squares for Intra‐Fraction Lung Tumor Motion Modeling36(2009); http://dx.doi.org/10.1118/1.3182261View Description Hide Description
Purpose: For accurate operation, real‐time tumortracking devices for radiation therapy require the real‐time position of the radiation target. In this study, we assess Ordinary‐Least‐Squares (OLS) and Partial‐Least‐Squares (PLS) modeling methods for inferring intra‐fraction motion from external markers. Method and Materials: We obtained the concurrent 3D positions of three optically tracked external markers affixed to the skin and the 3D centroid position of a set of three internal fiducials implanted in lungtumors localized with fluoroscopy by the Cyberknife system. We analyzed 134 treatment fractions from 63 patients, each including 40–112 (mean=62) samples spaced at approximately 1–2min. For each fraction, we used a randomly selected subset of N (4–35) points to train OLS and PLS models to infer tumor motion from the positions of the optical markers, and we repeated this process 40 times for each fraction and each N. We then tested the models against the remaining datapoints in that fraction to determine the position error. Results: The PLS mean(±standard deviation) errors decreased monotonically as N increased, from 0.3±2.6cm at N=4 to 0.2±1.4cm at N=35. In contrast, the OLS error peaked (mean=5.3cm) at N=10 training samples, a consequence of the Moore‐Penrose pseudo‐inverse regression technique. OLS errors at N=4 and N=35 were 0.4±4.6cm and 0.2±1.7cm, respectively. PLS and OLS mean and maximum errors converged for large N (approximately N⩾20). To achieve mean errors less than 0.25cm or 0.20cm over the entire dataset with PLS, at least 8 or 18 training samples, respectively, must be used. Conclusion: The results of this study indicate that PLS shows potential as an efficient (few image acquisitions) and accurate (2–3mm) intra‐fraction lungtumor motion modeling technique. Future work will focus on investigating methods for and consequences of non‐random training sample selection.
36(2009); http://dx.doi.org/10.1118/1.3182262View Description Hide Description
Purpose: Teletherapy of an unconstrained eye poses risks because of the likelihood of substantial eye motion. A new 100kVp tabletop stereotactic radiosurgical unit has recently been developed to deliver highly‐collimated beams of x‐rays to the retina. Here we evaluate the efficacy of a hardware and software motion‐management solution designed for that system. Method and Materials: To minimize the effects of motion during therapy, an eye fixation device (the “I‐Guide”) was developed. It consists of a contact‐lens, a central post, and a fixed, horizontal stabilizer bar. The contact lens is held to the eye with suction and attached to the central post. The post is mounted to stabilizer bar via a ball‐joint. The I‐Guide steadies the eye and incorporates reflective fiducials that are used by an imaging system to monitor the position and gaze‐angle of the eye. During treatment, software continuously evaluates the deviation of the eye from nominal position. When necessary, the beam is automatically gated off by an algorithm that incorporates time as well as displacement, allowing brief excursions while still maintaining effective targeting. Results: The motion‐management system was evaluated in a clinical trial of 40 patients. With the I‐Guide in place, rapid eye motion was damped considerably. Because of the time component, the gating algorithm permitted brief, random excursions while stopping the beam when excess drift was apparent. With a 3‐second x‐ray source ramp‐up, this flexibility was important for efficient treatment flow. Although patients' heads were immobilized, head mobility was found to be the dominant source of motion. However, the automatic gating‐off of the beam followed by repositioning kept the average motion error under 0.4mm in all cases. Conclusion: The combination of I‐Guide fixation, fiducial tracking, and automated beam gating enables precise targeting and dose conformality for ophthalmic radiotherapy.
Research sponsored by Oraya Therapeutics
MO‐EE‐A3‐05: Improving the Temporal Resolution of Dynamic MRI by Deformable Alignment of the Peripheral K‐Space36(2009); http://dx.doi.org/10.1118/1.3182263View Description Hide Description
Introduction Dynamic MRI may be used to image motion that is needed for accurate radiotherapytreatment planning. The tradeoff between temporal and spatial resolution has been a major issue affecting the use of dynamic MRI. Here we propose to improve temporal resolution by reconstructing high spatial resolution components with deformably registered motion data. Dynamic MRI with increased temporal resolution without scarifying spatial resolution for several sites are presenting. Methods and Materials High and low spatial resolution components of the image typically move similarly, i.e. share most of the motion vector information. The motion vectors can be obtained by deformable registration of the central k‐space, and applied to the high spatial resolution data. This allows one to acquire the peripheral k‐space at a lower temporal resolution than the central k‐space therefore increase the overall sampling rate. We tested the possible gain by this method for 2‐dimensional dynamic MRIs (SIEMENS 3T Verio) of swallowing, breathing and peristalsis. Results The mean error in the displacement vectors obtained using only the central 20% of k‐space vs. using the full k‐space was less than 0.5mm, which reduced to 0.2mm if 50% of the radius is used. Peristalsis motion with sudden changes shared least information between central and peripheral k‐space. The reconstructed images with as low as central 20% of the k‐space radius are very similar to imagesreconstructed with full k‐space. The resulting probability distance histograms and regions of interest show minimal variation. Conclusion Temporal resolution can be improved considerably without compromising the spatial resolution if the low‐resolution vector information is used to construct high spatial resolution data as demonstrated in this work. The deformable registration accuracy is highly critical for the overall performance of the method. Conflict of Interest: This work is supported partially by Siemens Healthcare.
MO‐EE‐A3‐06: Comparison of Pre‐ and Post‐Implant Prostate Volume Segmentation Using Trans‐Rectal Ultrasound and Computed Tomography36(2009); http://dx.doi.org/10.1118/1.3182264View Description Hide Description
Purpose: To quantitatively evaluate differences in prostate volume measurements via transrectal ultrasound(TRUS) and computed tomography(CT) used for dosimetric planning in prostate brachytherapy.Method and Materials: Prostate volume measurements were made in 83 prostate brachytherapy patients. One to two weeks prior to implantation, patients were placed in the dorsal lithotomy position, and prostate volumes were measured via TRUS. Immediately afterward, patients were rescanned using CT to allow for comparison of CT and TRUS measurements. On treatment day, patients were rescanned via TRUS before and after implantation and were rescanned by CT within an hour after procedure completion. The prostate and urethra were delineated by the same physician on all scans, and prostate volumes were calculated. Results: Substantial correlation was found between pre‐/post‐ implant prostate US volumes (R2=0.7998), pre‐/post‐ implant CT volumes (R2=0.7872), and pre‐implant US and post‐implant CT volumes (R2=0.8352). However, it was found that the difference between pre‐ and post‐implant volume measurements varied with the combination of modalities compared. For pre‐ vs post‐implant TRUS, post‐implant prostate volumes averaged 26.1% larger than those measured prior to treatment. For pre‐ vs. post‐ implant CT, post‐implant measurements averaged 58.9% greater. Post‐implant CT measurements averaged 23.8% greater than post‐implant TRUS measurements. Conclusion: While it would be ideal to use the same method to measure pre‐ and post‐implant volumes, this is not generally practiced. Prostate volumes measured using TRUS and CT show a distinct relationship, but it is important to note that this does not mean that the modalities provide the same values for the same prostate. In addition, the relationship between pre‐implant TRUS and CT is not identical to that between post‐implant TRUS and CT. Therefore, it is not possible to state that accurate comparison of pre‐ and post‐implant dosimetric plans can be accomplished when different imaging modalities are used for each.
- Moderated Poster ‐ Target and Normal Organ Delineation
MO‐FF‐A3‐01: Advantages of Multi‐Shot, Variable‐Density Spiral, Diffusion‐Weighted MR Imaging for Radiation Treatment Planning36(2009); http://dx.doi.org/10.1118/1.3182289View Description Hide Description
Purpose: The prerequisite for intensity‐modulated radiation therapy(IMRT) is accurate target and critical structure definition. Diffusion‐weighted MRI provides information that may improve target delineation and, therefore, radiation treatment outcome. However, geometric distortion artifacts have hampered the use of diffusion‐weighted MRI in radiation treatment planning. We demonstrate here a multi‐shot (i.e. interleaved), variable‐density spiral method that not only diminishes geometric distortion artifacts, but also permits acquisition of higher resolution diffusion‐weighted images for use in radiation treatment planning. Method and Materials: A custom, multi‐shot, spin‐echo, variable‐density spiral sequence with bipolar diffusion gradients was implemented on a 3.0T GEscanner. High‐resolution, multi‐shot, diffusion‐weighted images were acquired on three healthy volunteers (FOV: 240mm2, matrix: 2562, interleaves: 32, TE: 50 msec, TR: 3000 msec, scan time: 3.2 minutes). To demonstrate the inherent self‐navigating capabilities of variable‐density spiral, diffusion‐weighted images were reconstructed without and with self‐navigated phase correction. For comparison to the current state‐of‐the‐art method, diffusion‐weighted images were also collected using a clinical, single‐shot, spin‐echo echo‐planar imaging (EPI) sequence with ramp sampling, partial k‐space, and parallel imaging (FOV: 240mm2, matrix: 1282, TE: 64.7 msec, SENSE, R=2, scan time: ∼1 minute). Results: The inherent self‐navigating capability of variable‐density spiral was effective at diminishing motion‐induced phase errors, encoded by the diffusion gradients, that varied from shot‐to‐shot. The multi‐shot diffusion‐weighted images exhibited good signal‐to‐noise ratio despite the 47% reduction in voxel volume compared to the diffusion‐weighted EPI images. Off‐resonance effects resulted in geometric distortions in ventral and rostral brain regions on diffusion‐weighted images collected with EPI, but did not compromise the geometric integrity of diffusion‐weighted images acquired with multi‐shot spiral. Conclusions: Multi‐shot, variable‐density spiral permits acquisition of high‐resolution, diffusion‐weighted images that do not suffer from geometric distortions that have previously prohibited their use in radiation treatment planning.
36(2009); http://dx.doi.org/10.1118/1.3182290View Description Hide Description
Purpose: During a course of radiation therapy for the prostate, organs such as the bladder and rectum may change shape significantly from day to day, causing the delivereddose to differ from that planned. Deformable image registration has been used to overlap structures with different shape. In this project, we propose a method to validate the correspondence accuracy of a deformable registration algorithm using images of previously treated prostate patients. Method and Materials: The planning and one daily treatmentCTimages for each of five patients with three Calypso transponders implanted in their prostate were retrospectively selected for the study. Two CT volumes of each patient were pre‐processed to mask the fiducials. A fluid‐based image registration method was used to register images. The displacement fields obtained were applied on the original images with fiducials. The overlap of prostate volumes between the registered images was evaluated by the coincidence index (CI). The centroid positions of each transponder were calculated. The distance between the corresponding transponders was used to quantify the accuracy of correspondence of image registration.Results: The fluid deformable image registration produced satisfactory results of the overlap of the prostate volumes. The mean CI is 95.3%, with 91.5% as minimum and 98.7% as maximum. The average distance between the centroid of all pairs of transponders is 0.35 mm. The displacement is related to the quality of image registration. For image registration with CI 98.7%, all displacement of three pairs of transponders is within 0.22 mm. When less overlap is achieved, CI 91.5%, one pair of transponder is 0.78 mm apart. Conclusion: We have used real patient data to validate the correspondence of the fluid deformable image registration method. For the prostate, it has maintained sub‐millimeter mapping accuracy, which is dependent on image registration quality.