Volume 34, Issue 6, June 2007
Index of content:
- Joint Imaging/Therapy Scientific Session: Room M100J
- Motion Modeling
34(2007); http://dx.doi.org/10.1118/1.2761332View Description Hide Description
Purpose: To examine the influence of phase delays and baseline drifting on residual motion in gated radiotherapy.Method and Materials: Four patients with hepatocellular carcinoma were studied within a phase I clinical trial for gated proton beam radiotherapy. Between 3 and 5 gold fiducial markers were implanted in or near the tumor site to serve as internal surrogates for the treatment target. The Varian Real‐time Position Management (RPM) system was used as an external surrogate for respiration. The RPM was synchronized to an orthogonal bi‐plane fluoroscopic imaging system, and simulation sessions were conducted under free breathing. The three‐dimensional coordinates of each marker were tracked retrospectively, and compared with the synchronized RPM signal. The phase delay from internal position to external surrogate, and baseline drifting of the exhale position over the course of several minutes were measured, and their impact on gated treatment was assessed. Results: Phase delays were found, ranging from −0.02 seconds to +0.22 seconds, indicating that motion inside the liver may not be seen immediately on the external surface. Phase delays varied from marker to marker, and also over time during the course of 4 to 6 minutes. Baseline drifting was also found in two of the four patients, indicating that the exhale position of the liver may not be stable within the first several minutes after lying on the treatment couch. The baseline drift was usually from inferior to superior, with values ranging from −1 mm to 6 mm. The worst‐case residual motion of an amplitude‐gated treatment was estimated to be 7.4 mm. Conclusion: We observed phase delays of more than 100 ms in two of four patients, and baseline drifting of 5 mm or more in two of four patients. The utility of external surrogates for gating should be studied in more detail.
34(2007); http://dx.doi.org/10.1118/1.2761333View Description Hide Description
Purpose: To develop and evaluate a template matching algorithm for direct detection of a lungtumor trajectory in the raw projections acquired from a cone beam CT.Method and Materials: A planning respiration‐correlated CT (pRCCT) was used to construct digitally‐reconstructed radiographic(DRR) templates for registration with raw projection data of a free‐breathing cone beam CT (FB‐CBCT). The pRCCT and FB‐CBCT were first registered to obtain the approximate mean 3D position of the tumor. This position was applied as a shift to the planning isocenter in the pRCCT. Next, 72 DRRs were generated at 5 degree angular increments around this shifted position, for each phase in the pRCCT. The appropriate phase DRR template for registration was selected using the measured phase of the FB‐CBCT projection. This phase was measured by tracking the diaphragm apex position in the projections. Two template matching algorithms were compared: normalized cross correlation and a block normalized cross correlation. A dynamic phantom was induced to move reproducibly 2cm during planning and 3cm during online CBCTimaging. The phantom was imaged at isocenter and with a 1cm translational shift from isocenter. Results: The tumor excursion measured with the normalized cross correlation method was 28.7 mm, and 28.5 mm with the block normalized cross correlation. The error in detecting the daily standard deviation of tumor position was 0.1mm for both methods. The mean absolute difference in detected tumor position was 1.07 mm, with a 95% confidence value of 2.43 mm. Block normalized cross correlation reduced the maximum error in tumor position detection slightly, in relation to normalized cross correlation. Conclusion: Direct detection of the tumor position from cone beam CT projections is feasible with template matching, provided that appropriate and accurate templates are used. Conflict of Interest: Supported in part by NIH R01 CA116249 and Elekta, Inc.
TU‐C‐M100J‐03: Objective Assessment of Deformable Image Registration in Radiotherapy — a Multi‐Institution Study34(2007); http://dx.doi.org/10.1118/1.2761334View Description Hide Description
Purpose: The looming potential of deformable alignment tools to play an integral role in adaptive radiotherapy suggests a need for objective assessment of these complex algorithms. Current studies of accuracy require analytically generated deformations applied to sample image data, or use of contours or bifurcations, with inherent uncertainty as well as potential bias, as these visible features are likely to dominate the local deformation parameters. Method and Materials: A deformable phantom was embedded with 48 small radiopaque markers, placed in regions varying from high contrast to roughly uniform regional intensity, and small to large regional discontinuities in movement. CT volumes of this phantom were acquired in two different deformation states. After manual localization of marker coordinates, images were edited to remove the markers. The resulting image volumes were sent to 4 collaborating institutions, each of which has developed previously published deformable alignment tools routinely in use. Alignments were done, and applied to the list of reference coordinates. The transformed coordinates were compared to the actual marker locations. Results: 5 alignment techniques were tested from the 4 institutions. All algorithms performed generally well, as compared to previous publications. Average errors in predicted location ranged from 1.7–3.9 mm, depending on technique. No algorithm was uniformly accurate across all regions of the phantom, with maximum errors ranging from 5.5 – 15.4 mm. Larger errors were seen in regions near significant shape changes, as well as areas with uniform contrast but large local motion discontinuity.Conclusion: An objective test of deformable alignment is feasible. Reasonable accuracy was achieved, although variable errors in different regions suggest caution in globally accepting the results from deformable alignment. Valuable feedback from investigators is being applied to development of further generations of deformable phantoms for alignment experiments.
Sponsored by NIH P01‐CA59827.
TU‐C‐M100J‐04: Determination of Prospective Displacement Gate Threshold for Respiratory‐Gated Radiation Delivery From Retrospective Phase‐Based Gate Threshold Selected at 4D CT Simulation34(2007); http://dx.doi.org/10.1118/1.2761335View Description Hide Description
Purpose: 4D‐CT imaging is based on retrospective respiratory phase sorting of an external motion monitor. Subsets of 4D‐CT images, encompassing a few respiratory phases are increasingly being used in respiratory‐gated radiotherapy to reduce internal tumor displacement. Such subset selection is equivalent to selection of a retrospective phase‐based gate threshold on the external motion monitor. Respiratory‐gated radiation delivery, however, is based on a prospective respiratory displacement threshold. We aim to establish a relationship between the retrospective phase‐based gate threshold selected at simulation and prospective displacement‐based gate threshold that should be applied during treatment delivery. Method and Materials: Over 150 external respiratory motion traces, from 90 patients, with and without audio‐visual biofeedback were analyzed. Respiratory gating phase interval was chosen from the 4D‐CT‐derived tumor motion trajectory. Retrospective displacement gate threshold for the selected phase interval was defined as the average respiratory displacement within the phase interval. By including displacement values that were both within the selected phase interval and retrospective displacement gate threshold, a prospective displacement gate duty cycle was obtained. The external respiratory displacement value that yielded the prospective displacement gate duty cycle was the prospective displacement gate threshold for gated radiation delivery. Results: Phantom motion tests yielded coincidence of retrospective and prospective gate thresholds within 0.3%. Patient data analysis indicated that the average difference between prospective and retrospective gate thresholds was 8±15% and 4±7% of the total respiratory motion range for smaller and larger phase intervals respectively. With audio‐visual biofeedback, corresponding differences were 10±14% and 9±10% respectively. Conclusion: A relationship between retrospective phase‐based gate threshold at simulation and prospective displacement gate threshold at radiation delivery was established and validated for regular and non‐regular respiratory motion. Using this relationship, the prospective external displacement gate threshold to be applied at radiation delivery can be reliably estimated from 4D‐CT simulation.
TU‐C‐M100J‐05: Image Guided 4D Monte Carlo Study of the Dosimetric Effects of Intra/inter Fraction Motion in Lung Tumors34(2007); http://dx.doi.org/10.1118/1.2761336View Description Hide Description
Purpose: Application of an image guided 4D Monte Carlo framework to the evaluation of intra/inter fraction motion effects between patient organ‐movement and dynamic delivery effects Materials & Methods: An IGMC toolkit that accounts for time‐dependent geometries of organ motion and dynamic delivery and provides accurate dosimetry was developed for lungcancer patients. Fluence calculation for VARIAN‐2100C/D photon accelerator is performed with EGSnrc and for patient dose calculation we use dose‐planning‐method. The time dependent beam delivery information is obtained from the treatment‐planning program in the form of MLC leaf‐sequencing files, while the organ motion pattern was obtained with the use of RPM signal and 4D‐CT. A voxel displacement map is used to quantify the motion of the organ in the voxelized geometry. The study was performed for lung patient for several breathing phases. The data analysis was performed with the freely available package VODCA. Results: Differences between the conventional CORVUS dose plan and the Monte Carlo dose results suggest that Monte Carlo dose calculation is vital for assessing effects in lungtissue. For the free breathing CT, CORVUS was under‐dosing the lung‐tumor by approximately 3Gy. When accounting for temporal effects, comparison between IGMC dose distributions for free breathing, inhale and exhale phases, show that dose coverage of the primary/secondary tumors was significant worse in the inhale phase relative to the exhale/free‐breathing phases. For inhale phase, 5–10% of the volume was receiving 10Gy less than in exhale or free‐breathing phases, leading to large cold spots inside the tumor. This is mainly due to the motion of the diaphragm, which subsequently moves the tumor in the superior‐inferior by more than 2 cm. Conclusions:Image guided 4DMC methods can significantly improve the planning of lungtumor treatments by accurately modeling the motion of the tumors and large heterogeneity in the tumor region. Supported NIH/NCI‐R01/CA111590.
34(2007); http://dx.doi.org/10.1118/1.2761337View Description Hide Description
Purpose: To predict tumor position in order to compensate for temporal latency of a target tracking‐based dynamic radiation dose delivery system.Method and Materials: A linear adaptive filter was trained for real‐time prediction of tumor target position. Signal history was first filtered by a low‐pass filter (LPF). Template matching was used in order to select training examples that were closest to the current case, using the distance metric: where, This algorithm was applied to 160 3D tumor motion datasets acquired from 46 patients. Algorithm parameters were selected through a comprehensive sensitivity analysis performed on 10% of the datasets that exhibited maximum peak‐to‐peak motion. The resulting parameters were used to analyze algorithm performance over the remaining 90% of the datasets. Root mean square (RMS), maximum, mean and standard deviation of errors with and without prediction were calculated. The error distribution was tested against a Gaussian distribution using the Kolmogorov‐Smirnov (KS) test. Results: Sensitivity analyses showed that a third‐order Butterworth LPF having cutoff frequency of 2 Hz was the most suitable. Examination of error indicates that it has close‐to‐zero mean for all the three dimensions and standard deviation is of the order of 0.1 mm. Moreover, the D‐statistics of KS test are close to 0.1 (indicating a distribution that closely approaches a Gaussian) except for cases where tumor motion itself is very small (less than 5 mm). Compared to no prediction, the algorithm reduced RMS and maximum errors, on average, by 74.58% and 55.39% respectively. Conclusion: We have developed a robust and highly efficient algorithm for predicting respiratory tumor motion. Future work includes integration of the algorithm with the laboratory setup of a dynamic MLC‐based target trackingsystem.
TU‐C‐M100J‐07: A Systematic Study On the Sources of Drift in a Turbine‐Based Spirometer Using a Breathing Simulator34(2007); http://dx.doi.org/10.1118/1.2761338View Description Hide Description
Purpose: To systemically isolate and quantify the contributions of different sources of signal drift in turbine‐based spirometry. Method and Materials: To get a repeatable response from the VMM‐400 spirometer, we used a breathing simulator made of an airtight cylinder. The cylinder's piston was driven by a motor to force the air in/out of the cylinder to the spirometer. A heating blanket was used to heat the cylinder, so that the in/out air would have different temperature. The piston position, thereby the cylinder air volume, was determined using a position sensor.Results: Our data show that even when the piston was driven sinusoidally and the heating blanket was off, the spirometer exhibits a drift per cycle of 3.4% of the maximum tidal air volume due to the differential response of its turbine blade. Reversing the direction of in/out flow simply changes the drift direction. With the heating on, the drift accumulates an additional 4% per cycle. The added drift is due to the expanded air volume from the heating. The most significant drift was observed when the piston was driven in a saw‐tooth pattern, either with a fast inhale followed by a slow exhale or visa versa. The difference in the measured volume between the two breathing phases can be as large as 60% or more due to the failure of the spirometer to register the volume in the low flow‐rate phase and the air needed to be spent on reversing the blade at the end of the fast‐changing phase. Conclusions: The drift due to the blade asymmetry and temperature stays the same per breathing cycle (3.4% and 4%), and can be corrected in real‐time. The drift due to breathing asymmetry can vary between cycles because of patient irregular breathing; the correction would most likely be complex (i.e., nonlinear).
TU‐C‐M100J‐08: Objective Characterization, Estimation and Prediction for Modeling Breathing‐Related Movement34(2007); http://dx.doi.org/10.1118/1.2761339View Description Hide Description
Purpose: To propose a hierarchical model for estimation, tracking, and prediction of respiratory tumor motion. To incorporate modeling on different scales: semi‐reproducibility globally and slow frequency/displacement variation locally. Method and Materials: The problem is formulated with a hierarchy of scales: On the finer scale, a databased approach is used to estimate the local variation of both displacement and frequency, utilizing classic control and chaostheory. A warping procedure is used to “counteract” local variation, resulting in a much more regular post‐warping signal. On the global level, the post‐warping (phase‐synchronized) signal is modeled as a noisy observation of an intrinsic periodic system, and the best periodic pattern is estimated within a nonparametric optimization setting. For tracking and prediction purposes, the locally estimated warping map (together with proper interpolation or extrapolation, whichever applies) is used to un‐warp the globally obtained periodic pattern. A recursive method is devised to further improve the efficiency for real‐time processing. Results: The obtained estimation/prediction signal demonstrates similar local variation as the raw observation, while semi‐periodicity is incorporated to decrease its noise sensitivity and enhance prediction accuracy. Verification using RPM data shows that the proposed method reduces 1‐period look‐ahead prediction error (RMSE) by more than 50% compared to perfect periodic modeling. Conventional local linear models generally fail in such long‐term prediction tasks. Conclusions: This work provides an infrastructure for incorporating information on different knowledge levels, and offers the flexibility to adaptively balance the roles of physical prior knowledge and data fidelity. The model‐based method on the global level incorporates the well‐recognized semi‐periodicity pattern of respiratory motion, overcoming the myopia of local state models. Data‐driven local phase map estimation used in warping and un‐warping fully utilizes the observation and enjoys the freedom of nonparametric setup.
This work is sponsored by NIH P01‐CA59827.
34(2007); http://dx.doi.org/10.1118/1.2761340View Description Hide Description
Purpose: Accurate motion tracking can decrease normal tissuedose, PTV margins, and shorten treatment times required for radiation therapy. A Periodic Autoregressive Moving Average (PARMA) model was developed to predict one‐dimensional respiratory motion in human subjects. Methods and Materials: A PARMA prediction algorithm was applied to clinical respiration signals from three categories: free‐breathing, audio instruction, and audio‐visual coaching. The PARMA model of the respiration motion is a dual‐component signal, a partially correlated time‐series superimposed onto a periodic pattern. The periodic component was estimated by averaging over the inhale‐exhale cycles in a 60‐second sample of the respiration signal, and the autocorrelation in the residual signal was modeled by linear autoregressive moving average equations. The PARMA predictions were compared to the system latency errors produced when tumor tracking is done without forward predictions. Results: For audio‐visual coaching, the PARMA prediction errors (1σ) were 0.09 mm, 0.23 mm and 0.99 mm at 0.1 s, 0.2 s and 0.5 s respectively. A prediction error of 0.99 mm is equivalent to 7.2% of the average end‐inhale to end‐exhale motion. By comparison, the errors were 0.60 mm, 1.20, and 2.90 mm when the lagged signal is used as an estimate of current position at 0.1 s, 0.2 s and 0.5 s respectively. For free‐breathing respiration, PARMA predictions were accurate to within 0.10 mm (1σ) at 0.1 seconds, 0.24 mm at 0.2 seconds, and 1.07mm at 0.5 seconds. The system latency errors were also similar to the coached signals. Conclusion:Tumor tracking with dynamic MLC can theoretically achieve system latency times of 0.16 seconds with accuracy of about 1 mm. The prediction errors of the PARMA algorithm was well under one‐third of the calculated system latency error for all lag times investigated.
34(2007); http://dx.doi.org/10.1118/1.2761341View Description Hide Description
Purpose: The purpose of this study was to develop a novel technique for dynamically predicting respiration motion and uncertainty up to 1.5 seconds in the future in real‐time using Time‐Delay Kernel Regression (TDKR) modeling. Unlike neural network based prediction techniques, kernel regression models are continuously learning new respiration cycles for each patient without the need for computationally and time intensive re‐training. Method and Materials: Recorded respiration data from a real‐time respiratory gating system was used to develop a model to predict the amplitude of the marker block at a future time. An empirical TDKR model was designed to compensate for the latency that occurs between the acquisition of the image and the point at which the beam actually turns on/off in beam tracking systems. This non‐parametric model incorporates the temporal information present in the input data. The model was tested using respiration data from 4 patients. Results: The rootmean squared error (RMSE) between the model prediction's and the measured data was computed for each patient at different latencies, and then the average was taken over all the patients. For predictions 1.5 seconds into the future the average RMSE was 1.4%. For predictions 1 second into the future, the RMSE dropped to 1.2%, and for 0.5 seconds it was only 0.7%. The average uncertainty for the predictions at 0.5, 1, and 1.5 seconds into the future was 2.4%, 3.2%, and 3.4%, respectively. Conclusion: This study proves that a TDKR model can learn the relationships present in respiration data. The reported results showed that the TDKR model has the same, if not better predictive performance, as previously studied parametric models. However, because TDKR is non‐parametric, it has several distinct advantages over these models that make it more suited for respiratory gating applications.
- Correction Strategies, Interventional Procedures
34(2007); http://dx.doi.org/10.1118/1.2761649View Description Hide Description
Purpose: To investigate the optimum requirements of magnetically shielding a Linac from an MRI coupled to it for the purpose of real‐time image‐guided adaptive radiotherapy. In addition, to investigate the effects on the MRI'smagnetic field homogeneity due to the presence of the required shielding. Method and Materials:Finite element method was used (COMSOL MultiPhysics) to model a Linac coupled to a 0.2 T bi‐planar MRI system, and to calculate the magnetic fields in its vicinity. Associated shielding was optimized by varying sheets of shielding material (eg, iron, Mu Metal) to keep the MRI's magnetic fringe fields within the waveguide to less than 0.5 Gauss. Results: We found that to reduce the MRI's magnetic fringe fields within the waveguide from 20 Gauss (unshielded waveguide) to the specific requirement of less than 0.5 Gauss, we require a 5 cm thick iron plate between the Linac and the MRI, and a 1 mm Mu Metal wrapping around the waveguide. An additional 5 cm plate is placed on the opposing side to improve magnetic field symmetry. The magnet's field inhomogenieties resulting from the shielding were calculated to be less than 142 ppm, and thus easily “shimmable”. Conclusion:Magnetic shielding is needed for operating a Linac coupled to a 0.2T MRI — system. We have optimized the shielding required to reduce the MRI's magnetic fringe fields within the waveguide to less than 0.5 Gauss by using simple and realizable passive shielding. Field inhomogeneities due to the presence of our shielding are sufficiently small that they can be easily corrected though conventional shimming techniques.
34(2007); http://dx.doi.org/10.1118/1.2761650View Description Hide Description
Purpose: Topotherapy is a new delivery technique that uses the binary MLC to implement static beam IMRT. We present a simple approach for real time motion adaptive delivery (MAD) of topotherapy plan using the current delivery system. Methods: Topotherapy uses fixed jaws and binary MLC for intensity modulation and its projection rate is a few Hz. Topotherapy delivery is controlled by a planned projection‐wised leaf sequence, which is optimized assuming stationary target. The couch proceeds in constant speed. To compensate longitudinal tumor motion in real time, instead of sequential execution of the planned leaf sequence, the projections are out of order executed. That is, a previous or future projection that with the same planned target position as the current status, is chosen instead. The transversal target motion is further compensated by shifting and scaling the MLC leaf open time at the chosen projection. Results: Extensive simulations with realistic topotherapy parameters and respiration motions validate MAD technique. The delivered dose conformed well to the target regardless how target moved during delivery. Significant margin reduction could be approached provided that real time tumor localization is achievable. Discussions and Conclusions: We present a novel technique for real time MAD in topotherapy. Unlike the dynamic MLC based tracking method, this technique requires neither the whole target motion trajectory nor the velocity of target motion. Instead, it only requires instantaneous target positions, which greatly simplifies the system implementation. This technique re‐uses the planned leaf sequence by re‐ordering its projection and leaf sequence. It does not involve time consuming re‐optimization. The simulations validated the presented technique. Experimental validation and extension to helical tomotherapy delivery is presented in another paper of this conference.
34(2007); http://dx.doi.org/10.1118/1.2761651View Description Hide Description
Purpose: The purpose of this study was to evaluate the daily setup error of small peripheral lungtumors based on implanted markers and stereoscopic x‐rays. Method and Materials:Twenty‐three patients with small peripheral lungtumors were implanted with radiographic markers for localized radiation therapy. Exhale CT scans were used for planning. At treatment, the patients were aligned to skin marks followed by exhale synchronized stereoscopic x‐rays (Exactrac™, BrainLAB). Setup error is evaluated as the shift from initial skin alignment to the implanted x‐ray alignment. The mean, σ mean, and RMS of σ for this data are reported. Results: A total of 790 daily treatment sessions on 23 patients were recorded. In all patients, an evaluation CT of the implanted marker was performed prior to treatment to verify marker location. The mean and σ for the 23 patients from initial setup to x‐ray localization were 0.2 ± 3.8 mm laterally, 1.6 ± 5.3 mm longitudinally, and −0.7 ± 7.8 mm vertically. The RMS error was Lat 5.5 mm, Long 7.7, and Vert 6.4 mm. Conclusion: A large patient study using implanted markers for lungtumor alignment indicated large initial setup shifts from skin marks. Daily random errors on the order of 8mm were reported. The range of these shifts is as large as 3.0 cm in some instances indicating that a small target would potentially be missed without better setup or with localization. The large systematic error is most likely due to the difference in tumor position from exhale breath hold (CT) relative to free breathing exhale (X‐Ray)..Conflict of Interest: This work is supported in part from a research grant from BrainLAB, Inc.
34(2007); http://dx.doi.org/10.1118/1.2761652View Description Hide Description
Purpose: For on‐line positioning corrections based on fiducial markers we have developed Stereographic Targeting (SGT). SGT is described and clinical results for prostate patients are reported. Method and Materials: SGT entails (1) acquisition of a MV and orthogonal kV image in rapid succession without gantry rotation, (2) marker detection and calculation of center of mass (COM) translation and rotations, (3) correction execution by remote couch control and (4) optional repetition of previous steps for verification. MV images are acquired with 2–6 MU of a treatment beam and kV images with Elekta Synergy®. Performance was measured for 10 prostate patients with 3–4 markers who had daily SGT with verification imaging plus an extra kV image at the end of each fraction to study intrafraction motion. Results: Compared with manual analysis, automatically determined marker positions in SGT were accurate within 0.5, 0.2, and 0.3 mm (SD) for the lateral (LR), cranial‐caudal (CC), and anteriorposterior (AP) directions respectively and within 1°(SD) for rotations. SGT added < 1 min to treatment time while it reduced COM systematic errors (Σ) from 4 mm to 0.5 mm (SD) and random errors (σ) from 3 mm to < 0.8 mm on all axes. The largest prostate rotations were about the LR axis: Σ = 3.4° and σ = 4.5°. The start‐to‐end intrafraction motion was 0.5 mm (Σ) and 1.5 mm (σ) for AP and 0.5 mm (Σ) and 1.4 mm (σ) for CC. Conclusion: SGT allows for fast, precise positioning (3D deviation < 3 mm within one minute) and is now routinely used for our prostate treatments. Taking into account rotations and the mainly random intrafraction motion, planning margins <=5 mm are obtained. While intrafraction motion can be reduced by repetitive SGT within one fraction, we develop a strategy to counter rotations as well.
34(2007); http://dx.doi.org/10.1118/1.2761653View Description Hide Description
Purpose: While ART has been studied for years, the specific quantitative implementation details have not. In order for this new scheme of radiation therapy to reach its potential, an effective ART planning strategy capable of taking into account the dosedelivery history and patient's on‐treatment geometric model must be in place. This work performs a study of dynamic closed‐loop control algorithms for ART and demonstrates their utilities with data from phantom and clinical cases with on‐treatment cone‐beam CTimages.Method: In closed‐loop control, the controller is not run just once but repeatedly, each time receiving the current state of the system as its input. To meet the requirements of different clinical applications, two classes of algorithms are developed: those Adapting to Changing Geometry (ACG) and those Adapting to Geometry and DeliveredDose (AGDD). The former takes into account organ deformations found just before treatment. The latter optimizes the dose distribution accumulated over the entire course of treatment by adapting at each fraction not only to the anatomic information just before treatment but also to the dosedelivery history. The closed‐loop algorithms are showcased by phantom and clinical cases. Results: A comparison of the approaches with conventional open‐loop IMRT without adaptively incorporating feedback information indicates that closed‐loop ART may significantly improve the current practice. In both phantom and clinical studies, AGDD outperforms ACG algorithms in three aspects: target dose coverage, sensitive‐structure sparing, and steeper gradients around the tumor. Within the AGDD formalism, it is beneficial not to correct all the previous dosimetric errors at once right after the information is available but over a number of fractions until next set of feedback data is available. Conclusion:ART with closed‐loop dynamic algorithms substantially improve the dose distribution. In addition, the differing performance of the specific implementations shows that the algorithmic details matter.
34(2007); http://dx.doi.org/10.1118/1.2761654View Description Hide Description
Purpose: A prototype mobile C‐arm capable of real‐time fluoroscopy,tomosynthesis, and cone‐beam CT(CBCT) has been developed for intraoperative guidance of minimally invasive procedures. This work investigates the imaging performance, radiation dose, and image acquisition / reconstruction time associated with a range of 3D imaging techniques. In addition, protocols for intraoperative use are developed for implementation in clinical research trials. Method and Materials:Imaging performance was assessed in terms of spatial resolution as well as soft‐tissue and bony detail visualization in phantoms and cadavers. Image reconstruction via 3D filtered backprojection was performed across a range of tomosynthesis angle (10°–178°) and number of projections (5–200). The corresponding radiation dose was evaluated analytically and experimentally (using a Farmer chamber in cylindrical head and body phantoms) to quantify peripheral, central, and organ dose (e.g., dose to the eyes). Acquisition / reconstruction times were benchmarked in relation to operational time constraints. Results:CBCTimages exhibited sub‐mm spatial resolution and soft‐tissue visibility (∼20 HU) sufficient for interventional guidance at a central dose of 10 mGy. Tomosynthesis offered a useful, high‐speed adjuvant to CBCT, providing visualization of high‐contrast features at a fairly limited arc — e.g., visualization of the clivus (skull base) achieved at angles down to ∼30° (central dose ∼1.6 mGy). Reconstruction times were ∼62 sec (full CBCT volume, 0.8 mm voxels) and ∼20 sec (tomosynthesis “slab,” 0.4 mm voxels), respectively.Conclusion:C‐arm tomosynthesis and CBCT provide a valuable addition to the image guidance arsenal. The former offers fast, low‐dose 3D images sufficient for guidance with respect to high‐contrast bony features. The latter offers sub‐mm spatial resolution and soft‐tissue visibility at the cost of time and dose. Deployment of the C‐arm prototype in clinical research trials promises increased surgical precision, with protocols for intraoperative imaging consistent with operational time and dose constraints.
34(2007); http://dx.doi.org/10.1118/1.2761655View Description Hide Description
Purpose: Intra‐operative dosimetryoptimization of TRUS‐guided prostate brachytherapy requires localization of seeds relative to the prostate [Todor et al. PMB 48(9):1153–71]. Method and Materials: Seeds were reconstructed from C‐arm fluoroscopy, registered to TRUS, and exported to commercial brachytherapy system for dosimetryoptimization. Technical obstacles included pose‐dependant C‐arm calibration; distortion correction; pose estimation of C‐arm images; registering C‐arm to TRUS; and seed reconstruction. We proved that calibration is not critical and distortion correction in AP suffices. A radiographic tracking fiducial was attached in known pose over the template to recover the C‐arm pose from a single image, relative to the template. The 3D coordinates of the segmented seeds are calculated upon resolving the correspondence of seeds in the multiple C‐arm images, by formalizing seed‐matching as a network‐flow problem [Jain et al. Med Phys, 32(11):3475–92]. The seed locations are exported in template coordinates to the Interplant® commercial system for dosimetry analysis and optimization.Results: In precision‐machined hard phantoms with 40–100 seeds, we correctly reconstructed 98.5% seeds with a mean 3D accuracy of 0.63mm (0.91mm error for mismatched seeds). In soft tissue phantoms with 45–87 seeds and clinically realistic 15° C‐arm motion, we correctly reconstructed 100% seeds with 1.5mm absolute accuracy (0.25mm relative accuracy), and registered them to the prostate segmented from TRUS with an accuracy of 3.4mm (0.82mm relative). In a Phase‐1 clinical trial, so far on 4 patients with 66–84 seeds, we achieved intra‐operative monitoring of seed distribution and dosimetry. We optimized the V100 dose by inserting 3–9 additional seeds. Conclusion: Intra‐operative doseoptimization is possible with an average C‐arm, at negligible additional cost to existing clinical installations. Conflict of Interest: The work has been done in collaboration with Acoustic MedSystems for seed import into the Interplant®. Keywords: Prostate Brachytherapy, Intra‐operative Dosimetry.Support: DoD/PC‐050170, DoD/PC‐050042, NIH‐1R43CA099374, NSF‐9731478.
TH‐C‐M100J‐08: Dosimetric Effects of a 4D Magnetic Localization System for LINAC Beam Gating On Prostate and Lung Radiation Therapy34(2007); http://dx.doi.org/10.1118/1.2761656View Description Hide Description
Purpose: The Calypso® Medical 4D Localization System is capable of tracking continuous dynamic motion, and has been FDA cleared for use in the prostate. To date, automatic intervention for the measured motion is not a function of the system. We investigated use of the system to gate radiation therapydelivery on a motion phantom for both a lung and prostate fraction Materials and Methods: A Calypso gating prototype system with an optically isolated relay was connected to the BEAM_HOLD interface of a Varian Trilogy linac. A sample 3D lung plan and SMLC prostate plan were randomly selected from the active patient list. The Washington University 4D phantom was programmed to move a film box through 2 patient‐measured prostate and lung trajectories. Radiation therapy was delivered to the static phantom, to the phantom undergoing each motion trajectory without gating, and to the phantom undergoing motion trajectories while the Calypso System was gating the linac. A 4×4×4 mm gating window centered on isocenter was used for the prostate, while a linear gating window corresponding to exhalation (ie 2 mm<y<5 mm) was used for the lung.Results: Without gating beam delivery, prostate motion during treatment delivery caused approximately 10% underdosing and overdosing due to the superior/inferior and anterior/posterior prostate motion. The Calypso System triggering gated therapy delivery reduced these areas significantly, as seen in difference images using film dosimetry. Likewise, lungtumor motion caused a geometric mismatch between static and motion delivery. The Calypso System with prototype gating showed a decrease in this mismatch. Conclusions: A wireless electromagnetic implanted transponder system for linac gating was effective in reducing dosimetric errors caused by prostate and lung motion. More work is needed to define optimal gating windows to provide maximal clinical efficiency with minimal delivery error.
This work was supported by Calypso Medical Technologies.
TH‐C‐M100J‐09: MR Guided Focused Ultrasound (MRgFU) For The Treatmentment of Prostate Cancer: A Feasibility Study of Increasing Cellular Uptake of AS‐MDM2 in Vivo34(2007); http://dx.doi.org/10.1118/1.2761657View Description Hide Description
Purpose: MDM2 is an oncogene and overexpressed in 30–40% of prostate cancer. Antisense MDM2 oligonucleotide (AS‐MDM2) inhibits MDM2 expression, and enhances the effects of radiation and chemotherapy on prostate cancer. The purpose of this study is to investigate the feasibility of increasing the cellular uptake of AS‐MDM2 using MR guided High Intensity Focused Ultrasound (MRgFU). Materials and Methods: A HIFU system (InSightec ExAblate 2000) and a 1.5T MR scanner(GE) were used for this study. Human prostate cancer cells LNCaP 105, were grown orthotopically in the prostates of 11 nude mice. Extensive experiments were performed to determine the optimal MR parameters for target delineation and the optimal ultrasound parameters for animal treatment studies using an acoustic phantom. The mice bearing implanted prostate tumors (61±22mm3) were treated under general anesthesia using pulsed focused ultrasound with the output acoustic power of 4W, pulse width of 100msec and either 300 or 900 pulses in one sonication. The focal region is cigar shaped, about 2mm in diameter and 10mm in focal length. The focal peak was set within the target under the MR guidance. Two to four sonications were used to cover the whole tumor. Immediately after the treatment 0.1ml of AS‐MDM2, dissolved in PBS, was given by tail vein injection at doses 25mg/kg. After 24hr, the animals were sacrificed and tumors were removed. The expression levels of p53 proteins were analyzed by immunohistochemical staining. Results: Our preliminary results showed that the animals tolerated well the HIFU treatment. With 300 pulses per site in each sonication blood cell extravasation on H&E staining and an increase in p53 expression (5.0±2.4%) in the treatedtumor bearing LNCaP cells were observed as compared to the control group (1.3 ± 0.5%). Conclusion: MRgFU can be used an alternative treatment modality for prostate cancertreatment.
34(2007); http://dx.doi.org/10.1118/1.2761658View Description Hide Description
Purpose: To develop and employ novel image guidance methods for targeting in the stereotactic functional procedure of deep brain stimulation (DBS) in regions that are poorly defined with anatomic imaging using a deformable atlas and functional imaging.Method and Materials: An image guidance system was developed to enhance targeting for stereotactic DBS surgeries. An atlas of the structures in the basal ganglia was created from the Schaltenbrand‐Bailey series of histologically stained sagittal and axial sections. By defining a surface that connects each plane, a voxelized binary atlas was created and smoothed to reduce inconsistencies. A set of programs were created using Matlab to allow for user driven linear atlas deformation to match the atlas with patient specific anatomy and landmarks. An additional set of programs were created to record intraoperative microelectrode recording (MER) maps and to visualize these maps through sagittal and coronal cuts in the patient deformed atlas. To add additional functional information, a high resolution functional MRI (fMRI) protocol was developed that allows for localization of motor, sensory, language, and emotional regions in the basal ganglia. Software to visualize the deformed atlas, MRI and fMRI all together was created to allow for target definition and planning based off multiple sources of information simultaneously. Results: The developed atlas‐based image guidance system has been used as a clinical tool for several months and now allows physicians the ability to deform an anatomic atlas to patient specific anatomy and also obtain and view electrode tracks through the atlas in oblique angles. fMRI data on initial subjects has shown good qualitative agreement with expected physiological locations and MER maps in patients. Conclusion: This work allows for improved targeting in DBS based off the simultaneous usage of a 3D deformed atlas, microelectrode recording maps, and fMRI data.