W Lu

Development of the 4D Phantom for patient-specific end-to-end radiation therapy QA

In many patients respiratory motion causes motion artifacts in CT images, thereby inhibiting precise treatment planning and lowering the ability to target radiation to tumors. The 4D Phantom, which includes a 3D stage and a 1D stage that each are capable of arbitrary motion and timing, was developed to serve as an end-to-end radiation therapy QA device that could be used throughout CT imaging, radiation therapy treatment planning, and radiation therapy delivery. The dynamic accuracy of the system was measured with a camera system. The positional error was found to be equally likely to occur in the positive and negative directions for each axis, and the stage was within 0.1 mm of the desired position 85% of the time. In an experiment designed to use the 4D Phantoms encoders to measure trial-to-trial precision of the system, the 4D Phantom reproduced the motion during variable bag ventilation of a transponder that had been bronchoscopically implanted in a canine lung. In this case, the encoder readout indicated that the stage was within 10 microns of the sent position 94% of the time and that the RMS error was 7 microns. Motion artifacts were clearly visible in 3D and respiratory-correlated (4D) CT scans of phantoms reproducing tissue motion. In 4D CT scans, apparent volume was found to be directly correlated to instantaneous velocity. The system is capable of reproducing individual patient-specific tissue trajectories with a high degree of accuracy and precision and will be useful for end-to-end radiation therapy QA.

Quantitative prediction of respiratory tidal volume based on the external torso volume change: a potential volumetric surrogate

An external respiratory surrogate that not only highly correlates with but also quantitatively predicts internal tidal volume should be useful in guiding four-dimensional computed tomography (4DCT), as well as 4D radiation therapy (4DRT). A volumetric surrogate should have advantages over external fiducial point(s) for monitoring respiration-induced motion of the torso, which deforms in synchronization with a patient-specific breathing pattern. This study establishes a linear relationship between the external torso volume change (TVC) and lung air volume change (AVC) by validating a proposed volume conservation hypothesis (TVC = AVC) throughout the respiratory cycle using 4DCT and spirometry. Fourteen patients torso 4DCT images and corresponding spirometric tidal volumes were acquired to examine this hypothesis. The 4DCT images were acquired using dual surrogates in ciné mode and amplitude-based binning in 12 respiratory stages, minimizing residual motion artifacts. Torso and lung volumes were calculated using threshold-based segmentation algorithms and volume changes were calculated relative to the full-exhalation stage. The TVC and AVC, as functions of respiratory stages, were compared, showing a high correlation (r = 0.992 +/- 0.005, p < 0.0001) as well as a linear relationship (slope = 1.027 +/- 0.061, R(2) = 0.980) without phase shift. The AVC was also compared to the spirometric tidal volumes, showing a similar linearity (slope = 1.030 +/- 0.092, R(2) = 0.947). In contrast, the thoracic and abdominal heights measured from 4DCT showed relatively low correlation (0.28 +/- 0.44 and 0.82 +/- 0.30, respectively) and location-dependent phase shifts. This novel approach establishes the foundation for developing an external volumetric respiratory surrogate.

Dosimetric variances anticipated from breathing- induced tumor motion during tomotherapy treatment delivery

In their classic paper, Yu et al (1998 Phys. Med. Biol. 43 91) investigated the interplay between tumor motion caused by breathing and dynamically collimated, intensity-modulated radiation delivery. The papers analytic model assumed an idealized, sinusoidal pattern of motion. In this work, we investigate the effect of tumor motion based on patients breathing patterns for typical tomotherapy treatments with field widths of 1.0 and 2.5 cm. The measured breathing patterns of 52 lung- and upper-abdominal-cancer patients were used to model a one-dimensional motion. A convolution of the measured beam-dose profiles with the motion model was used to compute the dose-distribution errors, and the positive and negative dose errors were recorded for each simulation. The dose errors increased with increasing motion magnitude, until the motion was similar in magnitude to the field width. For the 1.0 cm and 2.5 cm field widths, the maximum dose-error magnitude exceeded 10% in some simulations, even with breathing-motion magnitudes as small as 5 mm and 10 mm, respectively. Dose errors also increased slightly with increasing couch speed. We propose that the errors were due to subtle drifts in the amplitude and frequency of breathing motion, as well as changes in baseline (exhalation) position, causing both over- and under-dosing of the target. The results of this study highlight potential breathing-motion-induced dose delivery errors in tomotherapy. However, for conventionally fractionated treatments, the dose delivery errors may not be co-located and may average out over many fractions, although this may not be true for hypofractionated treatments.

A novel analytical approach to the prediction of respiratory diaphragm motion based on external torso volume change.

An analytical approach to predict respiratory diaphragm motion should have advantages over a correlation-based method, which cannot adapt to breathing pattern changes without re-calibration for a changing correlation and/or linear coefficient. To quantitatively calculate the diaphragm motion, a new expandable piston respiratory (EPR) model was proposed and tested using 4DCT torso images of 14 patients. The EPR model allows two orthogonal lung motions (with a few volumetric constraints): (1) the lungs expand (DeltaV(EXP)) with the same anterior height variation as the thoracic surface, and (2) the lungs extend (DeltaV(EXT)) with the same inferior distance as the volumetrically equivalent piston diaphragm. A volume conservation rule (VCR) established previously (Li et al 2009 Phys. Med. Biol. 54 1963-78) was applied to link the external torso volume change (TVC) to internal lung volume change (LVC) via lung air volume change (AVC). As the diaphragm moves inferiorly, the vacant space above the diaphragm inside the rib cage should be filled by lung tissue with a volume equal to DeltaV(EXT) (=LVC-DeltaV(EXP)), while the volume of non-lung tissues in the thoracic cavity should conserve. It was found that DeltaV(EXP) accounted for 3-24% of the LVC in these patients. The volumetric shape of the rib cage, characterized by the variation of cavity volume per slice over the piston motion range, deviated from a hollow cylinder by -1.1% to 6.0%, and correction was made iteratively if the variation is >3%. The predictions based on the LVC and TVC (with a conversion factor) were compared with measured diaphragm displacements (averaged from six pivot points), showing excellent agreements (0.2 +/- 0.7 mm and 0.2 +/- 1.2 mm, respectively), which are within clinically acceptable tolerance. Assuming motion synchronization between the piston and points of interest along the diaphragm, point motion was estimated but at higher uncertainty ( approximately 10% +/- 4%). This analytical approach provides a patient-independent technique to calculate the patient-specific diaphragm motion, using the anatomical and respiratory volumetric constraints.

MO-D-I-611-06: Reduction of Motion Blurring Artifact Using Respiratory Gated CT: A Quantitative Evaluation

Purpose: To develop a technique for reducing respiratory motion blurring artifacts using respiratory-gated CT, and to quantitatively evaluate the artifact reduction. Method and Materials: Similar to electrocardiogram (ECG) gated imaging for the heart, a synthetic sinogram was built from multiple scans intercepting a respiration gated window. A gated CT image was then reconstructed by the filtered back-projection algorithm. CT images of wedge phantoms moving at different speeds, and 13 patients were taken with synchronized respiratory motion measurement. The scanner was operated in cine mode with 100 and 15 scans (0.5 s rotation) acquired consecutively at each couch position for phantoms and patients, respectively. Two error functions were fit to the CT profile across the air-phantom or lung-diaphragm boundaries for a quantitative evaluation of the blurring artifact. Results: The blurring artifact was reduced significantly at the air-phantom boundaries in the gated image. The gated image of phantoms with a motion of 20 mm/s showed similar blurring artifacts as the non-gated image of phantoms with a motion of 10 mm/s. The blurring artifact had a linear relationship with both the speed and the tangent of the wedge angles. The blurring artifacts were also reduced at the lung-diaphragm boundaries for patients. Centers of the two fitted error functions provided a reliable measure of large blurring, and were found equivalent to 25% and 75% locations of the CT profile. Conclusion: The respiratory gated CT imaging reduced the blurring artifacts for both moving phantoms and patients. This technique may be applied for other tomographic imaging modalities that require long imaging times with significant motion blurring artifacts, such as PET.

TU-D-J-6C-01: A Comparison Between Amplitude Sorting and Phase Sorting Using External Respiratory Measurements for 4D CT

Purpose: To compare amplitude sorting and phase sorting techniques using external respiratory measurements for 4D‐CT for patients undergoing quiet breathing. Method and Materials: We have developed a 4D CT technique for mapping respiratory motion in radiotherapytreatment planning. A 16‐slice CT scanner was operated in cine mode to acquire 25 scans consecutively at each couch position. The scans were sorted into 12 respiratory‐windows based on the amplitude and direction (inhalation or exhalation), and on the phase (0–360°) of a synchronized external respiratory measurement. An air content measure (the amount of air in a 16‐slice CT segment, used as a surrogate for internal motion) was correlated to the respiratory amplitude and phase throughout the lung.Imagesreconstructed based on the two sorting techniques were displayed for a qualitative comparison. Also, the variations in the amplitude of the respiratory measurement during the entire scan session were compared using 8, 12, 24, and 48 respiratory windows. Results: The air content showed a higher correlation with the respiratory amplitude than with the respiratory phase for most cases. Imagesreconstructed based on the amplitude sorting technique displayed fewer artifacts, especially at the lung‐diaphragm boundaries, than imagesreconstructed based on the phase sorting technique. The variations in the respiratory amplitude were much smaller with amplitude sorting than those with phase sorting. These variations decreased significantly with finer amplitude respiratory windows while showed insignificant changes with finer phase respiratory windows. Conclusion: The amplitude sorting was generally better than phase sorting, especially for patients whose breathing was less reproducible. The use of finer respiratory windows did not improve the consistency for phase sorting. Keywords: 4‐D CT, respiratory sorting, motion, radiotherapy.

SU‐E‐J‐16: Noise Reduction with Detail Preservation for Low‐Dose KV CBCT Using Non‐Local Means: Simulated Patient Study

Purpose: With the frequent uses of CBCT in IGRT, the cumulative imaging dose to normal tissues may not be insignificant. A lower mAs protocol in CBCT acquisition reduces the dose, but dramatically degrades image quality due to excessive noise. The purpose of this study is to examine the effectiveness of the nonlocal means (NL‐means) denoising algorithm in reducing noise while preserving details in simulated low‐dose patient CBCT. Method and Materials:NL‐means algorithm estimates the true value of a pixel as a weighted average of all pixels in the image, where the weights depend on the similarity between the pixels. Compared with the local smoothing or filtering methods, NL‐means can reduce noise while preserve details. Low‐dose patient CBCTimages were generated by adding noise to the normal‐dose images based on stochastic property of incident photons from the x‐ray source, simulating data acquisition with reduced mAs. The low‐dose images were normalized to be within [0 1] and then fed to the NL‐means for denoising. The optimal parameters for NL‐means were experimentally determined. Results: The simulated low‐dose CBCT had only 6.3% of the patient CBCT dose. NL‐means clearly reduced the noise without obvious blurring, and the images appear to have similar quality as normal‐dose images. The suppressed noise resembled the desired white noise except at sharp edges. The mean signal‐to‐noise ratio in homogeneous regions was increased from 7.6 to 28.1. NL‐means preserved anatomic details as fine as 2 pixels but blurred single‐pixel details. Conclusion: Results on the patient data demonstrated that NL‐means effectively reduced the noise while preserving most fine details in simulated low‐dose patient CBCT, consistent with results on phantom study. A single set of optimal parameters was suitable for various CBCTs. This post‐processing method can be straightforwardly implemented in OBI software.

WE‐C‐BRD‐01: Automated Segmentation for Radiotherapy Volume Definition

One of the more exciting advances in radiation therapy is the potential for customizing therapy for each patient through the model of adaptive radiation therapy(ART). This includes obtaining 3‐dimensional functional and anatomic images throughout the course of therapy using both in‐room and conventional imaging platforms. Image acquisition is only the first step in monitoring the quality and efficacy of treatments. The images need to be segmented to allow dosimetric evaluations and comparisons with earlier image datasets. The comparisons will be conducted both to evaluate dose distribution delivery and radiation response and may require acquisition and analysis of multiple image datasets throughout therapy. One of the greatest time consuming aspects of radiation therapy treatment planning is the process of segmenting tumors and normal organs. The expectation that the manpower bandwidth will be available to significantly increase this workload is not realistic, so automated and validated technologies will be necessary to enable wide‐spread implementation of ART. One of the greatest challenges for ART will be to develop efficient and effective methods for reviewing the automated segmentation output. The current paradigm of evaluating structure contours on a slice‐by‐slice basis is too time consuming and does not take advantage of the natural anatomic characteristics of the structures being reviewed. A more efficient method for segmentation review will be required. Automated techniques will also be necessary to map tissue deformation that occurs due to normal day‐to‐day setup and internal organ variations as well as radiation response and disease progression or regression. This symposium will present the state‐of‐the art in automated segmentation, deformable image registration, and segmentation review technologies, technologies that will be critical to the effective implementation of ART.

TU-D-M100F-09: Breathing Motion-Induced Dose Delivery Error Evaluations as Applied to Tomotherapy Dose Delivery

Purpose: To develop a method for evaluating breathing motion‐induced dose delivery errors in Tomotherapy dose delivery. Methods and Materials: Dosimetric inaccuracy can result from breathing‐induced tumor motion in Tomotherapy treatment delivery. Patient breathing motion patterns were simulated using quantitative spirometry‐measured patient tidal volumes and converting the tidal volume to tumor motion by varying the ratio of tumor motion to tidal volume for 34 patients. Simulations of Tomotherapy deliveries were conducted modifying the previously published techniques by using measured beam profiles instead of step‐function fluences, and couch speeds typical of Tomotherapy treatments. Radiochromic film and our in‐house 4D phantom were used to verify the algorithm. Results: As expected, the breathing motion blurred the dose distributions, but slow drifts in the average tissue position caused detectable dose errors. The dose errors were expectedly largest with the smallest (1.0cm) field size, and could be >10% for motion amplitudes comparable to the field size. As the field width increased relative to the motion amplitude, and as couch‐speed (pitch) decreased, the error also decreased, and as such these settings may be preferable for patient treatments. These slow drifts occurred over time periods that were coincident with the amount of time required for the field to pass a stationary point. Measurements agreed with the simulation. Conclusions: Previous breathing motion studies did not use real patient breathing patterns and therefore did not consider the impact of slow, relatively small drifts in those patterns. The drifts change direction during the breathing measurement, causing dose errors that are both positive and negative. While the individual dose fraction errors can be >10%, they are unlikely to occur in the same place each day, so the average dose is likely to be consistent with earlier studies. Conflict of Interest: This work supported in part by a grant from Tomotherapy, Inc.

SU‐FF‐J‐04: A Computerized Method for Peak and Valley Detection in Respiratory Waveforms Without Flow Measurement

Purpose: To develop a computerized method for reliable peak (end inspiration) and valley (end expiration) detection in respiratory waveforms for gated radiotherapy and 4D CT. Method and Materials: A computerized method for peak (end inspiration) and valley (end expiration) detection in respiratory waveforms without flow measurement is described. The respiratory period T was estimated by applying a Fast Fourier Transform. The intercepts of the respiratory waveform with a moving average curve were determined with an averaging width of 2T. Peaks and valleys were defined respectively as the maximum and minimum between pairs of interwoven intercepts. While this method worked well the majority of the time, both automatic corrections and manual user interventions were employed to correct errors and adjust the results. Results: The method was implemented in MATLAB on a PC with a 3.0 GHz Pentium IV CPU and 2.0 GB RAM. On average, the respiratory waveform was 575.3 s long and contained a total of 307 peaks and valleys. For each patient, 99% of all peaks and valleys were correctly located by the automatic algorithm in 2.8 s. Only three (1%) points required manual user adjustment. A user spent 66.8 s for reviewing, and manually adding or deleting points. For nine of the 20 patients, all peaks and valleys were automatically detected. The high efficiency of the automatic algorithm is clear. Conclusion: The results demonstrated that this method was reliable and efficient for peak detection in respirator waveforms with noise and large variations in baseline level, amplitude and period.