4DCT Reconstruction Research Articles

SU-D-WAB-06: Sources of Motion Artifacts in 4DCT Reconstruction

Purpose: To identify sources of motion artifacts (MA) in 4-D CT reconstruction when patient breathing is irregular. Methods: A simulation study with a wire phantom was designed to quantify MA in phase-sorted 4D CT. Motion in the inferior-superior with cos4tɷt variation and 1 cm motion range was assumed. Motion frequency, expressed in breaths per minute (BPM) was assumed to be 15 for most of our cases, and always higher than 10 in order to exclude undersampling artifacts due to slow breathing and violate data sufficiency. The simulation included variations in motion period, amplitude and baseline drift; these variations were assumed to be affecting phantom and surrogate motion simultaneously. A varying phase shift between the phantom motion and the surrogate motion was also studied. The effect on MA derived from the variation of the above parameters was evaluated. Results: No noticeable MA was observed when the breathing period is changing during scanning. Variations of the amplitude and the base line drift play little role in generating MA. MA only becomes noticeable when the variation of the amplitude becomes larger than 30% of the motion range - or 3mm in our simulation. MA, however are sensitive to the variation in the phase difference between surrogate and phantom motion. Change of only 10 degrees in phase or 0.1 seconds shift during scanning generated significant MA. Conclusion: Change of phase between the breathing surrogate and the organ motion during scanning is the main source of the MA if patient breathing period is not too slow, i.e. if the data sufficiency condition is not violated. NIH R01 CA133539

EP-1237: The effect of Varian's normal breathing predictive filter in 4D-CT reconstruction.

EP-1237: The effect of Varian's normal breathing predictive filter in 4D-CT reconstruction.

Improving image‐guided radiation therapy of lung cancer by reconstructing 4D‐CT from a single free‐breathing 3D‐CT on the treatment day

One of the major challenges of lung cancer radiation therapy is how to reduce the margin of treatment field but also manage geometric uncertainty from respiratory motion. To this end, 4D-CT imaging has been widely used for treatment planning by providing a full range of respiratory motion for both tumor and normal structures. However, due to the considerable radiation dose and the limit of resource and time, typically only a free-breathing 3D-CT image is acquired on the treatment day for image-guided patient setup, which is often determined by the image fusion of the free-breathing treatment and planning day 3D-CT images. Since individual slices of two free breathing 3D-CTs are possibly acquired at different phases, two 3D-CTs often look different, which makes the image registration very challenging. This uncertainty of pretreatment patient setup requires a generous margin of radiation field in order to cover the tumor sufficiently during the treatment. In order to solve this problem, our main idea is to reconstruct the 4D-CT (with full range of tumor motion) from a single free-breathing 3D-CT acquired on the treatment day. We first build a super-resolution 4D-CT model from a low-resolution 4D-CT on the planning day, with the temporal correspondences also established across respiratory phases. Next, we propose a 4D-to-3D image registration method to warp the 4D-CT model to the treatment day 3D-CT while also accommodating the new motion detected on the treatment day 3D-CT. In this way, we can more precisely localize the moving tumor on the treatment day. Specifically, since the free-breathing 3D-CT is actually the mixed-phase image where different slices are often acquired at different respiratory phases, we first determine the optimal phase for each local image patch in the free-breathing 3D-CT to obtain a sequence of partial 3D-CT images (with incomplete image data at each phase) for the treatment day. Then we reconstruct a new 4D-CT for the treatment day by registering the 4D-CT of the planning day (with complete information) to the sequence of partial 3D-CT images of the treatment day, under the guidance of the 4D-CT model built on the planning day. We first evaluated the accuracy of our 4D-CT model on a set of lung 4D-CT images with manually labeled landmarks, where the maximum error in respiratory motion estimation can be reduced from 6.08 mm by diffeomorphic Demons to 3.67 mm by our method. Next, we evaluated our proposed 4D-CT reconstruction algorithm on both simulated and real free-breathing images. The reconstructed 4D-CT using our algorithm shows clinically acceptable accuracy and could be used to guide a more accurate patient setup than the conventional method. We have proposed a novel two-step method to reconstruct a new 4D-CT from a single free-breathing 3D-CT on the treatment day. Promising reconstruction results imply the possible application of this new algorithm in the image guided radiation therapy of lung cancer.

Respiratory motion sampling in 4DCT reconstruction for radiotherapy

Phase-based and amplitude-based sorting techniques are commonly used in four-dimensional CT (4DCT) reconstruction. However, effect of these sorting techniques on 4D dose calculation has not been explored. In this study, the authors investigated a candidate 4DCT sorting technique by comparing its 4D dose calculation accuracy with that for phase-based and amplitude-based sorting techniques. An optimization model was formed using organ motion probability density function (PDF) in the 4D dose convolution. The objective function for optimization was defined as the maximum difference between the expected 4D dose in organ of interest and the 4D dose calculated using a 4DCT sorted by a candidate sampling method. Sorting samples, as optimization variables, were selected on the respiratory motion PDF assessed during the CT scanning. Breathing curves obtained from patients' 4DCT scanning, as well as 3D dose distribution from treatment planning, were used in the study. Given the objective function, a residual error analysis was performed, and k-means clustering was found to be an effective sampling scheme to improve the 4D dose calculation accuracy and independent with the patient-specific dose distribution. Patient data analysis demonstrated that the k-means sampling was superior to the conventional phase-based and amplitude-based sorting and comparable to the optimal sampling results. For phase-based sorting, the residual error in 4D dose calculations may not be further reduced to an acceptable accuracy after a certain number of phases, while for amplitude-based sorting, k-means sampling, and the optimal sampling, the residual error in 4D dose calculations decreased rapidly as the number of 4DCT phases increased to 6. An innovative phase sorting method (k-means method) is presented in this study. The method is dependent only on tumor motion PDF. It could provide a way to refine the phase sorting in 4DCT reconstruction and is effective for 4D dose accumulation. Optimized sorting techniques could achieve acceptable residuals (less than 0.5% of the prescription dose) using 6 sorting samples, which is much better than amplitude-based or phase-based sorting. Further increase in sorting phase number exceeding 6 or more may not be necessary when using the k-means sampling or optimal sampling points.

Clinical evaluations of an amplitude-based binning algorithm for 4DCT reconstruction in radiation therapy

Phase-binning algorithms are commonly utilized in 4DCT image reconstruction for characterization of tumor or organ shape and respiration motion, but breathing irregularities occurring during 4DCT acquisition can cause considerable image distortions. Recently, amplitude-binning algorithms have been evaluated as a potential improvement to phase-binning algorithms for 4DCT image reconstruction. The purpose of this study was to evaluate the performance of the first commercially available on-line retrospective amplitude-binning algorithm for comparison to the traditional phase-binning algorithm. Both phantom and clinical data were used for evaluation. A phantom of known geometry was mounted on a 4D motion platform programmed with seven respiratory waves (two computer generated and five patient trajectories) and scanned with a Philips Brilliance Big bore 16-slice CT simulator. 4DCT images were reconstructed using commercial amplitude- and phase-binning algorithms. Image quality of the amplitude- and phase-binned image sets was compared by evaluation of shape and volume distortions in reconstructed images. Clinical evaluations were performed on 64 4DCT patient image sets in a blinded review process. The amplitude- and phase-binned 4DCT maximum intensity projection (MIP) images were further evaluated for 28 stereotactic body radiation therapy (SBRT) cases of total 64 cases. A preliminary investigation of the effects of respiratory amplitude and pattern irregularities on motion artifact severity was conducted. The phantom experiments illustrated that, as expected, maximum inhalation occurred at the 0% amplitude and maximum exhalation occurred at the 50% amplitude of the amplitude-binned 4DCT image sets. The phantom shape distortions were more severe in the images reconstructed from the phase-binning algorithm. In the clinical study, compared to the phase-binning algorithm, the amplitude-binning algorithm yielded fewer or less severe motion artifacts in 37.5% of the cases (24∕64), comparable artifacts in 54.7% of the cases (35∕64), and slightly greater artifacts in 7.8% of the cases (5∕64). Evaluation of SBRT cases demonstrated that the reconstructed tumor sizes and locations were comparable in 96% (1∕28) of the MIP image pairs generated from both amplitude- and phase-binning algorithms. In this case the amplitude-binned image set rendered a smaller tumor size, which was likely due to very shallow respiratory amplitudes occurring over several breathing cycles. Overall, the amplitude-binning algorithm for 4DCT reconstruction reduced the severity of tumor distortion and image artifacts compared to the phase-binning algorithm. However, the full range of motion may not be characterized using amplitude-binning algorithms. Despite superior performance, amplitude binning can still be susceptible to motion artifacts caused by large variations in amplitude of respiratory waves.

Improving image-guided radiation therapy of lung cancer by reconstructing 4D-CT from a single free-breathing 3D-CT on the treatment day: Treatment day 4D-CT reconstruction

Improving image-guided radiation therapy of lung cancer by reconstructing 4D-CT from a single free-breathing 3D-CT on the treatment day: Treatment day 4D-CT reconstruction

MO-A(SAM)-BRC-01: 4DCT: Methods and Artifacts

4DCT is routinely used in the treatment planning of moving targets. 4D scanning provides information essential for planning, specifically the size, shape and location of the target and organs at risk as a function of respiratory phase. 4D scanning also provides quantitative data on radiological pathlength needed for radiation dose calculations. The amount of scan data has increased by a factor of ∼10 over helical scans. Given the role of 4DCT in aperture design and dose calculation, it is important to analyze how accurately this scanning technique provides quantitative data and what the sources of uncertainty are. The learning objectives of this talk include: 1. Understand methods of 4DCT reconstruction commonly used today 2. Describe the residual artifacts in 4DCT due to irregularities in respiration 3. Describe advances in visualization that

SU‐E‐T‐287: The Clinical Evaluation of a Novel Amplitude‐Based Binning Algorithm for 4D CT Reconstruction

Purpose: To evaluate the performance of the first commercial amplitude‐binning algorithm for 4DCT reconstruction, to compare its reconstructed image quality to that of the phase‐binning algorithm, and to investigate possible clinical effects from changing algorithms. Methods: A department customized 4DCT motion phantom was utilized to evaluate the geometry accuracy (correct phase sequencing and motion) of the amplitude‐binning algorithm. 64 4DCT patient cases, in which 60(4) cases were acquired with 4D chest (abdomen) protocol, were collected and evaluated on the respiratory motion artifact severity. The MIP images generated from both algorithms for 10 SBRT cases were further evaluated by two radiation oncologists regarding the tumor sizes and locations. Results: The phantom experiments illustrated that, as expected, maximum inhalation occurred between 90%–10% phases, while maximum exhalation occurred between 40%–60% phases. The amplitude‐binning algorithm provides reasonable phase sequencing and motion within the 4D images. The amplitude‐binning algorithm yielded fewer respiratory motion artifacts on 24/64 cases (37.5%), comparable artifacts to the phase‐binning algorithm on 35/64 cases (54.7%), and slightly greater artifacts on 5/64 cases (7.8%). The 10 SBRT cases indicated that the reconstructed tumor sizes and locations were comparable on 9 MIP image pairs generated from amplitude‐binning and phase‐binning, respectively. However, amplitude‐binning yielded smaller tumor size on one case probably due to the very shallow respiratory amplitude occurring over several breathing cycles. Conclusions: Overall, amplitude‐binning algorithm yielded fewer or equivalent artifacts on 92.2% of cases (59/64 cases), and no dramatic difference on reconstructed tumor sizes and locations from the MIP images based on amplitude and phase binning algorithms. We suggest utilizing the amplitude‐binning 4DCT images for clinical practice, especially for SBRT cases. If very shallow respiratory waves or severe motion artifacts were found on specific cases, physicists should work with the physicians to provide the best achievable 4D images.

Low‐dose 4DCT reconstruction via temporal nonlocal means

Four-dimensional computed tomography (4DCT) has been widely used in cancer radiotherapy for accurate target delineation and motion measurement for tumors in the thorax and upper abdomen areas. However, its prolonged scanning duration causes a considerable increase of radiation dose compared to conventional CT, which is a major concern in its clinical application. This work is to develop a new algorithm to reconstruct 4DCT images from undersampled projections acquired at low mA s levels in order to reduce the imaging dose. Conventionally, each phase of 4DCT is reconstructed independently using the filtered backprojection (FBP) algorithm. The basic idea of the authors' new algorithm is that by utilizing the common information among different phases, the input information required to reconstruct the image of high quality, and thus the imaging dose, can be reduced. The authors proposed a temporal nonlocal means (TNLM) method to explore the interphase similarity. All phases of the 4DCT images are reconstructed simultaneously by minimizing a cost function consisting of a data fidelity term and a TNLM regularization term. The authors utilized a modified forward-backward splitting algorithm and a Gauss-Jacobi iteration method to efficiently solve the minimization problem. The algorithm was also implemented on a graphics processing unit (GPU) to improve the computational speed. The authors' reconstruction algorithm has been tested on a digital NCAT thorax phantom in three low dose scenarios: All projections with low mA s level, undersampled projections with high mA s level, and undersampled projections with low mA s level. In all three low dose scenarios, the new algorithm generates visually much better CT images containing less image noise and streaking artifacts compared to the standard FBP algorithm. Quantitative analysis shows that by comparing the authors' TNLM algorithm to the standard FBP algorithm, the contrast-to-noise ratio has been improved by a factor of 3.9-10.2 and the signal-to-noise ratio has been improved by a factor of 2.1-5.9, depending on the cases. In the situation of undersampled projection data, the majority of the streaks in the images reconstructed by FBP can be suppressed using the authors' algorithm. The total reconstruction time for all ten phases of a slice ranges from 40 to 90 s on an NVIDIA Tesla C1060 GPU card. The experimental results indicate that the authors' new algorithm outperforms the conventional FBP algorithm in effectively reducing the image artifacts due to undersampling and suppressing the image noise due to the low mA s level.

TU‐D‐204B‐04: 4DCT Reconstruction from Undersampled Projections Using Edge‐Preserving Total Variation and Non Local Means

Purpose: Four‐dimensional computed tomography (4DCT) is widely used in radiotherapy planning for thorax and upper abdomen to take respiratory motion into consideration. However, its high radiation dose becomes a major concern and impedes its wide application. In this work, we proposed an iterative 4DCT reconstruction algorithm with both an Edge‐Preserving Total Variation (EPTV) regularization and a Non‐Local Means (NLM) regularization to accurately reconstruct 4DCT images from highly undersampled projections to reduce imaging dose. Methods and Materials: In 4DCT, images of neighboring phases usually share similar structures, though these structures are apart from each other spatially in neighboring phases due to breathing motion. To take this temporal correlation into the reconstruction process, we propose to use NLM as a temporal regularization in 4DCT reconstruction. For this purpose, each phase of the 4DCT is iteratively reconstructed by a modified total variation method, namely EPTV method. NLM regularization is imposed in each iteration step to enhance image quality. Specifically, we consider every possible small patch in all phases. Patches spatially close to each other in successive phases are weighted averaged, with high weights assigned to similar patches. We tested our reconstruction algorithm on a NCAT phantom at thorax region and a real patient 4DCT thorax data. 4DCT images obtained from our EPTV+NLM approach are compared with those obtained from the EPTV method. Relative errors are computed and are used to evaluate the reconstruction quality. Results: About 40 x‐ray projections for each breathing phase are enough to accurately reconstruct the 4DCT image in both test cases. The images reconstructed with NLM regularization are found to have lower relative error and hence better image quality than those reconstructed otherwise. Conclusions: Our proposed EPTV+NLM method is a robust approach for the 4DCT reconstruction, where NLM as a temporal regularization can considerably enhance the reconstructed image quality.

WE‐C‐204B‐02: Disparity between 4DCT and CBCT Reconstruction of Internal Target Volumes

Purpose: Internal target volumes (ITVs) generated from different imaging modalities may not be identical. These differences could be due to inherent limitations in image reconstruction, and in this study we investigate the differences between 4DCT‐MIP generated ITVs and CBCT reconstructions of moving targets. Methods/Materials: A cork lung phantom with three embedded targets was attached to a programmable motion platform to simulate lung and tumor motion. Ideal sinusoidal motion and other simulated patterns were used to compare ITV dimensions reconstructed from 4DCT and CBCT. Four 4DCT and ten CBCT generated ITVs from four lung cancer patients were also compared using a physician's contour as ground truth. Results: CBCT targets along the direction of motion were consistently smaller than the 4DCT equivalents (by 0.49 +/− 0.50 cm) in the phantom measurements. The amplitude was varied during a single scan from 1 to 4 cm, and both 4DCT and CBCT under‐represented the ground truth dimension by 2.13 (+/−0.64) cm and 2.69 (+/−0.50) cm. Patient lung tumors did not reveal a similar pattern of smaller ITVs from CBCT. Lateral motion due to proximity to the heart resulted in lateral discrepancies for two patients, with CBCT ITV 0.32 cm and 0.68 cm larger than the 4DCT ITV. For the two other patients, the CBCT ITV was larger along the tumor's primary direction of motion. Conclusions: Even under ideal periodic motion, CBCT generated ITVs can deviate from their 4DCT counterparts. These variations are often assumed to be due to physiological changes, but may actually occur due to limitations in CBCT reconstruction. If irregular motion in actual patient cases is observed, the discrepancy between the 4DCT ITV and the CBCT counterpart can be greater than 5mm and adaptive re‐optimization based on CBCT ITV may not be beneficial.

Poster — Wed Eve—37: Phantom Study Comparing Image Quality of Slow‐CT and Average CT Dataset from 4DCT Radiotherapy Planning Acquisitions

The purpose of this work was to compare slow‐CT () and average CT () datasets with free‐breathing helical CT () for contouring of organs at risk (OAR) and radiation treatment planning in patients receiving stereotactic body radiation therapy in lung. A quantitative examination of image quality parameters obtained from 4DCT‐derived datasets was performed using a mobile image quality phantom. The rigid phantom was translated in the superior‐inferior and anterior‐posterior directions through the CT scanning plane. Measurements of noise, low and high‐contrast resolution, spatial linearity, sensitometry, and observation of image artifacts resulting from phantom motion were recorded. In addition to these measurements, we computed the mean CT image () by increasing the number of breathing phases in the average CT calculation, and compared those image quality parameters with , and . While the image quality parameters were statistically the same in the phantom study, datasets had fewer image artifacts than the datasets. The image quality of the datasets improved greatly when the number of datasets used to generate the CTN was larger than the maximum permissible bins on the 4DCT reconstruction software (10 phases). or datasets may be used in place of for OAR contouring and potentially dose calculations, without significant compromise in image quality. This work demonstrates that when 4DCT is available, it may not be necessary to acquire a separate scan for OAR contouring.

SU‐FF‐J‐39: The Impact of Improving Techniques for Phase Reassignment in 4D CT Reconstruction

Purpose: To develop the respiratory signal analysis program for a phase‐based retrospective 4D CT using RPM, spirometer, stereo‐camera and removed the irregular peaks of respiratory signals for accurate phase assignment in 4D CT. Methods: In this study, an in‐house respiratory signal analysis program was developed for the phase reassignment and the analysis of the irregular respiratory signals. Various irregular respiratory patterns were obtained from clinical experimental volunteers. After then, the in‐house program analyzed the factors affecting to phase assignment which is directly related to irradiate sector. The algorithms for separating individual respiratory signals are presented, which made use of baseline signals of input signal to identify the nodes between baseline signals and input signal. Therefore, we could achieve irregular peak reduction and noise elimination in phase assignment. Subsequently, accuracy of phase assignment was improved with removal of irregular signals by self‐developed algorithm. Furthermore, this program also has a function to detect period and amplitude automatically. Result: The standard deviation (SD) of breathing period representing the baseline method was shown stable compare with gradient method, but the difference was not statistically significant. Gradient method shows inaccurate phase assignment when all irregular peaks were detected and assigned. On the other hand, baseline method which was self‐developed algorithm confirmed accurate phase assignment with removal of error peak detection due to tussis. It was also verified superior peak detection with the suggested baseline method at respiratory signals which contained irregular period changes and drifts of amplitude. Conclusion: This study is considered to be useful for not only image reconstruction and elevation of irradiating accuracy through phase assignment of RPM system but also analysis of respiratory signals. Moreover, this respiratory analysis program can be applied to the 4D CT reconstruction using various respiratory sensors for the reduction of motion artifact.

TU-D-304A-05: An Improved Phase-Based 4DCT Reconstruction Using Local Breathing Variation Without RPM

Purpose: To develop a 4DCT sorting method based on the 2DCT breathing volume variation. Method and Materials: Current 4DCT techniques require an external surrogate (e.g., the Varian RPM system) to retrospectively sort 2DCT slices. Artifacts are often observed due to the imperfect correlation between the RPM and breathing motion. Our new method uses local breathing volume variations at each scan position as a surrogate. Six patients with severe 4DCT artifacts based on RPM sorting were retrospectively studied. These patients acquired 22–26 scans per couch position using a 16-slice CT scanner operated in Cine mode. For each couch position, the patient body contour was segmented using thresholding and region growing. The number of the pixels inside the body contour was counted as breathing volume, indicating the current breathing phase. For each scan position, standard respiratory phases were generated based on the nearest breathing volumes. 4DCT sorting was accomplished using CT images at the standard breathing phases. Results: For 6 patients studied, 4DCT images sorted from the RPM and our method were reviewed by experts. Most artifacts observed in RPM based 4DCT were removed by our method especially near the diaphragm and regions where the RPM signal was less correlated. For one patient, the RPM showed irregular breathing for one slice position resulting in severe artifacts in reconstructed 4DCT. However, local volume variation indicated regular breathing and the reconstructed 4DCT images were artifact free using our new method. For some scan positions where large irregular breathing occurred, our method, as well as the RPM method, did not work well and a rescan was recommended. Conclusion: The proposed sorting method provides a highly correlated individual respiratory signal for each scan position. The proposed method has the advantages of (1) significantly reduced 4DCT artifacts, (2) without RPM and (3) without complex organ segmentation.

SU‐FF‐T‐122: Investigation of Clinical Performance of Two 4DRT Techniques — RPM and Bellows System

Purpose The objective of this study was to systematically evaluate the clinical performance of two 4DRT techniques, RPM and Bellows system. Materials and Methods An RPM system was mounted at the distal end of a Philips CT simulator couch. A marker block was positioned on the sternum of a newly released commercial lung phantom. A Bellows sensor was tied around the thorax of the phantom next to the marker block to ensure equal breathing amplitudes. The air pressure generated by an air pump inflated and deflated the lung phantom periodically. It also drove a spherical target in a translational fashion with an 8° rotation at the end of the motion. The performance of the two systems was investigated by applying different motion depths, 0.3, 0.5, 0.75, and 1.0, and motion patterns, sin2(x), sin4(x), and sin6(x), to the lung phantom. The pressure wave was detected by the Bellows sensor and fed into the CT scanner for 4DCT reconstruction. Results The motion wave forms detected by RPM and Bellows system were almost identical in cases of sin2(x), sin4(x), and sin6(x) with a motion depth of 1. At large depths, both systems provided equal sensitivity in detecting breathing signals. At shallow breathing depths, the wave forms detected by RPM were degraded and the peaks were somewhat flattened, while the Bellows system preserved a relatively good wave form. At extremely shallow depths, RPM failed to track the breathing motion while the Bellows system continued to perform normally, providing a superior sensitivity over the RPM. Conclusions RPM failed to detect the breathing signal when the motion amplitude became very small. Bellows System was able to track the respiratory motion continuously even when it is extremely weak.

TU-D-304A-01: Development and Testing of a Novel, 4D Maximum A Posteriori (MAP) Image Reconstruction Algorithm

Purpose: Deformable image registration has been proven to be useful in tracking organ motion for dose calculation using artifact‐free 4D RCCT images. Such methods are challenged in the presence of image artifacts. We present an alternative method which avoids binning artifacts by directly estimating organ deformation during the reconstruction process. Method and Materials: We have developed a maximum a posteriori (MAP) algorithm for tracking organ motion that uses raw time‐stamped data to reconstruct the images and estimate deformations in anatomy simultaneously. Since the algorithm does not rely on a binning process, binning artifacts are avoided. Signal‐to‐noise ratio (SNR) is also increased since the algorithm uses all of the collected data. The increased SNR provides the opportunity to reduce dose to the patient during scanning. This framework also facilitates the incorporation of fundamental physical properties such as the conservation of local tissue volume during the estimation of organ motion. In order to validate the accuracy of the 4D reconstruction algorithm, a phantom study was performed using the CIRS anthropomorphic thorax phantom in a CT scanner. An improvement in image quality was also demonstrated by application of the algorithm to data from a real liver stereotactic body radiation therapy(SBRT) patient. Results: The algorithm accurately estimated the known motion of the anthropomorphic phantom. Additionally, a significant SNR increase was observed when using 4D reconstruction over binning, even for a scan with X‐ray tube current reduced to 10%. Conclusion: A novel method of fully 4D CTreconstruction was presented. The geometric accuracy of the estimated deformation was validated in phantom. A marked improvement in image quality was observed when applying the algorithm to image data from a real liverSBRT patient. The method allows reduction of X‐ray tube current during scanning while simultaneously improving motion estimates for use in dose calculation.

Reconstruction of 4D deformed CT for moving anatomy

Purpose To develop a 4DCT reconstruction technique that improves time resolution when the anatomy moves with respiration.

4D-CT Reconstruction Based on Accurate Vector Field Inter/Extrapolation

Four-dimensional computed tomography (4D-CT) generated using amplitude binning performs better and contains fewer artifacts, compared to those based on phase binning. However, missing amplitude values may appear during image acquisition using the external respiratory signal provided by a Real-time Position Management system (RPM) Here, we propose a method to overcome possible missing amplitude values based on accurate motion inter/extrapolation.

SU-GG-J-21: Accuracy in the Localization of Thoracic Tumors Using Respiratory Displacement, Velocity, and Phase

Purpose: Current 4D CT reconstruction techniques are retrospectively based on phase or displacement of the respiratory signal. Both have known limitations. The aim of this work was to evaluate the reproducibility of tumor position as function of displacement, phase, and velocity of the respiratory signal, respectively, in order to determine the optimum combination of parameters for real‐time 4D CTimage sorting. Method and Materials: Estimated tumor centroid position and respiratory signal data were acquired with the Cyberknife Synchrony system for 26 patients (52 fractions) thoracic radiotherapy patients. A reference respiratory cycle was calculated for each patient. This reference cycle was used to calculate displacement, phase, and velocity of 10 image bins. The tumor position was then sorted into these bins if the phase, displacement (inhale and exhale separately), simultaneous displacement and phase, or simultaneous displacement and velocity of the respiratory signal were within tolerances of 0.5 mm and 0.5 mm/s for displacement and velocity corresponding to the bins, respectively. For phase sorting, phase closest to the reference phase for each bin was chosen. Results: The mean of the standard deviations of tumor positions over all bins and all fractions for the superior‐inferior direction were: 1.65 ± 0.74 mm for phase sorting, 1.00 ± 0.59 mm for displacement sorting, 0.98 ± 0.52 mm for simultaneous displacement and phase sorting, and 0.91 ± 0.53 mm for simultaneous displacement and velocity sorting. The same trend was observed for the anterior‐posterior and left‐right directions. Conclusion: This study illustrates that position of a tumor can be determined more accurately if displacement and velocity are used simultaneously as sorting parameters. A real‐time displacement and velocity based 4D CTimage sorting method may therefore produce smaller artifacts in 4D CTimages than current retrospective sorting methods. Conflict‐of‐Interest: Research supported by 1P01CA116602.

Inter fractional variability of breathing phase definition as determined by fiducial location

Reconstruction of four-dimensional (4D) imaging typically requires an externally measurable surrogate to represent the real-time relative phase of respiration. A common method is to use a reflective marker on the external surface of the patient which moves with respiration and can be tracked in real time. The location of the marker is often chosen to maximize the observable motion, though this location may not be at the region of interest. We evaluate the importance of infrared (IR) marker placement location on breathing phase definition for the purpose of respiratory gating and 4D computed tomography (CT) image reconstruction. Data were collected for ten patients enrolled on an approved IRB protocol. Real-time position data were collected during CT imaging and daily treatments for two external IR reflective markers: one placed near the xyphoid and another at the approximate location of the treatment isocenter. Motion traces from the markers were compared using cross-correlation coefficient and by estimating the relative respiratory phase, based on either marker, as would be used for 4D-CT reconstruction. Cross-correlation analysis revealed differences in the motion waveform, as well as phase differences, both of which were variable between patients as well as day to day for the same patient. Estimated relative phases from each marker were compared by the percentage amount of time the estimated phase for each marker was different, binned based on increments of 10% of a full cycle. For all collected data combined, the frequency with which breathing phase mismatch led to different bin allocation in steps of 10% was as follows: T0%-10% = 65.1%, T10%-20% = 25.3%, T20%-30% = 7.8%, T30%-40% = 1.5% and T40%-50% = 0.4%. Based on ten images per cycle, this indicates that 4D reconstructions would be influenced, depending on which marker was used, by at least 1 bin 34.9% of the time. This number was noticeably higher for some patients; the maximum was 71% of the time for one patient of ten. In conclusion, the respiratory amplitude and relative phase depend significantly on the location of the IR marker used to monitor respiration. For some patients the xyphoid and isocentric markers may be completely out of phase. More importantly, this relationship varies day to day, suggesting that a single marker may be inadequate for the purposes of respiratory gating.