Alen Docef

Predicting respiratory motion for four-dimensional radiotherapy

Adapting radiation delivery to respiratory motion is made possible through corrective action based on real-time feedback of target position during respiration. The advantage of this approach lies with its ability to allow tighter margins around the target while simultaneously following its motion. A significant hurdle to the successful implementation of real-time target-tracking-based radiation delivery is the existence of a finite time delay between the acquisition of target position and the mechanical response of the system to the change in position. Target motion during the time delay leads to a resultant lag in the systems response to a change in tumor position. Predicting target position in advance is one approach to ensure accurate delivery. The aim of this manuscript is to estimate the predictive ability of sinusoidal and adaptive filter-based prediction algorithms on multiple sessions of patient respiratory patterns. Respiratory motion information was obtained from recordings of diaphragm motion for five patients over 60 sessions. A prediction algorithm that employed both prediction models-the sinusoidal model and the adaptive filter model-was developed to estimate prediction accuracy over all the sessions. For each session, prediction error was computed for several time instants (response time) in the future (0-1.8 seconds at 0.2-second intervals), based on position data collected over several signal-history lengths (1-7 seconds at 1-second intervals). Based on patient data included in this study, the following observations are made. Qualitative comparison of predicted and actual position indicated a progressive increase in prediction error with an increase in response time. A signal-history length of 5 seconds was found to be the optimal signal history length for prediction using the sinusoidal model for all breathing training modalities. In terms of overall error in predicting respiratory motion, the adaptive filter model performed better than the sinusoidal model. With the adaptive filter, average prediction errors of less than 0.2 cm (1sigma) are possible for response times less than 0.4 seconds. In comparing prediction error with system latency error (no prediction), the adaptive filter model exhibited lesser prediction errors as compared to the sinusoidal model, especially for longer response time values (>0.4 seconds). At smaller response time values (<0.4 seconds), improvements in prediction error reduction are required for both predictive models in order to maximize gains in position accuracy due to prediction. Respiratory motion patterns are inherently complex in nature. While linear prediction-based prediction models perform satisfactorily for shorter response times, their prediction accuracy significantly deteriorates for longer response times. Successful implementation of real-time target-tracking-based radiotherapy requires response times less than 0.4 seconds or improved prediction algorithms.

Dosimetric impact of geometric errors due to respiratory motion prediction on dynamic multileaf collimator-based four-dimensional radiation delivery.

The synchronization of dynamic multileaf collimator (DMLC) response with respiratory motion is critical to ensure the accuracy of DMLC-based four dimensional (4D) radiation delivery. In practice, however, a finite time delay (response time) between the acquisition of tumor position and multileaf collimator response necessitates predictive models of respiratory tumor motion to synchronize radiation delivery. Predicting a complex process such as respiratory motion introduces geometric errors, which have been reported in several publications. However, the dosimetric effect of such errors on 4D radiation delivery has not yet been investigated. Thus, our aim in this work was to quantify the dosimetric effects of geometric error due to prediction under several different conditions. Conformal and intensity modulated radiation therapy (IMRT) plans for a lung patient were generated for anterior-posterior/posterior-anterior (AP/PA) beam arrangements at 6 and 18 MV energies to provide planned dose distributions. Respiratory motion data was obtained from 60 diaphragm-motion fluoroscopy recordings from five patients. A linear adaptive filter was employed to predict the tumor position. The geometric error of prediction was defined as the absolute difference between predicted and actual positions at each diaphragm position. Distributions of geometric error of prediction were obtained for all of the respiratory motion data. Planned dose distributions were then convolved with distributions for the geometric error of prediction to obtain convolved dose distributions. The dosimetric effect of such geometric errors was determined as a function of several variables: response time (0-0.6 s), beam energy (6∕18MV), treatment delivery (3D∕4D), treatment type (conformal/IMRT), beam direction (AP/PA), and breathing training type (free breathing/audio instruction/visual feedback). Dose difference and distance-to-agreement analysis was employed to quantify results. Based on our data, the dosimetric impact of prediction (a) increased with response time, (b) was larger for 3D radiation therapy as compared with 4D radiation therapy, (c) was relatively insensitive to change in beam energy and beam direction, (d) was greater for IMRT distributions as compared with conformal distributions, (e) was smaller than the dosimetric impact of latency, and (f) was greatest for respiration motion with audio instructions, followed by visual feedback and free breathing. Geometric errors of prediction that occur during 4D radiation delivery introduce dosimetric errors that are dependent on several factors, such as response time, treatment-delivery type, and beam energy. Even for relatively small response times of 0.6 s into the future, dosimetric errors due to prediction could approach delivery errors when respiratory motion is not accounted for at all. To reduce the dosimetric impact, better predictive models and/or shorter response times are required.

4D Cone-beam CT reconstruction using a motion model based on principal component analysis.

PURPOSE To provide a proof of concept validation of a novel 4D cone-beam CT (4DCBCT) reconstruction algorithm and to determine the best methods to train and optimize the algorithm. METHODS The algorithm animates a patient fan-beam CT (FBCT) with a patient specific parametric motion model in order to generate a time series of deformed CTs (the reconstructed 4DCBCT) that track the motion of the patient anatomy on a voxel by voxel scale. The motion model is constrained by requiring that projections cast through the deformed CT time series match the projections of the raw patient 4DCBCT. The motion model uses a basis of eigenvectors that are generated via principal component analysis (PCA) of a training set of displacement vector fields (DVFs) that approximate patient motion. The eigenvectors are weighted by a parameterized function of the patient breathing trace recorded during 4DCBCT. The algorithm is demonstrated and tested via numerical simulation. RESULTS The algorithm is shown to produce accurate reconstruction results for the most complicated simulated motion, in which voxels move with a pseudo-periodic pattern and relative phase shifts exist between voxels. The tests show that principal component eigenvectors trained on DVFs from a novel 2D/3D registration method give substantially better results than eigenvectors trained on DVFs obtained by conventionally registering 4DCBCT phases reconstructed via filtered backprojection. CONCLUSIONS Proof of concept testing has validated the 4DCBCT reconstruction approach for the types of simulated data considered. In addition, the authors found the 2D/3D registration approach to be our best choice for generating the DVF training set, and the Nelder-Mead simplex algorithm the most robust optimization routine.

Reconstruction of a cone-beam CT image via forward iterative projection matching

Purpose: To demonstrate the feasibility of reconstructing a cone-beam CT (CBCT) image by deformably altering a prior fan-beam CT (FBCT) image such that it matches the anatomy portrayed in the CBCT projection data set. Methods: A prior FBCT image of the patient is assumed to be available as a source image. A CBCT projection data set is obtained and used as a target image set. A parametrized deformation model is applied to the source FBCT image, digitally reconstructed radiographs (DRRs) that emulate the CBCT projection image geometry are calculated and compared to the target CBCT projection data, and the deformation model parameters are adjusted iteratively until the DRRs optimally match the CBCT projection data set. The resulting deformed FBCT image is hypothesized to be an accurate representation of the patient’s anatomy imaged by the CBCT system. The process is demonstrated via numerical simulation. A known deformation is applied to a prior FBCT image and used to create a synthetic set of CBCT target projections. The iterative projection matching process is then applied to reconstruct the deformation represented in the synthetic target projections; the reconstructed deformation is then compared to the known deformation. The sensitivity of the process to the number of projections and the DRR∕CBCT projection mismatch is explored by systematically adding noise to and perturbing the contrast of the target projections relative to the iterated source DRRs and by reducing the number of projections. Results: When there is no noise or contrast mismatch in the CBCT projection images, a set of 64 projections allows the known deformed CT image to be reconstructed to within a nRMS error of 1% and the known deformation to within a nRMS error of 7%. A CT image nRMS error of less than 4% is maintained at noise levels up to 3% of the mean projection intensity, at which the deformation error is 13%. At 1% noise level, the number of projections can be reduced to 8 while maintaining CT image and deformation errors of less than 4% and 13%, respectively. The method is sensitive to contrast mismatch between the simulated projections and the target projections when the soft-tissue contrast in the projections is low. Conclusions: By using prior knowledge available in a FBCT image, the authors show that a CBCT image can be iteratively reconstructed from a comparatively small number of projection images, thus saving acquisition time and reducing imaging dose. This will enable more frequent daily imaging during radiation therapy. Because the process preserves the CT numbers of the FBCT image, the resulting 3D image intensities will be more accurate than a CBCT image reconstructed via conventional backprojection methods. Reconstruction errors are insensitive to noise at levels beyond what would typically be found in CBCT projection data, but are sensitive to contrast mismatch errors between the CBCT projection data and the DRRs.

Performance Analysis of the JPEG 2000 Image Coding Standard

Some of the major objectives of the JPEG 2000 still image coding standard were compression and memory efficiency, lossy to lossless coding, support for continuous-tone to bi-level images, error resilience, and random access to regions of interest. This paper will provide readers with some insight on various features and functionalities supported by a baseline JPEG 2000-compliant codec. Three JPEG 2000 software implementations (Kakadu, JasPer, JJ2000) are compared with several other codecs, including JPEG, JBIG, JPEG-LS, MPEG-4 VTC and H.264 intra coding. This study can serve as a guideline for users to estimate the effectiveness of JPEG 2000 for various applications, and to select optimal parameters according to specific application requirements.

Reconstruction of 4D deformed CT for moving anatomy

PurposeTo develop a 4DCT reconstruction technique that improves time resolution when the anatomy moves with respiration.MethodA cone-beam CT (CBCT) scan is performed and a breathing signal is acquired. At the same time a set of simulated CBCT projections is created from a prior source CT deformed by a time-dependent parametric deformation model. The model parameters are iteratively adjusted until the simulated projections optimally resemble the acquired images. This concept was tested with three different simulated deformation scenarios approximating a moving lung tumor with rigid and elastic deformation and a 3D anatomical expansion/contraction.ResultsThe known deformation was accurately reconstructed in all three scenarios. The method is reasonably robust to noise and contrast mismatch in the projectionsConclusionMatching simulated to actual CBCT projections can adequately constrain a 4D model of breathing-induced motion that occurs during acquisition of the CBCT data.

On-line versus off-line accelerated kernel feature analysis: Application to computer-aided detection of polyps in CT colonography

A semi-supervised learning method, the on-line accelerated kernel feature analysis (On-line AKFA) is presented. In On-line AKFA, features are extracted while data are being fed to the algorithm in small batches as the algorithm proceeds. The paper compares and contrasts the use of On-line AKFA and Off-line AKFA in CT colonography. On-line AKFA provides the flexibility to allow the feature space to dynamically adjust to changes in the input data with time during the training phase. The computational time, reconstruction accuracy, projection variance, and classification performance of the proposed method are experimentally evaluated for kernel principal component analysis (KPCA), Off-line AKFA, and On-line AKFA. Experimental results demonstrate a significant reduction in computation time for On-line AKFA compared to the other feature extraction methods considered in this paper.

Motion-based projection generation for 4D-CT reconstruction

A method for reducing streaking artifacts in 4D-CT reconstruction by generating additional projections is proposed. This method uses optical flow to track anatomy motion across the complete set of projections and then uses this information to compute interpolated projections while compensating for breathing motion. Original and interpolated projections, all belonging to one respiratory phase, are used to reconstruct a 4D-CT volume. Experimental results showed that the proposed method reduces artifacts and blurring in reconstructed 4D-CT volumes and improves the image quality.

TH-C-BRC-05: 4DCBCT Reconstruction Using a PCA Based DVF Motion Model

Purpose: To develop a 4D cone beam CT (4DCBCT) reconstruction algorithm that produces a time series of images with high spatial and temporal resolution, minimal artifact, and no time averaging of data as compared to standard 4DCBCT reconstruction techniques. Methods: Our algorithm works to constrain a motion model by using it to deform a static prior CTimage at multiple time steps. It then adjusts the motion model parameters until digitally reconstructed radiographs (DRRs) taken through the deformed prior optimally match the projections of the raw 4DCBCT data‐set. Once the motion model parameters have been optimized the motion model can be used to animate the static CT, in essence producing a 3D video that serves as the reconstructed 4DCBCT. Our motion model uses principal components analysis (PCA) of a DVF training set to generate eigenvectors characterizing de‐correlated modes of anatomic motion. It produces new DVFs by computing weighted sums of the eigenvectors. Each eigenvector weight, or principal coefficient, varies in time uniquely according to a parameterized version of the patient breathing trace. Results: Our current focus is on proof of concept testing using numerically generated data. We plan to test the algorithm on simulated data with different degrees of motion complexity. In addition, we will test the effect on eigenvector quality of different DVF training set acquisition methods. Early results indicate that when supplied with an ideal DVF training set the model can reproduce simple motion to a high degree of accuracy. Conclusions: Our current task is to test our algorithms reconstruction quality with numerically simulated data of varying complexity. Beyond that we plan to test on phantom and patient data. If our reconstruction algorithm is successful it will produce 4DCBCTs with high spatial and temporal resolution, low artifact, and no time averaging. This work supported by NIH P01 CA116602.

SU‐GG‐I‐46: Reconstruction of a Cone‐Beam CT Image Via Forward Iterative Projection Matching

PURPOSE To demonstrate the feasibility of reconstructing a cone-beam CT (CBCT) image by deformably altering a prior fan-beam CT (FBCT) image such that it matches the anatomy portrayed in the CBCT projection data set. METHODS A prior FBCT image of the patient is assumed to be available as a source image. A CBCT projection data set is obtained and used as a target image set. A parametrized deformation model is applied to the source FBCT image, digitally reconstructed radiographs (DRRs) that emulate the CBCT projection image geometry are calculated and compared to the target CBCT projection data, and the deformation model parameters are adjusted iteratively until the DRRs optimally match the CBCT projection data set. The resulting deformed FBCT image is hypothesized to be an accurate representation of the patients anatomy imaged by the CBCT system. The process is demonstrated via numerical simulation. A known deformation is applied to a prior FBCT image and used to create a synthetic set of CBCT target projections. The iterative projection matching process is then applied to reconstruct the deformation represented in the synthetic target projections; the reconstructed deformation is then compared to the known deformation. The sensitivity of the process to the number of projections and the DRR/CBCT projection mismatch is explored by systematically adding noise to and perturbing the contrast of the target projections relative to the iterated source DRRs and by reducing the number of projections. RESULTS When there is no noise or contrast mismatch in the CBCT projection images, a set of 64 projections allows the known deformed CT image to be reconstructed to within a nRMS error of 1% and the known deformation to within a nRMS error of 7%. A CT image nRMS error of less than 4% is maintained at noise levels up to 3% of the mean projection intensity, at which the deformation error is 13%. At 1% noise level, the number of projections can be reduced to 8 while maintaining CT image and deformation errors of less than 4% and 13%, respectively. The method is sensitive to contrast mismatch between the simulated projections and the target projections when the soft-tissue contrast in the projections is low. CONCLUSIONS By using prior knowledge available in a FBCT image, the authors show that a CBCT image can be iteratively reconstructed from a comparatively small number of projection images, thus saving acquisition time and reducing imaging dose. This will enable more frequent daily imaging during radiation therapy. Because the process preserves the CT numbers of the FBCT image, the resulting 3D image intensities will be more accurate than a CBCT image reconstructed via conventional backprojection methods. Reconstruction errors are insensitive to noise at levels beyond what would typically be found in CBCT projection data, but are sensitive to contrast mismatch errors between the CBCT projection data and the DRRs.