State-of-the-art reconstruction algorithms for medical helical cone-beam Computed Tomography (CT) are of type non-exact Filtered Backprojection (FBP). They are attractive because of their simplicity and low computational cost, but they produce sub-optimal images with respect to artifacts, resolution, and noise. This thesis deals with possibilities to improve the image quality by means of iterative techniques. The first algorithm, Regularized Iterative Weighted Filtered Backprojection (RIWFBP), is an iterative algorithm employing the non-exact Weighted FilteredBackprojection (WFBP) algorithm [Stierstorfer et al., Phys. Med. Biol. 49, 2209-2218, 2004] in the update step. We have measured and compared artifact reduction as well as resolution and noise properties for RIWFBP and WFBP. The results show that artifacts originating in the non-exactness of the WFBP algorithm are suppressed within five iterations without notable degradation in terms of resolution versus noise. Our experiments also indicate that the number of required iterations can be reduced by employing a technique known as ordered subsets. A small modification of RIWFBP leads to a new algorithm, the Weighted Least Squares Iterative Filtered Backprojection (WLS-IFBP). This algorithm has a slightly lower rate of convergence than RIWFBP, but in return it has the attractive property of converging to a solution of a certain least squares minimization problem. Hereby, theory and algorithms from optimization theory become applicable. Besides linear regularization, we have examined edge-preserving non-linear regularization.In this case, resolution becomes contrast dependent, a fact that can be utilized for improving high contrast resolution without degrading the signal-to-noise ratio in low contrast regions. Resolution measurements at different contrast levels and anthropomorphic phantom studies confirm this property. Furthermore, an even morepronounced suppression of artifacts is observed. Iterative reconstruction opens for more realistic modeling of the input data acquisition process than what is possible with FBP. We have examined the possibility to improve the forward projection model by (i) multiple ray models, and (ii) calculating strip integrals instead of line integrals. In both cases, for linearregularization, the experiments indicate a trade off: the resolution is improved atthe price of increased noise levels. With non-linear regularization on the other hand, the degraded signal-to-noise ratio in low contrast regions can be avoided. Huge input data sizes make experiments on real medical CT data very demanding. To alleviate this problem, we have implemented the most time consuming parts of the algorithms on a Graphics Processing Unit (GPU). These implementations are described in some detail, and some specific problems regarding parallelism and memory access are discussed.

<p>Onboard cone beam computed tomography (CBCT) has become a widespread means of three-dimensional target localization for radiation therapy; however, it is susceptible to metal artifacts and beam-hardening artifacts that can hinder visualization of low contrast anatomy. Dual-CBCT provides easy access to techniques that may reduces such artifacts. Additionally, dual-CBCT can decrease imaging time and provide simultaneous orthogonal projections which may also be useful for fast target localization. However, dual-CBCT will suffer from large increases in scattered radiation due to the addition of the second source.</p><p>An experimental bench top dual CBCT system was constructed so that each imaging chain in the dual CBCT system mimics the geometry of gantry-mounted CBCT systems commonly used in the radiation therapy room. The two systems share a common axis of rotation and are mounted orthogonally. Custom control software was developed to ensure reproducible exposure and rotation timings. This software allows the implementation of the acquisition sequences required for the cross scatter avoidance and correction strategies studied.</p><p>Utilizing the experimental dual CBCT system cross scatter was characterized from 70-145 kVp in projections and reconstructed images using this system and three cylindrical phantoms (15cm, 20cm, and 30cm) with a common Catphan core. A novel strategy for avoiding cross-scatter in dual-CBCT was developed that utilized interleaved data acquisition on each imaging chain. Contrast and contrast-to-noise-ratio were measured in reconstructions to evaluate the effectiveness of this strategy to avoid the effects of cross scatter.</p><p>A novel correction strategy for cross scatter was developed wherein the cross scatter was regularly sampled during the course of data acquisition and these samples were used as the basis for low- and high- frequency corrections for the cross-scatter in projections. The cross scatter sampling interval was determined for an anthropomorphic phantom at three different sites relevant to radiation therapy by estimating the angular Nyquist frequency. The low frequency portion of the cross scatter distribution is interpolated between samples to provide an estimate of the cross scatter distribution at every projection angle and was then subtracted from the projections.</p><p>The high-frequency portion of the correction was applied after the low-frequency correction was applied. The novel high-frequency correction utilizes the fact that a direct estimate of the high-frequency components was obtained in the cross scatter samples. The high-frequency components of the measured cross scatter were subtracted from the projections in the Fourier domain, a process referred to as spectral subtraction. Each projection is corrected using the cross scatter sample taken at the closest projection angle. In order to apply this correction in the Fourier domain the high-frequency component of the cross scatter must be approximately stationary. To improve the stationarity of the high-frequency cross scatter component a novel two-dimensional, overlapping window was developed. The spectral subtraction was then applied in each window and the results added to form the final image.</p><p>The effectiveness of the correction techniques were evaluated by measuring the contrast and contrast-to-noise-ratio in an image quality phantom. Additionally, the effect of the high-frequency correction on resolution was measured using a line pair phantom.</p><p>Cross scatter in dual CBCT was shown for large phantoms to be much higher than forward scatter which has long been known to be one of the largest degrading factors of image quality in CBCT. This results in large losses of contrast and CNR in reconstructed images. The interleaving strategy for avoiding cross scatter during projection acquisition showed similar performance to cross scatter free acquisitions, however, does not acquire projections at the maximum possible rate. For those applications in which maximizing the acquisition rate of projections is important, the low- and high-frequency corrections effectively mitigated the effects of cross scatter in the dual CBCT system.</p> / Dissertation

Interstitial and intracavitary brachytherapy plays an essential role in management of several malignancies. However, the achievable accuracy of brachytherapy treatment for prostate and cervical cancer is limited due to the lack of intraoperative planning and adaptive replanning. A major problem in implementing TRUS-based intraoperative planning is an inability of TRUS to accurately localize individual seed poses (positions and orientations) relative to the prostate volume during or after the implantation. For the locally advanced cervical cancer patient, manual drawing of the source positions on orthogonal films can not localize the full 3D intracavitary brachytherapy (ICB) applicator geometry. A new iterative forward projection matching (IFPM) algorithm can explicitly localize each individual seed/applicator by iteratively matching computed projections of the post-implant patient with the measured projections. This thesis describes adaptation and implementation of a novel IFPM algorithm that addresses hitherto unsolved problems in localization of brachytherapy seeds and applicators. The prototype implementation of 3-parameter point-seed IFPM algorithm was experimentally validated using a set of a few cone-beam CT (CBCT) projections of both the phantom and post-implant patient’s datasets. Geometric uncertainty due to gantry angle inaccuracy was incorporated. After this, IFPM algorithm was extended to 5-parameter elongated line-seed model which automatically reconstructs individual seed orientation as well as position. The accuracy of this algorithm was tested using both the synthetic-measured projections of clinically-realistic Model-6711 125I seed arrangements and measured projections of an in-house precision-machined prostate implant phantom that allows the orientations and locations of up to 100 seeds to be set to known values. The seed reconstruction error for simulation was less than 0.6 mm/3o. For the physical phantom experiments, IFPM absolute accuracy for position, polar angle, and azimuthal angel were (0.78 ± 0.57) mm, (5.8 ± 4.8)o, and (6.8 ± 4.0)o, respectively. It avoids the need to match corresponding seeds in each projection and accommodates incomplete data, overlapping seed clusters, and highly-migrated seeds. IFPM was further generalized from 5-parameter to 6-parameter model which was needed to reconstruct 3D pose of arbitrary-shape applicators. The voxelized 3D model of the applicator was obtained from external complex combinatorial geometric modeling. It is then integrated into the forward projection matching method for computing the 2D projections of the 3D ICB applicators, iteratively. The applicator reconstruction error for simulation was about 0.5 mm/2o. The residual 2D registration error (positional difference) between computed and actual measured applicator images was less than 1 mm for the intrauterine tandem and about 1.5 mm for the bilateral colpostats in each detector plane. By localizing the applicator’s internal structure and the sources, the effect of intra and inter-applicator attenuation can be included in the resultant dose distribution and CBCT metal streaking artifact mitigation. The localization accuracy of better than 1 mm and 6o has the potential to support more accurate Monte Carlo-based or 2D TG-43 dose calculations in clinical practice. It is hoped the clinical implementation of IFPM approach to localize elongated line-seed/applicator for intraoperative brachytherapy planning may have a positive impact on the treatment of prostate and cervical cancers.

A utilização da tomografia computadorizada de feixe cônico (TCFC) permite avaliar a altura e espessura do osso remanescente para a instalação de implantes. As mandíbulas atrésicas necessitam de maior grau de detalhamento para obtenção de medidas. A autora, no presente estudo, avaliou a influência do tamanho dos voxels (0,125mm e 0,25mm) nas medidas de altura e espessura em 10 mandíbulas desdentadas e atrésicas. Para determinação dos cortes tomográficos no estudo foram utilizados fios metáticos nas regiões posteriores. Após o exame as mandíbulas foram seccionadas e as medidas diretas de altura e espessura foram obtidas pelo uso de paquímetro digital. Os resultados pelo teste ICC demonstraram não haver diferenças estatisticamente significantes entre as medidas realizadas pelos dois examinadores com paquímetro e as tomografias. Concluiu-se que a TCFC é confiável para realização de medidas na mandíbula e que o tamanho do voxel não interferiu no planejamento cirúrgico. / The use of cone beam computed tomography (CBCT) allows to evaluate the height and thickness of the remaining bone for implant placement. The atresic mandible requires higher level of detail for obtaining measurements. The author in the present study evaluates the influence of the voxel size (0.125 mm and 0.25 mm) in thickness and height measurements using 10 edentulous dry mandibles. For determination of the tomographic study were used metallic wires in the posterior regions. After examining, the mandibles were sectioned and direct measures of height and thickness were obtained through digital caliper. The results by ICC test showed no statistically significant differences between the measurements with caliper and tomograms made by two examiners. It was concluded that CBCT is trusted to carry out measurements in mandible and the voxel size did not affect the surgical planning.

A utilização da tomografia computadorizada de feixe cônico (TCFC) permite avaliar a altura e espessura do osso remanescente para a instalação de implantes. As mandíbulas atrésicas necessitam de maior grau de detalhamento para obtenção de medidas. A autora, no presente estudo, avaliou a influência do tamanho dos voxels (0,125mm e 0,25mm) nas medidas de altura e espessura em 10 mandíbulas desdentadas e atrésicas. Para determinação dos cortes tomográficos no estudo foram utilizados fios metáticos nas regiões posteriores. Após o exame as mandíbulas foram seccionadas e as medidas diretas de altura e espessura foram obtidas pelo uso de paquímetro digital. Os resultados pelo teste ICC demonstraram não haver diferenças estatisticamente significantes entre as medidas realizadas pelos dois examinadores com paquímetro e as tomografias. Concluiu-se que a TCFC é confiável para realização de medidas na mandíbula e que o tamanho do voxel não interferiu no planejamento cirúrgico. / The use of cone beam computed tomography (CBCT) allows to evaluate the height and thickness of the remaining bone for implant placement. The atresic mandible requires higher level of detail for obtaining measurements. The author in the present study evaluates the influence of the voxel size (0.125 mm and 0.25 mm) in thickness and height measurements using 10 edentulous dry mandibles. For determination of the tomographic study were used metallic wires in the posterior regions. After examining, the mandibles were sectioned and direct measures of height and thickness were obtained through digital caliper. The results by ICC test showed no statistically significant differences between the measurements with caliper and tomograms made by two examiners. It was concluded that CBCT is trusted to carry out measurements in mandible and the voxel size did not affect the surgical planning.

<p>In this project several tests were performed to evaluate the performance of an On-Board Imager® (OBI) mounted on a clinical linear accelerator. The measurements were divided into three parts; geometric accuracy, image registration and couch shift accuracy, and image quality. A cube phantom containing a radiation opaque marker was used to study the agreement with treatment isocenter for both kV-images and cone-beam CT (CBCT) images. The long term stability was investigated by acquiring frontal and lateral kV images twice a week over a 3 month period. Stability in vertical and longitudinal robotic arm motion as well as the stability of the center-of-rotation was evaluated. Further, the agreement of kV image and CBCT center with MV image center was examined.</p><p>A marker seed phantom was used to evaluate and compare the three applications in image registration; 2D/2D, 2D/3D and 3D/3D. Image registration using kV-kV image sets were compared with MV MV and MV-kV image sets. Further, the accuracy in 2D/2D matches with images acquired at non-orthogonal gantry angles was evaluated. The image quality in CBCT images was evaluated using a Catphan® phantom. Hounsfield unit (HU) uniformity and linearity was compared with planning CT. HU accuracy is crucial for dose verification using CBCT data.</p><p>The geometric measurements showed good long term stability and accurate position reproducibility after robotic arm motions. A systematic error of about 1 mm in lateral direction of the kV-image center was detected. A small difference between kV and CBCT center was observed and related to a lateral kV detector offset. The vector disagreement between kV- and MV-image centers was  2 mm at some gantry angles. Image registration with the different match applications worked sufficiently. 2D/3D match was seen to correct more accurately than 2D/2D match for large translational and rotational shifts. CBCT images acquired with full-fan mode showed good HU uniformity but half fan images were less uniform. In the soft tissue region the HU agreement with planning CT was reasonable while a larger disagreement was observed at higher densities. This work shows that the OBI is robust and stable in its performance. With regular QC and calibrations the geometric precision of the OBI can be maintained within 1 mm of treatment isocenter.</p>

In this project several tests were performed to evaluate the performance of an On-Board Imager® (OBI) mounted on a clinical linear accelerator. The measurements were divided into three parts; geometric accuracy, image registration and couch shift accuracy, and image quality. A cube phantom containing a radiation opaque marker was used to study the agreement with treatment isocenter for both kV-images and cone-beam CT (CBCT) images. The long term stability was investigated by acquiring frontal and lateral kV images twice a week over a 3 month period. Stability in vertical and longitudinal robotic arm motion as well as the stability of the center-of-rotation was evaluated. Further, the agreement of kV image and CBCT center with MV image center was examined. A marker seed phantom was used to evaluate and compare the three applications in image registration; 2D/2D, 2D/3D and 3D/3D. Image registration using kV-kV image sets were compared with MV MV and MV-kV image sets. Further, the accuracy in 2D/2D matches with images acquired at non-orthogonal gantry angles was evaluated. The image quality in CBCT images was evaluated using a Catphan® phantom. Hounsfield unit (HU) uniformity and linearity was compared with planning CT. HU accuracy is crucial for dose verification using CBCT data. The geometric measurements showed good long term stability and accurate position reproducibility after robotic arm motions. A systematic error of about 1 mm in lateral direction of the kV-image center was detected. A small difference between kV and CBCT center was observed and related to a lateral kV detector offset. The vector disagreement between kV- and MV-image centers was  2 mm at some gantry angles. Image registration with the different match applications worked sufficiently. 2D/3D match was seen to correct more accurately than 2D/2D match for large translational and rotational shifts. CBCT images acquired with full-fan mode showed good HU uniformity but half fan images were less uniform. In the soft tissue region the HU agreement with planning CT was reasonable while a larger disagreement was observed at higher densities. This work shows that the OBI is robust and stable in its performance. With regular QC and calibrations the geometric precision of the OBI can be maintained within 1 mm of treatment isocenter.

No description available.

Modern image-guided surgery aids minimally-invasive, high-precision procedures that increase efficacy of treatment. This thesis investigates two research aims to improve precision and integration of intraoperative cone-beam CT (CBCT) imaging in guidance of head and neck (H&N) surgery. First, marker configurations were examined to identify arrangements that minimize target registration error (TRE). Best arrangements minimized the distance between the configuration centroid and surgical target while maximizing marker separation. Configurations of few markers could minimized TRE with more markers providing improved uniformity. Second, an algorithm for automatic registration of image and world reference frames was pursued to streamline integration of CBCT with real-time tracking and provide automatic updates per scan. Markers visible to the tracking and imaging systems are automatically co-localized and registered with equivalent accuracy and superior reproducibility compared to conventional registration. Such work helps the implementation of CBCT in H&N surgery to maximize surgical precision and exploit intraoperative image guidance.

Modern image-guided surgery aids minimally-invasive, high-precision procedures that increase efficacy of treatment. This thesis investigates two research aims to improve precision and integration of intraoperative cone-beam CT (CBCT) imaging in guidance of head and neck (H&N) surgery. First, marker configurations were examined to identify arrangements that minimize target registration error (TRE). Best arrangements minimized the distance between the configuration centroid and surgical target while maximizing marker separation. Configurations of few markers could minimized TRE with more markers providing improved uniformity. Second, an algorithm for automatic registration of image and world reference frames was pursued to streamline integration of CBCT with real-time tracking and provide automatic updates per scan. Markers visible to the tracking and imaging systems are automatically co-localized and registered with equivalent accuracy and superior reproducibility compared to conventional registration. Such work helps the implementation of CBCT in H&N surgery to maximize surgical precision and exploit intraoperative image guidance.