Accuracy of model-based tracking of knee kinematics and cartilage contact measured by dynamic volumetric MRI

The purpose of this study was to determine the accuracy of knee kinematics and cartilage contact measured by volumetric dynamic MRI. A motor-actuated phantom drove femoral and tibial bone segments through cyclic 3D motion patterns. Volumetric images were continuously acquired using a 3D radially undersampled cine spoiled gradient echo sequence (SPGR-VIPR). Image data was binned based on position measured via a MRI-compatible rotary encoder. High-resolution static images were segmented to create bone models. Model-based tracking was performed by optimally registering the bone models to the volumetric images at each frame of the SPGR-VIPR series. 3D tibiofemoral translations and orientations were reconstructed, and compared to kinematics obtained by tracking fiducial markers. Imaging was repeated on a healthy subject who performed cyclic knee flexion-extension. Cartilage contact for the subject was assessed by measuring the overlap between articular cartilage surfaces. Model-based tracking was able to track tibiofemoral angles and translations with precisions less than 0.8° and 0.5 mm. These precisions resulted in an uncertainty of less than 0.5 mm in cartilage contact location. Dynamic SPGR-VIPR imaging can accurately assess in vivo knee kinematics and cartilage contact during voluntary knee motion performed in a MRI scanner. This technology could facilitate the quantitative investigation of links between joint mechanics and the development of osteoarthritis.     

Precision assessment of model-based RSA for a total knee prosthesis in a biplanar set-up

Model-based Roentgen Stereophotogrammetric Analysis (RSA) was recently developed for the measurement of prosthesis micromotion. Its main advantage is that markers do not need to be attached to the implants as traditional marker-based RSA requires. Model-based RSA has only been tested in uniplanar radiographic set-ups. A biplanar set-up would theoretically facilitate the pose estimation algorithm, since radiographic projections would show more different shape features of the implants than in uniplanar images. We tested the precision of model-based RSA and compared it with that of the traditional marker-based method in a biplanar set-up. Micromotions of both tibial and femoral components were measured with both the techniques from double examinations of patients participating in a clinical study. The results showed that in the biplanar set-up model-based RSA presents a homogeneous distribution of precision for all the translation directions, but an inhomogenous error for rotations, especially internal-external rotation presented higher errors than rotations about the transverse and sagittal axes. Model-based RSA was less precise than the marker-based method, although the differences were not significant for the translations and rotations of the tibial component, with the exception of the internal-external rotations. For both prosthesis components the precisions of model-based RSA were below 0.2 mm for all the translations, and below 0.3 degrees for rotations about transverse and sagittal axes. These values are still acceptable for clinical studies aimed at evaluating total knee prosthesis micromotion. In a biplanar set-up model-based RSA is a valid alternative to traditional marker-based RSA where marking of the prosthesis is an enormous disadvantage.     

Error performances of a model-based biplane fluoroscopic system for tracking knee prosthesis during treadmill gait task

Roentgen stereophotogrammetry analysis technique allows an accurate measurement of knee joint prosthesis position and orientation using two X-ray images. Although this technique is used generally during static procedure, it is possible to use it with a biplane fluoroscopic system to measure the prosthesis kinematics during functional tasks (e.g., gait, squat, jump) performed in a laboratory environment. However, the performance of the system in terms of errors for the measurements and the model-based matching algorithm are not well known for dynamic tasks such as walking. The goal of this study was to estimate the static and dynamic errors of a model-based biplane fluoroscopic system for a treadmill gait task and analyze the error performance according to the speed and location of the knee joint prosthesis relative to X-ray sources. The results show a static maximum error (RMSE) of 0.13° for orientation and 0.06 mm for position for prosthesis components. The dynamic errors were different for each axis of the acquisition system and each prosthesis component. The largest dynamic error was along the vertical axis for the position (RMSE = 2.42 mm) and along the medio-lateral axis (perpendicular to movement) for the orientation (RMSE = 0.95°). As expected, the error depends on the distance between the prosthesis and the source in the acquisition system as well as the linear and angular velocity of the movement. The most accurate dynamic measure was around the centroid of the acquisition system, while kinematics measurements close to the X-rays detectors gave the worst errors.     

The accuracy and repeatability of an automatic 2D-3D fluoroscopic image-model registration technique for determining shoulder joint kinematics

Fluoroscopic imaging, using single plane or dual plane images, has grown in popularity to measure dynamic in vivo human shoulder joint kinematics. However, no study has quantified the difference in spatial positional accuracy between single and dual plane image-model registration applied to the shoulder joint. In this paper, an automatic 2D-3D image-model registration technique was validated for accuracy and repeatability with single and dual plane fluoroscopic images. Accuracy was assessed in a cadaver model, kinematics found using the automatic registration technique were compared to those found using radiostereometric analysis. The in vivo repeatability of the automatic registration technique was assessed during the dynamic abduction motion of four human subjects. The in vitro data indicated that the error in spatial positional accuracy of the humerus and the scapula was less than 0.30mm in translation and less than 0.58° in rotation using dual plane images. Single plane accuracy was satisfactory for in-plane motion variables, but out-of-plane motion variables on average were approximately 8 times less accurate. The in vivo test indicated that the repeatability of the automatic 2D-3D image-model registration was 0.50mm in translation and 1.04° in rotation using dual images. For a single plane technique, the repeatability was 3.31mm in translation and 2.46° in rotation for measuring shoulder joint kinematics. The data demonstrate that accurate and repeatable shoulder joint kinematics can be obtained using dual plane fluoroscopic images with an automatic 2D-3D image-model registration technique; and that out-of-plane motion variables are less accurate than in-plane motion variables using a single plane technique.     

A fast, accurate, and automatic 2D-3D image registration for image-guided cranial radiosurgery

The authors developed a fast and accurate two-dimensional (2D)-three-dimensional (3D) image registration method to perform precise initial patient setup and frequent detection and correction for patient movement during image-guided cranial radiosurgery treatment. In this method, an approximate geometric relationship is first established to decompose a 3D rigid transformation in the 3D patient coordinate into in-plane transformations and out-of-plane rotations in two orthogonal 2D projections. Digitally reconstructed radiographs are generated offline from a preoperative computed tomography volume prior to treatment and used as the reference for patient position. A multiphase framework is designed to register the digitally reconstructed radiographs with the x-ray images periodically acquired during patient setup and treatment. The registration in each projection is performed independently; the results in the two projections are then combined and converted to a 3D rigid transformation by 2D-3D geometric backprojection. The in-plane transformation and the out-of-plane rotation are estimated using different search methods, including multiresolution matching, steepest descent minimization, and one-dimensional search. Two similarity measures, optimized pattern intensity and sum of squared difference, are applied at different registration phases to optimize accuracy and computation speed. Various experiments on an anthropomorphic head-and-neck phantom showed that, using fiducial registration as a gold standard, the registration errors were 0.33 +/- 0.16 mm (s.d.) in overall translation and 0.29 degrees +/- 0.11 degrees (s.d.) in overall rotation. The total targeting errors were 0.34 +/- 0.16 mm (s.d.), 0.40 +/- 0.2 mm (s.d.), and 0.51 +/- 0.26 mm (s.d.) for the targets at the distances of 2, 6, and 10 cm from the rotation center, respectively. The computation time was less than 3 s on a computer with an Intel Pentium 3.0 GHz dual processor.     

The quality of bone surfaces may govern the use of model based fluoroscopy in the determination of joint laxity

The assessment of knee joint laxity is clinically important but its quantification remains elusive. Calibrated, low dosage fluoroscopy, combined with registered surfaces and controlled external loading may offer possible solutions for quantifying relative tibio-femoral motion without soft tissue artefact, even in native joints. The aim of this study was to determine the accuracy of registration using CT and MRI derived 3D bone models, as well as metallic implants, to 2D single-plane fluoroscopic datasets, to assess their suitability for examining knee joint laxity.Four cadaveric knees and one knee implant were positioned using a micromanipulator. After fluoroscopy, the accuracy of registering each surface to the 2D fluoroscopic images was determined by comparison against known translations from the micromanipulator measurements. Dynamic measurements were also performed to assess the relative tibio-femoral error. For CT and MRI derived 3D femur and tibia models during static testing, the in-plane error was 0.4 mm and 0.9 mm, and out-of-plane error 2.6 mm and 9.3 mm respectively. For metallic implants, the in-plane error was 0.2 mm and out-of-plane error 1.5 mm. The relative tibio-femoral error during dynamic measurements was 0.9 mm, 1.2 mm and 0.7 mm in-plane, and 3.9 mm, 10.4 mm and 2.5 mm out-of-plane for CT and MRI based models and metallic implants respectively. The rotational errors ranged from 0.5° to 1.9° for CT, 0.5–4.3° for MRI and 0.1–0.8° for metallic implants.The results of this study indicate that single-plane fluoroscopic analysis can provide accurate information in the investigation of knee joint laxity, but should be limited to static or quasi-static evaluations when assessing native bones, where possible. With this knowledge of registration accuracy, targeted approaches for the determination of tibio-femoral laxity could now determine objective in vivo measures for the identification of ligament reconstruction candidates as well as improve our understanding of the consequences of knee joint instability in TKA.     

Hierarchical model-based tracking of cervical vertebrae from dynamic biplane radiographs

We present a novel approach for automatically, accurately and reliably determining the 3D motion of the cervical spine from a series of stereo or biplane radiographic images. These images could be acquired through a variety of different imaging hardware configurations. We follow a hierarchical, anatomically-aware, multi-bone approach that takes into account the complex structure of cervical vertebrae and inter-vertebrae overlapping, as well as the temporal coherence in the imaging series. These significant innovations improve the speed, accuracy, reliability and flexibility of the tracking process. Evaluation on cervical data shows that the approach is as accurate (average precision 0.3 mm and 1°) as the expert human-operator driven method that was previously state of the art. However, unlike the previously used method, the hierarchical approach is automatic and robust; even in the presence of implanted hardware. Therefore, the method has solid potential for clinical use to evaluate the effectiveness of surgical interventions.     

Performance evaluation of a CyberKnife G4 image-guided robotic stereotactic radiosurgery system

The aim of the current work was to present the performance evaluation procedures implemented at our department for the commissioning of a G4 CyberKnife system. This system consists of a robotic manipulator, a target-locating system and a lightweight 6-MV linac. Individual quality assurance procedures were performed for each of the CyberKnife subsystems. The system was checked for the mechanical accuracy of its robotic manipulator. The performance of the target-locating system was evaluated in terms of mechanical accuracy of both cameras' alignment and quality assurance tests of the x-ray generators and the flat-panel detectors. The traditional linac 6-MV beam characteristics and beam output parameters were also measured. Results revealed a manipulator mechanical mean accuracy of approximately 0.1 mm, with individual maximum position uncertainties less than 0.25 mm. The target-locating system mechanical accuracy was found within the acceptance limits. For the most clinically used parameters in the CyberKnife practice, e.g. 100-120 kV and 50-200 ms, kV and exposure time accuracy error were measured as less than 2%, while the precision error of the kV was determined as less than 1%. The acquired images of the ETR grid pattern revealed no geometrical distortion while the critical frequency f(50) values for cameras A and B were calculated as 1.5 lp mm(-1) and 1.4 lp mm(-1), respectively. Dose placement measurements were performed in a head and neck phantom. Results revealed sub-millimeter beam delivery precision whereas the total clinical accuracy of the system was measured equal to 0.44 +/- 0.12 mm, 0.29 +/- 0.10 mm and 0.53 +/- 0.16 mm for the skull, fiducial and Xsight spine tracking methods, respectively. The results of this work certify the G4 CyberKnife SRS system capable of delivering high dose distributions with sub-millimeter accuracy and precision to intracranial and extracranial lesions. Moreover, total clinical accuracy of the investigated G4 system was found to be improved for the skull and fiducial tracking methods and was comparable for Xsight spine tracking method compared with the earlier generation of the instrument.     

A phantom study on the positioning accuracy of the Novalis Body system

A phantom study was conducted to investigate inherent positioning accuracy of an image-guided patient positioning system-the Novalis Body system for three-dimensional (3-D) conformal radiotherapy. This positioning system consists of two infrared (IR) cameras and one video camera and two kV x-ray imaging devices. The initial patient setup was guided by the IR camera system and the target localization was accomplished using the kV x-ray imaging system. In this study, the IR marker shift and phantom rotation were simulated and their effects on the positioning accuracy were examined by a Rando phantom. The effects of CT slice thickness and treatment sites on the positioning accuracy were tested. In addition, the internal target shift was simulated and its effect on the positioning accuracy was examined by a water tank. With the application of the Novalis Body system, the positioning error of the planned isocenter was significantly reduced. The experimental results with the simulated IR marker shifts indicated that the positioning errors of the planned isocenter were 0.6 +/- 0.3, 0.5 +/- 0.2, and 0.7 +/- 0.2 mm along the lateral, longitudinal, and vertical axes, respectively. The experimental results with the simulated phantom rotations indicated that the positioning errors of the planned isocenter were 0.6 +/- 0.3, 0.7 +/- 0.2, and 0.5 +/- 0.2 mm along the three axes, respectively. The experimental results with the simulated target shifts indicated that the positioning errors of the planned isocenter were 0.6 +/- 0.3, 0.7 +/- 0.2, and 0.5 +/- 0.2 mm along the three axes, respectively. On average, the positioning accuracy of 1 mm for the planned isocenter was achieved using the Novalis Body system.     

Accuracy of cranial coplanar beam therapy using an oblique, stereoscopic x-ray image guidance system

A system for measuring two-dimensional (2D) dose distributions in orthogonal anatomical planes in the cranium was developed and used to evaluate the accuracy of coplanar conformal therapy using ExacTrac image guidance. Dose distributions were measured in the axial, sagittal, and coronal planes using a CIRS (Computerized Imaging Reference Systems, Inc.) anthropomorphic head phantom with a custom internal film cassette. Sections of radiographic Kodak EDR2 film were cut, processed, and digitized using custom templates. Spatial and dosimetric accuracy and precision of the film system were assessed. BrainScan planned a coplanar-beam treatment to conformally irradiate a 2-cm-diameter x 2-cm-long cylindrical planning target volume. Prior to delivery, phantom misalignments were imposed in combinations of +/- 8 mm offsets in each of the principal directions. ExacTrac x-ray correction was applied until the phantom was within an acceptance criteria of 1 mm/1 degrees (first two measurement sets) or 0.4 mm/0.4 degrees (last two measurement sets). Measured dose distributions from film were registered to the treatment plan dose calculations and compared. Alignment errors, displacement between midpoints of planned and measured 70% isodose contours (Deltac), and positional errors of the 80% isodose line were evaluated using 49 2D film measurements (98 profiles). Comparison of common, but independent measurements of Deltac showed that systematic errors in the measurement technique were 0.2 mm or less along all three anatomical axes and that random error averaged [formula: see text] 0.29+/-0.06 mm for the acceptance criteria of 1 mm/1 degrees and 0.15 +/- 0.02 mm for the acceptance criteria of 0.4 mm/0.4 degrees. The latter was consistent with independent estimates that showed the precision of the measurement system was 0.3 mm (2sigma). Values of Deltac were as great as 0.9, 0.3, and 1.0 mm along the P-A, R-L, and I-S axes, respectively. Variations in Deltac along the P-A axis were correlated to misalignments between laser isocenter and radiation isocenter as documented by daily clinical Lutz tests. Based on results of comparisons of measured with calculated positions of the 80% dose lines along the major anatomical axes, a 1.25, 1.0, and 1.0 mm (0.75, 0.5, and 0.25 mm) gross tumor volume (GTV)-planning target volume (PTV) margin to account for delivery error would be appropriate for the P-A, R-L, and I-S axes, respectively, for an acceptance criteria of 1 mm/1 degrees (0.4 mm/0.4 degrees). It typically took 2 (3) ExacTrac x-ray image sets to achieve and verify acceptance criteria of 1 mm/1 degrees (0.4 mm/0.4 degrees). Our results demonstrated a measurement technique using a CIRS anthropomorphic head phantom with a modified film cassette, radiographic film (Kodak EDR2) with a custom film cutting template, and film dosimetry software has been developed and successfully applied to our clinic. It is recommended that a third party offer this service. Our goal of achieving accuracy of delivery of 1 mm or better in each of the three major anatomical axes was almost, but not quite achieved, not because of the accuracy of the image guidance system, but likely due to inaccuracy of laser isocenter and other systematic errors.     

Model-based compensation for quantitative 123I brain SPECT imaging

Previously we have developed a model-based method that can accurately estimate downscatter contamination from high-energy photons in 123I imaging. In this work we combined the model-based method with iterative reconstruction-based compensations for other image-degrading factors such as attenuation, scatter, the collimator-detector response function (CDRF) and partial volume effects to form a comprehensive method for performing quantitative 123I SPECT image reconstruction. In the model-based downscatter estimation method, photon scatter inside the object was modelled using the effective source scatter estimation (ESSE) technique, including contributions from all the photon emissions. The CDRFs, including the penetration and scatter components due to the high-energy 123I photons, were estimated using Monte Carlo (MC) simulations of point sources in air at various distances from the face of the collimator. The downscatter contamination was then compensated for during the iterative reconstruction by adding the estimated results to the projection steps. The model-based downscatter compensation (MBDC) was evaluated using MC simulated and experimentally acquired projection data. From the MC simulation, we found about 39% of the total counts in the energy window of 123I were attributed to the downscatter contamination, which reduced image contrast and caused a 1.5% to 10% overestimation of activities in various brain structures. Model-based estimates of the downscatter contamination were in good agreement with the simulated data. Compensation using MBDC removed the contamination and improved the image contrast and quantitative accuracy to that of the images obtained from 159 keV photons. The errors in absolute quantitation were reduced to within +/-3.5%. The striatal specific binding potential calculated based on the activity ratio to the background was also improved after MBDC. The errors were reduced from -4.5% to -10.93% without compensation to -0.55% to 4.87% after compensation. The model-based method provided accurate downscatter estimation and, when combined with iterative reconstruction-based compensations, accurate quantitation was obtained with minimal loss of precision.     

Accurate localization of intracavitary brachytherapy applicators from 3D CT imaging studies

PURPOSE: To present an accurate method to identify the positions and orientations of intracavitary (ICT) brachytherapy applicators imaged in 3D CT scans, in support of Monte Carlo photon-transport simulations, enabling accurate dose modeling in the presence of applicator shielding and interapplicator attenuation. MATERIALS AND METHODS: The method consists of finding the transformation that maximizes the coincidence between the known 3D shapes of each applicator component (colpostats and tandem) with the volume defined by contours of the corresponding surface on each CT slice. We use this technique to localize Fletcher-Suit CT-compatible applicators for three cervix cancer patients using post-implant CT examinations (3 mm slice thickness and separation). Dose distributions in 1-to-1 registration with the underlying CT anatomy are derived from 3D Monte Carlo photon-transport simulations incorporating each applicator's internal geometry (source encapsulation, high-density shields, and applicator body) oriented in relation to the dose matrix according to the measured localization transformations. The precision and accuracy of our localization method are assessed using CT scans, in which the positions and orientations of dense rods and spheres (in a precision-machined phantom) were measured at various orientations relative to the gantry. RESULTS: Using this method, we register 3D Monte Carlo dose calculations directly onto post insertion patient CT studies. Using CT studies of a precisely machined phantom, the absolute accuracy of the method was found to be +/-0.2 mm in plane, and +/-0.3 mm in the axial direction while its precision was +/-0.2 mm in plane, and +/-0.2 mm axially. CONCLUSION: We have developed a novel, and accurate technique to localize intracavitary brachytherapy applicators in 3D CT imaging studies, which supports 3D dose planning involving detailed 3D Monte Carlo dose calculations, modeling source positions, shielding and interapplicator shielding, accurately.     

The effects of stereo shift angle,geometric magnification and display zoom on depth measurements in digital stereomammography

We are developing virtual three-dimensional (3-D) cursors for measuring depths in digital stereomammograms. We performed a study to investigate the effects of stereo shift angle, geometric magnification, and display zoom on the accuracy of depth measurements made with a virtual 3-D cursor. A phantom containing 50 low contrast fibrils at depths ranging from 1 to 11 mm was imaged with a full-field digital mammography system. Left- and right-eye images were generated at stereo shift angles of +/-3 degrees and +/-6 degrees, using either contact or 1.8x geometric magnification geometry. The images were viewed on a high-resolution stereoscopic display system in normal and 2x zoom mode. Observers viewed the images with stereo glasses and adjusted the depth of a cross-shaped virtual cursor to best match the perceived depth of each fibril. The results for two trained observers with excellent stereo acuity were nearly identical when viewing the same images. The average root mean square errors for the two observers were 1.2 mm (+/-3 degrees contact, no zoom), 1.3 mm (+/-3 degrees contact zoom), 0.8 mm (+/-6 degrees contact, no zoom), 0.6 mm (+/-6 degrees contact, zoom), 0.8 mm (+/-3 magnification, no zoom), 0.7 mm (+/-3 degrees magnification, zoom), and 0.2 mm (+/-6 degrees magnification, no zoom). One observer repeated the entire study for two additional fibril phantom configurations. Combining all the results, we found that for the contact geometry increasing the stereo shift angle from +/-3 degrees to +/-6 degrees improved the depth measurement accuracy by factors of about 1.2-4.0. Zooming did not provide observable improvement in the depth measurement accuracy; sometimes having no effect, sometimes improving the accuracy, and other times reducing the accuracy, with no general trends. Its effect is likely within experimental errors. However, the stereo effect was more readily visualized in the zoom mode. Geometric magnification improved the depth measurement accuracy. The best accuracy among all cases was about 0.2 mm, obtained with geometric magnification using a stereo angle of +/-6 degrees. This is the mode we recommend for obtaining accurate depth measurements with virtual cursors in stereomammograms.     

Evaluation of a new electromagnetic tracking system using a standardized assessment protocol

This note uses a published protocol to evaluate a newly released 6 degrees of freedom electromagnetic tracking system (Aurora, Northern Digital Inc.). A practice for performance monitoring over time is also proposed. The protocol uses a machined base plate to measure relative error in position and orientation as well as the influence of metallic objects in the operating volume. Positional jitter (E(RMS)) was found to be 0.17 mm +/- 0.19 mm. A relative positional error of 0.25 mm +/- 0.22 mm at 50 mm offsets and 0.97 mm +/- 1.01 mm at 300 mm offsets was found. The mean of the relative rotation error was found to be 0.20 degrees +/- 0.14 degrees with respect to the axial and 0.91 degrees +/- 0.68 degrees for the longitudinal rotation. The most significant distortion caused by metallic objects is caused by 400-series stainless steel. A 9.4 mm maximum error occurred when the rod was closest to the emitter, 10 mm away. The improvement compared to older generations of the Aurora with respect to accuracy is substantial.     

IMRT verification with a camera-based electronic portal imaging system

An evaluation of the capabilities of a commercially available camera-based electronic portal imaging system for intensity-modulated radiotherapy verification is presented. Two modifications to the system are demonstrated which use a novel method to tag each image acquired with the delivered dose measured by the linac monitor chamber and reduce optical cross-talk in the imager. A detailed performance assessment is presented, including measurements of the optical decay characteristics of the system. The overall geometric accuracy of the system is determined to be +/-2.0 mm, with a dosimetric accuracy of +/-1.25 MU. Finally a clinical breast IMRT treatment, delivered by dynamic multileaf collimation, is successfully verified both by tracking the position of each leaf during beam delivery and recording the integrated intensity observed over the entire beam.     

Adaptive prediction of respiratory motion for motion compensation radiotherapy

One potential application of image-guided radiotherapy is to track the target motion in real time, then deliver adaptive treatment to a dynamic target by dMLC tracking or respiratory gating. However, the existence of a finite time delay (or a system latency) between the image acquisition and the response of the treatment system to a change in tumour position implies that some kind of predictive ability should be included in the real-time dynamic target treatment. If diagnostic x-ray imaging is used for the tracking, the dose given over a whole image-guided radiotherapy course can be significant. Therefore, the x-ray beam used for motion tracking should be triggered at a relatively slow pulse frequency, and an interpolation between predictions can be used to provide a fast tracking rate. This study evaluates the performance of an autoregressive-moving average (ARMA) model based prediction algorithm for reducing tumour localization error due to system latency and slow imaging rate. For this study, we use 3D motion data from ten lung tumour cases where the peak-to-peak motion is greater than 8 mm. Some strongly irregular traces with variation in amplitude and phase were included. To evaluate the prediction accuracy, the standard deviations between predicted and actual motion position are computed for three system latencies (0.1, 0.2 and 0.4 s) at several imaging rates (1.25-10 Hz), and compared against the situation of no prediction. The simulation results indicate that the implementation of the prediction algorithm in real-time target tracking can improve the localization precision for all latencies and imaging rates evaluated. From a common initial setting of model parameters, the predictor can quickly provide an accurate prediction of the position after collecting 20 initial data points. In this retrospective analysis, we calculate the standard deviation of the predicted position from the twentieth position data to the end of the session at 0.1 s interval. For both regular and irregular lung tumour motions, with prediction the range of average errors is 0.4-2.5 mm in the SI direction from shorter to longer latency, corresponding to a range of 0.8-4.3 mm without prediction; for the AP direction a range of 0.3-1.6 mm is obtained with prediction, corresponding to a range of 0.6-3.0 mm without prediction. For 0.2 s and 0.4 s system latency, with prediction the localization based on a relatively slow imaging rate (2.5 Hz) can achieve a better or similar precision compared with no prediction but on a fast imaging rate (10 Hz). This means that precise localization can be realized at a slow imaging rate. This is important for the application of kV x-ray imaging systems and EPID-based systems in image-guided radiotherapy. In conclusion, the adaptive predictor can successfully predict irregular respiratory motion, and the adaptive prediction of respiration motion can effectively improve the delivery precision of real-time motion compensation radiotherapy.     

Phantom validation of coregistration of PET and CT for image-guided radiotherapy

Radiotherapy treatment planning integrating positron emission tomography (PET) and computerized tomography (CT) is rapidly gaining acceptance in the clinical setting. Although hybrid systems are available, often the planning CT is acquired on a dedicated system separate from the PET scanner. A limiting factor to using PET data becomes the accuracy of the CT/PET registration. In this work, we use phantom and patient validation to demonstrate a general method for assessing the accuracy of CT/PET image registration and apply it to two multi-modality image registration programs. An IAEA (International Atomic Energy Association) brain phantom and an anthropomorphic head phantom were used. Internal volumes and externally mounted fiducial markers were filled with CT contrast and 18F-fluorodeoxyglucose (FDG). CT, PET emission, and PET transmission images were acquired and registered using two different image registration algorithms. CT/PET Fusion (GE Medical Systems, Milwaukee, WI) is commercially available and uses a semi-automated initial step followed by manual adjustment. Automatic Mutual Information-based Registration (AMIR), developed at our institution, is fully automated and exhibits no variation between repeated registrations. Registration was performed using distinct phantom structures; assessment of accuracy was determined from registration of the calculated centroids of a set of fiducial markers. By comparing structure-based registration with fiducial-based registration, target registration error (TRE) was computed at each point in a three-dimensional (3D) grid that spans the image volume. Identical methods were also applied to patient data to assess CT/PET registration accuracy. Accuracy was calculated as the mean with standard deviation of the TRE for every point in the 3D grid. Overall TRE values for the IAEA brain phantom are: CT/PET Fusion = 1.71 +/- 0.62 mm, AMIR = 1.13 +/- 0.53 mm; overall TRE values for the anthropomorphic head phantom are: CT/PET Fusion = 1.66 +/- 0.53 mm, AMIR = 1.15 +/- 0.48 mm. Precision (repeatability by a single user) measured for CT/PET Fusion: IAEA phantom = 1.59 +/- 0.67 mm and anthropomorphic head phantom = 1.63 +/- 0.52 mm. (AMIR has exact precision and so no measurements are necessary.) One sample patient demonstrated the following accuracy results: CT/PET Fusion = 3.89 +/- 1.61 mm, AMIR = 2.86 +/- 0.60 mm. Semi-automatic and automatic image registration methods may be used to facilitate incorporation of PET data into radiotherapy treatment planning in relatively rigid anatomic sites, such as head and neck. The overall accuracies in phantom and patient images are < 2 mm and < 4 mm, respectively, using either registration algorithm. Registration accuracy may decrease, however, as distance from the initial registration points (CT/PET fusion) or center of the image (AMIR) increases. Additional information provided by PET may improve dose coverage to active tumor subregions and hence tumor control. This study shows that the accuracy obtained by image registration with these two methods is well suited for image-guided radiotherapy.     

Stereotactic imaging for radiotherapy: accuracy of CT,MRI, PET and SPECT

CT, MRI, PET and SPECT provide complementary information for treatment planning in stereotactic radiotherapy. Stereotactic correlation of these images requires commissioning tests to confirm the localization accuracy of each modality. A phantom was developed to measure the accuracy of stereotactic localization for CT, MRI, PET and SPECT in the head and neck region. To this end. the stereotactically measured coordinates of structures within the phantom were compared with their mechanically defined coordinates. For MRI, PET and SPECT, measurements were performed using two different devices. For MRI, T1- and T2-weighted imaging sequences were applied. For each measurement, the mean radial deviation in space between the stereotactically measured and mechanically defined position of target points was determined. For CT, the mean radial deviation was 0.4 +/- 0.2 mm. For MRI, the mean deviations ranged between 0.7 +/- 0.2 mm and 1.4 +/- 0.5 mm, depending on the MRI device and the imaging sequence. For PET, mean deviations of 1.1 +/- 0.5 mm and 2.4 +/- 0.3 mm were obtained. The mean deviations for SPECT were 1.6 +/- 0.5 mm and 2.0 +/- 0.6 mm. The phantom is well suited to determine the accuracy of stereotactic localization with CT, MRI, PET and SPECT in the head and neck region. The obtained accuracy is well below the physical resolution for CT, PET and SPECT, and of comparable magnitude for MRI. Since the localization accuracy may be device dependent, results obtained at one device cannot be generalized to others.     

Detection of treatment setup errors between two CT scans for patients with head and neck cancer

Accuracy of treatment setup for head and neck patients undergoing intensity-modulated radiation therapy is of paramount importance. The conventional method using orthogonal portal images can only detect translational setup errors while the most frequent setup errors for head and neck patients could be rotational errors. With the rapid development of image-guided radiotherapy, three-dimensional images are readily acquired and can be used to detect both translational and rotational setup errors. The purpose of this study is to determine the significance of rotational variations between two planning CT scans acquired for each of eight head and neck patients, who experienced substantial weight loss or tumor shrinkage. To this end, using a rigid body assumption, we developed an in-house computer program that utilizes matrix transformations to align point bony landmarks with an incremental best-fit routine. The program returns the quantified translational and rotational shifts needed to align the scans of each patient. The program was tested using a phantom for a set of known translational and rotational shifts. For comparison, a commercial treatment planning system was used to register the two CT scans and estimate the translational errors for these patients. For the eight patients, we found that the average magnitudes and standard deviations of the rotational shifts about the transverse, anterior-posterior, and longitudinal axes were 1.7 +/- 2.3 degrees, 0.8 +/- 0.7 degrees, and 1.8 +/- 1.1 degrees, respectively. The average magnitudes and standard deviations of the translational shifts were 2.5 +/- 2.6 mm, 2.9 +/- 2.8 mm, 2.7 +/- 1.7 mm while the differences detected between our program and the CT-CT fusion method were 1.8 +/- 1.3 mm, 3.3 +/- 5.4 mm, and 3.0 +/- 3.4 mm in the left-right, anterior-posterior, and superior-inferior directions, respectively. A trend of larger rotational errors resulting in larger translational differences between the two methods was observed. In conclusion, conventional methods used for verifying patient positioning may misinterpret rotational shifts as translational shifts, and our study demonstrated that rotational errors may be significant in the treatment of head and neck cancer.     

In vitro quantification of the performance of model-based mono-planar and bi-planar fluoroscopy for 3D joint kinematics estimation

Model-based mono-planar and bi-planar 3D fluoroscopy methods can quantify intact joints kinematics with performance/cost trade-off. The aim of this study was to compare the performances of mono- and bi-planar setups to a marker-based gold-standard, during dynamic phantom knee acquisitions. Absolute pose errors for in-plane parameters were lower than 0.6 mm or 0.6° for both mono- and bi-planar setups. Mono-planar setups resulted critical in quantifying the out-of-plane translation (error < 6.5 mm), and bi-planar in quantifying the rotation along bone longitudinal axis (error < 1.3°). These errors propagated to joint angles and translations differently depending on the alignment of the anatomical axes and the fluoroscopic reference frames. Internal-external rotation was the least accurate angle both with mono- (error < 4.4°) and bi-planar (error < 1.7°) setups, due to bone longitudinal symmetries. Results highlighted that accuracy for mono-planar in-plane pose parameters is comparable to bi-planar, but with halved computational costs, halved segmentation time and halved ionizing radiation dose. Bi-planar analysis better compensated for the out-of-plane uncertainty that is differently propagated to relative kinematics depending on the setup. To take its full benefits, the motion task to be investigated should be designed to maintain the joint inside the visible volume introducing constraints with respect to mono-planar analysis.