Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Image Segmentation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,549 records · Page 86

Evaluation of magnetic shear in off-disk center active regions

The changes that projection effects produce in the evaluation of magnetic shear in off-disk center active regions by comparing angular shear calculated in image plane and heliographic coordinates are analyzed, and the procedure for properly evaluating magnetic shear by transforming the observed vector magnetic field into the heliographic system is described. This procedure is then used to evaluate magnetic shear along the magnetic neutral line in an active region that was observed on April 24, 1984 at a longitude offset of -45 deg. In particular, the number of 'critically sheared' pixels along an east-west directed segment of the neutral line in the leader sunspot group changes from 16 in the image plane magnetogram to 14 in the heliographic magnetogram. The critical shear as calculated in the image plane served as a good predictor for the location of flaring activity since the flare ribbons of the great flare of April 24 bracketed the inversion line where the critical shear was located. These results indicate that for this particular region, projection effects did not significantly affect the evaluation of critical shear.

Venkatakrishnan, P.↗

Method for implementation of recursive hierarchical segmentation on parallel computers

A method, computer readable storage, and apparatus for implementing a recursive hierarchical segmentation algorithm on a parallel computing platform. The method includes setting a bottom level of recursion that defines where a recursive division of an image into sections stops dividing, and setting an intermediate level of recursion where the recursive division changes from a parallel implementation into a serial implementation. The segmentation algorithm is implemented according to the set levels. The method can also include setting a convergence check level of recursion with which the first level of recursion communicates with when performing a convergence check.

Tilton, James C.↗

Estimating nonrigid motion from inconsistent intensity with robust shape features

Purpose: To develop a nonrigid motion estimation method that is robust to heterogeneous intensity inconsistencies amongst the image pairs or image sequence. Methods: Intensity and contrast variations, as in dynamic contrast enhanced magnetic resonance imaging, present a considerable challenge to registration methods based on general discrepancy metrics. In this study, the authors propose and validate a novel method that is robust to such variations by utilizing shape features. The geometry of interest (GOI) is represented with a flexible zero level set, segmented via well-behaved regularized optimization. The optimization energy drives the zero level set to high image gradient regions, and regularizes it with area and curvature priors. The resulting shape exhibits high consistency even in the presence of intensity or contrast variations. Subsequently, a multiscale nonrigid registration is performed to seek a regular deformation field that minimizes shape discrepancy in the vicinity of GOIs. Results: To establish the working principle, realistic 2D and 3D images were subject to simulated nonrigid motion and synthetic intensity variations, so as to enable quantitative evaluation of registration performance. The proposed method was benchmarked against three alternative registration approaches, specifically, optical flow, B-spline based mutual information, and multimodality demons. When intensity consistency was satisfied, all methods had comparable registration accuracy for the GOIs. When intensities among registration pairs were inconsistent, however, the proposed method yielded pronounced improvement in registration accuracy, with an approximate fivefold reduction in mean absolute error (MAE = 2.25 mm, SD = 0.98 mm), compared to optical flow (MAE = 9.23 mm, SD = 5.36 mm), B-spline based mutual information (MAE = 9.57 mm, SD = 8.74 mm) and mutimodality demons (MAE = 10.07 mm, SD = 4.03 mm). Applying the proposed method on a real MR image sequence also provided qualitatively appealing results, demonstrating good feasibility and applicability of the proposed method. Conclusions: The authors have developed a novel method to estimate the nonrigid motion of GOIs in the presence of spatial intensity and contrast variations, taking advantage of robust shape features. Quantitative analysis and qualitative evaluation demonstrated good promise of the proposed method. Further clinical assessment and validation is being performed.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Oriented Connectivity-Based Method for Segmenting Solar Loops

A method based on oriented connectivity that can automatically segnient arc-like structures (solar loops) from intensity images of the Sun's corona is introduced. The method is a constructive approach that uses model-guided processing to enable extraction of credible loop structures. Since the solar loops are vestiges of the solar magnetic field, the model-guided processing exploits external estimates of this field s local orientations that are derived from a physical magnetic field model. Empirical studies of the method s effectiveness are also presented. The Oriented Connectivity- Based Method is the first automatic method for the segmentation of solar loops.

Lee, J. K.↗

SU-E-J-209: Verification of 3D Surface Registration Between Stereograms and CT Images

Purpose: Stereography can provide a visualization of the skin surface for radiation therapy patients. The aim of this study was to verify the registration algorithm in a commercial image analysis software, 3dMDVultus, for the fusion of stereograms and CT images. Methods: CT and stereographic scans were acquired of a head phantom and a deformable phantom. CT images were imported in 3dMDVultus and the surface contours were generated by threshold segmentation. Stereograms were reconstructed in 3dMDVultus. The resulting surfaces were registered with Vultus algorithm and then exported to in-house registration software and compared with four algorithms: rigid, affine, non-rigid iterative closest point (ICP) and b-spline algorithm. RMS (root-mean-square residuals of the surface point distances) error between the registered CT and stereogram surfaces was calculated and analyzed. Results: For the head phantom, the maximum RMS error between registered CT surfaces to stereogram was 6.6 mm for Vultus algorithm, whereas the mean RMS error was 0.7 mm. For the deformable phantom, the maximum RMS error was 16.2 mm for Vultus algorithm, whereas the mean RMS error was 4.4 mm. Non-rigid ICP demonstrated the best registration accuracy, as the mean of RMS errors were both within 1 mm. Conclusion: The accuracy of registration algorithm in 3dMDVultus was verified and exceeded RMS of 2 mm for deformable cases. Non-rigid ICP and b-spline algorithms improve the registration accuracy for both phantoms, especially in deformable one. For those patients whose body habitus deforms during radiation therapy, more advanced nonrigid algorithms need to be used.

60 APPLIED LIFE SCIENCES↗

Pulmonary lobe segmentation based on ridge surface sampling and shape model fitting

Purpose: Performing lobe-based quantitative analysis of the lung in computed tomography (CT) scans can assist in efforts to better characterize complex diseases such as chronic obstructive pulmonary disease (COPD). While airways and vessels can help to indicate the location of lobe boundaries, segmentations of these structures are not always available, so methods to define the lobes in the absence of these structures are desirable. Methods: The authors present a fully automatic lung lobe segmentation algorithm that is effective in volumetric inspiratory and expiratory computed tomography (CT) datasets. The authors rely on ridge surface image features indicating fissure locations and a novel approach to modeling shape variation in the surfaces defining the lobe boundaries. The authors employ a particle system that efficiently samples ridge surfaces in the image domain and provides a set of candidate fissure locations based on the Hessian matrix. Following this, lobe boundary shape models generated from principal component analysis (PCA) are fit to the particles data to discriminate between fissure and nonfissure candidates. The resulting set of particle points are used to fit thin plate spline (TPS) interpolating surfaces to form the final boundaries between the lung lobes. Results: The authors tested algorithm performance on 50 inspiratory and 50 expiratory CT scans taken from the COPDGene study. Results indicate that the authors' algorithm performs comparably to pulmonologist-generated lung lobe segmentations and can produce good results in cases with accessory fissures, incomplete fissures, advanced emphysema, and low dose acquisition protocols. Dice scores indicate that only 29 out of 500 (5.85%) lobes showed Dice scores lower than 0.9. Two different approaches for evaluating lobe boundary surface discrepancies were applied and indicate that algorithm boundary identification is most accurate in the vicinity of fissures detectable on CT. Conclusions: The proposed algorithm is effective for lung lobe segmentation in absence of auxiliary structures such as vessels and airways. The most challenging cases are those with mostly incomplete, absent, or near-absent fissures and in cases with poorly revealed fissures due to high image noise. However, the authors observe good performance even in the majority of these cases.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Three-dimensional super line-localization in low signal-to-noise microscope images via prior-apprised unsupervised learning (PAUL)

Biological processes such as processive enzyme turnover and intracellular cargo tracking involve the dynamic motion of a small "article" along a curvilinear biopolymer track. To understand these processes that occur across multiple length and time scales, one must acquire both the trajectory of the particle and the position of the track along which it moves, possibly by combining high-resolution single-particle tracking with conventional microscopy. Yet, usually there is a significant resolution mismatch between these modalities: while the tracked particle is localized with a precision of 10 nm, the image of the surroundings is limited by optical difraction, with 200 nm lateral and 500 nm axial resolutions. Compared to the particle's trajectory, the surrounding curvilinear structure appears as a blurred and noisy image. This disparity in the spatial resolutions of the particle trajectory and the surrounding curvilinear structure image makes data reconstruction, as well as interpretation, particularly challenging. Analysis is further complicated when the curvilinear structures are oriented arbitrarily in 3D space. Here, we present a prior-apprised unsupervised learning (PAUL) approach to extract information from 3D images where the underlying features resemble a curved line such as a filament or microtubule. This three-stage framework starts with a Hessian-based feature enhancement, which is followed by feature registration, where local line segments are detected on repetitively sampled subimage tiles. In the final stage, statistical learning, segments are clustered based on their geometric relationships. Principal curves are then approximated from each segment group via statistical tools including principal component analysis, bootstrap and kernel transformation. This procedure is characterized on simulated images, where sub-voxel medium deviations from true curves have been achieved. The 3D PAUL approach has also been implemented for successful line localization in experimental 3D images of gold nanowires obtained using a multifocal microscope. Lastly, this work not only bridges the resolution gap between two microscopy modalities, but also allows us to conduct 3D super line-localization imaging experiments, without using super-resolution techniques.

42 ENGINEERING↗

SU-F-J-158: Respiratory Motion Resolved, Self-Gated 4D-MRI Using Rotating Cartesian K-Space Sampling

Purpose: Dynamic MRI has been used to quantify respiratory motion of abdominal organs in radiation treatment planning. Many existing 4D-MRI methods based on 2D acquisitions suffer from limited slice resolution and additional stitching artifacts when evaluated in 3D{sup 1}. To address these issues, we developed a 4D-MRI (3D dynamic) technique with true 3D k-space encoding and respiratory motion self-gating. Methods: The 3D k-space was acquired using a Rotating Cartesian K-space (ROCK) pattern, where the Cartesian grid was reordered in a quasi-spiral fashion with each spiral arm rotated using golden angle{sup 2}. Each quasi-spiral arm started with the k-space center-line, which were used as self-gating{sup 3} signal for respiratory motion estimation. The acquired k-space data was then binned into 8 respiratory phases and the golden angle ensures a near-uniform k-space sampling in each phase. Finally, dynamic 3D images were reconstructed using the ESPIRiT technique{sup 4}. 4D-MRI was performed on 6 healthy volunteers, using the following parameters (bSSFP, Fat-Sat, TE/TR=2ms/4ms, matrix size=500×350×120, resolution=1×1×1.2mm, TA=5min, 8 respiratory phases). Supplemental 2D real-time images were acquired in 9 different planes. Dynamic locations of the diaphragm dome and left kidney were measured from both 4D and 2D images. The same protocol was also performed on a MRI-compatible motion phantom where the motion was programmed with different amplitude (10–30mm) and frequency (3–10/min). Results: High resolution 4D-MRI were obtained successfully in 5 minutes. Quantitative motion measurements from 4D-MRI agree with the ones from 2D CINE (<5% error). The 4D images are free of the stitching artifacts and their near-isotropic resolution facilitates 3D visualization and segmentation of abdominal organs such as the liver, kidney and pancreas. Conclusion: Our preliminary studies demonstrated a novel ROCK 4D-MRI technique with true 3D k-space encoding and respiratory motion self-gating. The technique leads to high-resolution and artifacts-free 4D images for improved abdominal organ motion studies. K.S acknowledges funding support from NIH R01CA188300.

60 APPLIED LIFE SCIENCES↗

Technical Note: Suppression of artifacts arising from simultaneous cone-beam imaging and RF transponder tracking in prostate radiotherapy

Purpose: Artifacts in treatment-room cone-beam reconstructions have been observed at the authors' center when cone-beam acquisition is simultaneous with radio frequency (RF) transponder tracking using the Calypso 4D system (Calypso Medical, Seattle, WA). These artifacts manifest as CT-number modulations and increased CT-noise. The authors present a method for the suppression of the artifacts. Methods: The authors propose a three-stage postprocessing technique that can be applied to image volumes previously reconstructed by a cone-beam system. The stages are (1) segmentation of voxels into air, soft-tissue, and bone; (2) application of a 2D spatial-filter in the axial plane to the soft-tissue voxels; and (3) normalization to remove streaking along the axial-direction. The algorithm was tested on patient data acquired with Synergy XVI cone-beam CT systems (Elekta, Crawley, United Kingdom). Results: The computational demands of the suggested correction are small, taking less than 15 s per cone-beam reconstruction on a desktop PC. For a moderate loss of spatial-resolution, the artifacts are strongly suppressed and low-contrast visibility is improved. Conclusions: The correction technique proposed is fast and effective in removing the artifacts caused by simultaneous cone-beam imaging and RF-transponder tracking.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

WE-AB-207A-11: Respiratory Motion Guided 4DCBCT - A Step Towards Controlling 4DCBCT Image Quality

Purpose: We have developed a method, called respiratory motion guided 4DCBCT (RMG-4DCBCT), in which the gantry speed and projection frequency are varied in response to the patient’s real-time respiratory signal to eliminate streaking artifacts and to suppress duplicate projections in 4DCBCT images. In 2015, we realized RMG-4DCBCT on an Elekta Synergy linear accelerator with a mechanical relay to suppress projections and a potentiometer to adjust the gantry speed in response to the patient’s real-time respiratory signal. The aim of this study was to analyse the image quality to determine what can and cannot be controlled. Methods: Using RMG-4DCBCT, we acquired 40 (RMG-4DCBCT-40) and 60 (RMG-4DCBCT-60) equally spaced projections per respiratory phase of the CIRS dynamic thorax phantom with breathing periods from 2s to 8s and two breathing traces from lung cancer patients. The contrast to noise ratio (CNR) and edge response width (ERW) were used to compare image quality between RMG-4DCBCT and conventional 4DCBCT. Results: Regardless of the breathing period, for RMG-4DCBCT, the CNR is approximately 7 and 9 with RMG-4DCBCT-40 and RMG-4DCBCT-60 respectively. Conventional 4DCBCT has a CNR dropping from 20 down to 6 as the breathing period drops from 2s to 8s. With RMG-4DCBCT, the ERW, in the direction of phantom motion, ranges from 2.1mm to 2.5mm as the breathing period drops from 2s to 8s which compares to a higher range of 2.0mm to 2.5mm with conventional 4DCBCT. Images with similar quality to conventional 4DCBCT can be acquired with RMG-4DCBCT-40 which has a 70% reduction in imaging dose. Conclusion: The image contrast can be controlled with RMG-4DCBCT regardless of the patients breathing rate. However, although the image sharpness is better with RMG-4DCBCT, image sharpness has a small dependence on the breathing period; the accuracy of registration and segmentation will therefore vary with the patient’s breathing period. This project was supported by a National Health and Medical Research Council (NHMRC) project grant 1034060 and Cancer Australia grant number 1084566.

60 APPLIED LIFE SCIENCES↗

Scientific requirements for a Moderate-Resolution Imaging Spectrometer (MODIS) for EOS

The MODIS is an instrument planned for the sun-synchronous polar orbiting segment of the Space Station system. The radiometer is required to have 1 km resolution in terrestrial remote sensing applications. The monitoring program is targeted to last 10 yr in order to provide a sufficient database for discerning trends as opposed to natural variations. The study areas of interest include tropical deforestation, regrowth and areal distributions, acid rain effects on northern forests, desertification rates and locations, snow cover/albedo relationships and total biomass. MODIS will have 192 channels with 30 m spatial resolution and cover seven bands in the 3.5-12 microns interval for land viewing. Ocean studies will be carried out in 17 bands from 0.4-1.0 micron, and atmospheric scans will be performed over the land and ocean intervals at narrowband wavelengths (1.2 nm). Si detector arrays will be used and will be accompanied by an expected 600:1 SNR and produce data at a rate of 1.4-9.1 Mb/sec.

Barnes, W. L.↗

Earth Observing System instrument performance - Moderate Resolution Imaging Spectrometer case study

The requirements for characterization and calibration of instruments for the Earth Observing System space segment are defined in terms of tracking the Level 1 data for a sensor. The data must be traceable over the 15 year program, and must provide corrections for radiometric, spectral and geometric coefficients which capture the combined design and performance characteristics of the sensor. These requirements are then developed for MODIS-N to identify a set of techniques which can be used in the laboratory, spacecraft integrator's facility and in-orbit. The in-orbit of the so-called reflectance and radiance methods are reviewed, and the cross-calibrations available between HIRIS and MODIS through these techniques are listed. A scheme for the use of all available tools for instrument performance is provided as a Figure showing an estimated time interval over which each tool is expected to be useful. The final calibration strategy and plan must be developed through the understanding of a useful strategy for each of these candidate mechanisms.

Guenther, Bruce W.↗

Semiautomated segmentation of head and neck cancers in 18F-FDG PET scans: A just-enough-interaction approach

Purpose: The purpose of this work was to develop, validate, and compare a highly computer-aided method for the segmentation of hot lesions in head and neck 18F-FDG PET scans. Methods: A semiautomated segmentation method was developed, which transforms the segmentation problem into a graph-based optimization problem. For this purpose, a graph structure around a user-provided approximate lesion centerpoint is constructed and a suitable cost function is derived based on local image statistics. To handle frequently occurring situations that are ambiguous (e.g., lesions adjacent to each other versus lesion with inhomogeneous uptake), several segmentation modes are introduced that adapt the behavior of the base algorithm accordingly. In addition, the authors present approaches for the efficient interactive local and global refinement of initial segmentations that are based on the “just-enough-interaction” principle. For method validation, 60 PET/CT scans from 59 different subjects with 230 head and neck lesions were utilized. All patients had squamous cell carcinoma of the head and neck. A detailed comparison with the current clinically relevant standard manual segmentation approach was performed based on 2760 segmentations produced by three experts. Results: Segmentation accuracy measured by the Dice coefficient of the proposed semiautomated and standard manual segmentation approach was 0.766 and 0.764, respectively. This difference was not statistically significant (p = 0.2145). However, the intra- and interoperator standard deviations were significantly lower for the semiautomated method. In addition, the proposed method was found to be significantly faster and resulted in significantly higher intra- and interoperator segmentation agreement when compared to the manual segmentation approach. Conclusions: Lack of consistency in tumor definition is a critical barrier for radiation treatment targeting as well as for response assessment in clinical trials and in clinical oncology decision-making. The properties of the authors approach make it well suited for applications in image-guided radiation oncology, response assessment, or treatment outcome prediction.

60 APPLIED LIFE SCIENCES↗

Outer Wall Segmentation of Abdominal Aortic Aneurysm by Variable Neighborhood Search Through Intensity and Gradient Spaces

Aortic aneurysm segmentation remains a challenge. Manual segmentation is a time-consuming process which is not practical for routine use. To address this limitation, several automated segmentation techniques for aortic aneurysm have been developed, such as edge detection-based methods, partial differential equation methods, and graph partitioning methods. However, automatic segmentation of aortic aneurysm is difficult due to high pixel similarity to adjacent tissue and a lack of color information in the medical image, preventing previous work from being applicable to difficult cases. This paper uses uses a variable neighborhood search that alternates between intensity-based and gradient-based segmentation techniques. By alternating between intensity and gradient spaces, the search can escape from local optima of each space. The experimental results demonstrate that the proposed method outperforms the other existing segmentation methods in the literature, based on measurements of dice similarity coefficient and jaccard similarity coefficient at the pixel level. In addition, it is shown to perform well for cases that are difficult to segment.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Stellar Population Properties in the Stellar Streams around SPRC047

Abstract We have investigated the properties (e.g., age, metallicity) of the stellar populations of a ringlike tidal stellar stream (or streams) around the edge-on galaxy SPRC047 (z= 0.031) using spectral energy distribution (SED) fits to integrated broadband aperture flux densities. We used visual images in six different bands and Spitzer/IRAC 3.6μm data. We have attempted to derive best-fit stellar population parameters (metallicity, age) in three noncontiguous segments of the stream. Due to the very low surface brightness of the stream, we have performed a deconvolution with a Richardson–Lucy–type algorithm of the low spatial resolution 3.6μm IRAC image, thereby reducing the effect of the point-spread function aliasedemissionfrom the bright edge-on central galaxy at the locations of our three stream segments. Our SED fits that used several different star formation (SF) history priors, from an exponentially decaying SF burst to continuous SF, indicate that the age–metallicity–dust degeneracy is not resolved, most likely because of inadequate wavelength coverage and low signal-to-noise ratios of the low surface brightness features. We also discuss how future deep visual–near-infrared observations, combined with absolute flux calibration uncertainties at or below the 1% level, complemented by equally well absolute flux-calibrated observations in ultraviolet and mid-infrared bands, would improve the accuracy of broadband SED fitting results for low surface brightness targets, such as stellar streams around nearby galaxies that are not resolved into stars.

Astronomy & Astrophysics↗

Segmentation Of Multifrequency Complex-Amplitude SAR Data

Several mathematical models and associated algorithms implement method of segmenting multifrequency, highly speckled, high-resolution, complex-amplitude (amplitude and phase) synthetic-aperture-radar (SAR) digitized image into regions, within each of which radar backscattering characteristics are similar or homogeneous from place to place. Typically, each region represents different type of terrain or other surface; e.g., forest, agricultural land, sea ice, or water. Method of segmentation of SAR scene into regions is product of generalization, to multifrequency case, of single-frequency method described in "Algorithms for Segmentation of Complex-Amplitude SAR Data" (NPO-18524).

Rignot, Eric J.↗

Origins of Feeding Arteries of Hepatocellular Carcinoma Located Near the Umbilical Fissure of the Left Hepatic Lobe: Angiographic Evaluation

Purpose: To analyze the origins of the feeding arteries of hepatocellular carcinomas (HCCs) near the umbilical fissure of the left hepatic lobe. Methods: Twenty-eight HCCs with a mean {+-} SD tumor diameter of 3.4 {+-} 1.0 cm (range 1-4.4 cm) in contact with the right or left side of the umbilical fissure were treated by superselective transcatheter arterial chemoembolization (TACE). The origins of the tumor-feeding arteries were analyzed with arteriograms and computed tomography or cone-beam computed tomography images obtained during and 1 week after TACE.ResultsTwenty-one HCC lesions were located in segment 3 and seven were located in segment 4. Of 21 tumors in segment 3, 13 (61.9%) were supplied by the lateral inferior subsegmental artery (A3), three (14.3%) by the medial subsegmental artery (A4), three (14.3%) by both A4 and A3, one (4.8%) by a branch arising from the left lateral hepatic artery, and one (4.8%) by a branch of the right gastric artery. In particular, all tumor-feeding branches arising from A4 were the first branch of A4. Of seven tumors in segment 4, four (57.1%) were supplied by A4 and three (42.9%) by A3. In particular, all tumor-feeding branches arising from A3 were the first branch of A3. Conclusion: This study demonstrates crossover blood supply to HCC lesions located near the umbilical fissure, in addition to direct feeding from a separate branch. In particular, the first branch of the opposite subsegmental artery may feed tumors when crossover blood supply is present.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Detection of Cement Leakage After Vertebroplasty with a Non-Flat-Panel Angio Unit Compared to Multidetector Computed Tomography - An Ex Vivo Study

The purpose of this study was to investigate the detection of cement leakages after vertebroplasty using angiographic computed tomography (ACT) in a non-flat-panel angio unit compared to multidetector computed tomography (MDCT). Vertebroplasty was performed in 19 of 33 cadaver vertebrae (23 thoracic and 10 lumbar segments). In the angio suite, ACT (190{sup o}; 1.5{sup o} per image) was performed to obtain volumetric data. Another volumetric data set of the specimen was obtained by MDCT using a standard algorithm. Nine multiplanar reconstructions in standardized axial, coronal, and sagittal planes of every vertebra were generated from both data sets. Images were evaluated on the basis of a nominal scale with 18 criteria, comprising osseous properties (e.g., integrity of the end plate) and cement distribution (e.g., presence of intraspinal cement). MDCT images were regarded as gold standard and analyzed by two readers in a consensus mode. Rotational acquisitions were analyzed by six blinded readers. Results were correlated with the gold standard using Cohen's {kappa}-coefficient analysis. Furthermore, interobserver variability was calculated. Correlation with the gold standard ranged from no correlation (osseous margins of the neuroforamen, {kappa} = 0.008) to intermediate (trace of vertebroplasty canula; {kappa} = 0.615) for criteria referring to osseous morphology. However, there was an excellent correlation for those criteria referring to cement distribution, with {kappa} values ranging from 0.948 (paravertebral cement distribution) to 0.972 (intraspinal cement distribution). With a minimum of {kappa} = 0.768 ('good correlation') and a maximum of {kappa} = 0.91 ('excellent'), interobserver variability was low. In conclusion, ACT in an angio suite without a flat-panel detector depicts a cement leakage after vertebroplasty as well as MDCT. However, the method does not provide sufficient depiction of osseous morphology.

62 RADIOLOGY AND NUCLEAR MEDICINE↗