-
[show abstract]
[hide abstract]
ABSTRACT: The ability of micro-computed tomography (CT) to noninvasively evaluate allergic pulmonary inflammation in an experimental model was investigated. In addition, two image segmentation methods and the value of respiratory gating were investigated in the context of this model. Brown Norway rats were exposed to one of four doses of house dust mite (HDM) extract (0, 0.15, 15 or 150 microg) delivered intratracheally every 24 h for 10 days. CT scanning was performed at baseline and after several longitudinal HDM exposures. Both thoracic- and lung-segmentation methods yielded similar results when standardisation practices were employed. While tissue histology correlated well with CT images, cell counts from bronchoalveolar lavage depicted greater inflammation than did density measures from CT images. Evidence from representative CT slices and transaxial density distribution indicated that inflammation was primarily associated with major airways and extended into the periphery from these focal points. Respiratory gating demonstrated that images of the inspiratory state provided greater contrast of inflammatory processes. Lastly, decreases in tidal volumes indicated significant mechanical respiratory changes in animals exposed to both 15 and 150 microg. In summary, CT image segmentation can extract pertinent data on in vivo allergic airway/lung inflammation. Furthermore, respiratory gating provides additional contrast and insight into these quantification practices.
European Respiratory Journal 02/2009; 33(6):1437-47. · 5.89 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: In SPECT imaging, photon transport effects such as scatter, attenuation and septal penetration can negatively affect the quality of the reconstructed image and the accuracy of quantitation estimation. As such, it is useful to model these effects as carefully as possible during the image reconstruction process. Many of these effects can be included in Monte Carlo (MC) based image reconstruction using convolution-based forced detection (CFD). With CFD Monte Carlo (CFD-MC), often only the geometric response of the collimator is modeled, thereby making the assumption that the collimator materials are thick enough to completely absorb photons. However, in order to retain high collimator sensitivity and high spatial resolution, it is required that the septa be as thin as possible, thus resulting in a significant amount of septal penetration for high energy radionuclides. A method for modeling the effects of both collimator septal penetration and geometric response using ray tracing (RT) techniques has been performed and included into a CFD-MC program. Two look-up tables are pre-calculated based on the specific collimator parameters and radionuclides, and subsequently incorporated into the SIMIND MC program. One table consists of the cumulative septal thickness between any point on the collimator and the center location of the collimator. The other table presents the resultant collimator response for a point source at different distances from the collimator and for various energies. A series of RT simulations have been compared to experimental data for different radionuclides and collimators. Results of the RT technique matches experimental data of collimator response very well, producing correlation coefficients higher than 0.995. Reasonable values of the parameters in the lookup table and computation speed are discussed in order to achieve high accuracy while using minimal storage space for the look-up tables. In order to achieve noise-free projection images from MC, it -
-
is seen that the inclusion of the RT implementation for septal penetration increases the speed of the simulation by a factor of about 7,500 compared to the conventional SIMIND MC program.
IEEE Transactions on Nuclear Science 07/2008; · 1.45 Impact Factor
-
T H Farncombe
[show abstract]
[hide abstract]
ABSTRACT: In volumetric CT imaging of small animals, the breathing motion of the lungs during the image acquisition process results in inconsistent projection data being acquired. When reconstructed, these inconsistent data may produce images with reduced spatial resolution and image contrast. In order to minimize these effects, various approaches have been utilized to capture the respiratory signal of the animals under study and to only obtain CT data at specific moments in the respiratory cycle. These approaches typically utilize hardware-based physiological monitoring equipment in order to record the respiratory signal and either prospectively or retrospectively correlate this signal with acquired CT projection data. In this work, a new method of CT respiratory gating is described that does not rely on external physiological monitoring. Rather, determination of the respiratory phase of the animal is performed by postprocessing the acquired projection data. With this approach, any CT projection data can be respiratory gated with minimal effort. Validation of the method has been performed using a dynamic phantom and accuracy in tidal volumes determined to be within 16%. Rats and mice have been scanned and processed using the proposed method and compared to physiological-based measurement. With the proposed method, image quality is significantly improved in addition to providing quantitative information regarding tidal lung volumes.
Medical Physics 06/2008; 35(5):1785-92. · 2.83 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Monte Carlo (MC) is a well-utilized tool for simulating photon transport in single photon emission computed tomography (SPECT) due to its ability to accurately model physical processes of photon transport. As a consequence of this accuracy, it suffers from a relatively low detection efficiency and long computation time. One technique used to improve the speed of MC modeling is the effective and well-established variance reduction technique (VRT) known as forced detection (FD). With this method, photons are followed as they traverse the object under study but are then forced to travel in the direction of the detector surface, whereby they are detected at a single detector location. Another method, called convolution-based forced detection (CFD), is based on the fundamental idea of FD with the exception that detected photons are detected at multiple detector locations and determined with a distance-dependent blurring kernel. In order to further increase the speed of MC, a method named multiple projection convolution-based forced detection (MP-CFD) is presented. Rather than forcing photons to hit a single detector, the MP-CFD method follows the photon transport through the object but then, at each scatter site, forces the photon to interact with a number of detectors at a variety of angles surrounding the object. This way, it is possible to simulate all the projection images of a SPECT simulation in parallel, rather than as independent projections. The result of this is vastly improved simulation time as much of the computation load of simulating photon transport through the object is done only once for all projection angles. The results of the proposed MP-CFD method agrees well with the experimental data in measurements of point spread function (PSF), producing a correlation coefficient (r<sup>2</sup>) of 0.99 compared to experimental data. The speed of MP-CFD is shown to be about 60 times faster than a regular forced detection MC program with similar results.
IEEE Transactions on Nuclear Science 03/2008; · 1.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The purpose of this study was to investigate the importance of collimator-detector response compensation in single photon emission computed tomography (SPECT). The compensation methods included in this study addressed three important degrading factors, which are attenuation, scatter and collimator-detector response (CDR). Geometric response (GR) and septal penetration (SP) are two most important components of collimator-detector response. The compensation work has been divided into two general categories: one is GR compensation based reconstruction (GR-RECON), which includes attenuation, scatter and geometric response compensations, and the other one is GR-SP compensation based reconstruction (GR-SP-RECON), which consists of attenuation, scatter geometric response and septal penetration compensations. Ordered-subset expectation maximization (OS-EM) method is applied as the reconstruction framework. SIMIND Monte Carlo (MC) code is incorporated in the forward projection. Multiple projection sampling convolution-based forced detection (MP-CFD) is used here to accelerate the MC code. The convolution kernels in CFD are generated by the ray-tracing (RT) method. In order to further increase the convergence speed, the models of attenuation and collimator-detector response are included in the backprojection step. In order to assess the quantitative accuracy of the collimator-detector response compensation, the reconstruction images of cylinder, four different size spheres in various medium, and NURBS-based cardiac-torso (NCAT) phantom are evaluated using 1-131 and high energy general resolution (HEGR) collimators. The reconstruction images including GR compensation appear very obvious collimator dependent Gibbs ringing artifact. The quantitative estimation of the spheres and NURBS-based cardiac-torso (NCAT) phantom has denoted the high accuracy of GR-SP-RECON, whereas, a lot of quantitative activity will be mistakenly estimated in the background using GR-RECON.
Nuclear Science Symposium Conference Record, 2007. NSS '07. IEEE; 12/2007
-
[show abstract]
[hide abstract]
ABSTRACT: In single photon emission computed tomography (SPECT) imaging, photon transport effects such as scatter, attenuation, and septal penetration can negatively affect the quality of reconstructed image and quantitation. As such, it is useful to model these factors as careful as possible during image reconstruction in order to reduce the impact. Many of these effects can be included in Monte Carlo (MC) based image reconstruction using convolution-based forced detection (CFD-MC). However, in CFD-MC, often only the geometric response of collimators is modeled, thereby make the assumption that the collimator materials are thick enough to completely absorb photons traveling through them. However, in order to retain high collimator sensitivity and high spatial resolution, it is required that the septa be as thin as possible, thus resulting in a significant amount septal penetration for high energy radionuclides. A method for modeling the effects of collimator response including both collimator septal penetration and geometric response using ray tracing (RT) techniques has been performed and included into a CFD-MC program. Two lookup tables are precalculated based on the collimator parameters and radionu-clide, and subsequently incorporated into SIMIND MC program. One table consists of the cumulative septal thickness between any point on the collimator and the center, while the other table represents the resultant collimator response for a point source at differing distances from the collimator and for various energies. A series of RT simulations have been compared to experimental data for different radionuclides and collimators. The results of RT techniques matches the experimental data very well, producing correlation coefficients higher than 0.995. In order to achieve noise-free projection images from MC, it has been seen that the inclusion of the RT implementation for collimator response increases the speed of simulation by a factor of 7,500 compared to the conventional fo-
rced detection (FD) SIMIND MC program.
Nuclear Science Symposium Conference Record, 2006. IEEE; 12/2006
-
[show abstract]
[hide abstract]
ABSTRACT: In pelvic bone SPECT using Tc-99m labeled compounds, physical effects such as nonhomogeneous attenuation and the accumulation of activity into the bladder during the data acquisition process can often result in data inconsistencies. With filtered backprojection (FBP) reconstruction, this may result in streak artifacts. Various methods for attenuation compensation can be employed to account for some of these streaks, but if the rate of uptake of tracer in to the bladder is sufficient, the resultant streaks may be significant enough to impair lesion detection. We have investigated various reconstruction methods in an effort to reduce these artifacts. In order to assess the impact of inconsistent bladder activities on the detection of pelvic lesions, SPECT imaging was simulated using the Zubal voxelized phantom, for cases of both a static bladder activity and a changing activity distribution. Reconstructions were performed using FBP, ordered subset-expectation maximization (OSEM) and dynamic expectation maximization (dEM) and was assessed for lesion detectability using a channelized, nonprewhitening (CNPW) numerical observer model. This observer model was used to optimize reconstruction strategies for a human LROC observer study. The human LROC observer study was performed in order to assess the various reconstruction methods in terms of lesion detectability. Three human observers were used in this test. The results of this test indicate that FBP performs significantly worse than static OSEM iterative reconstruction with attenuation correction when assessed using the area under the LROC curve (maximal A<sub>LROC</sub>=0.47 for FBP versus 0.71 for OSEM). Visually, the dEM algorithm produces images with slightly reduced streak artifacts compared to OSEM, but this improvement was not reflected in significantly improved A<sub>LROC</sub> values. In fact, when assessed using human LROC methodology, detectability actually decreased slightly when using dEM (maximal A<sub>LROC<-
-
/sub>=0.71 for OSEM versus 0.66 for dEM), although this reduction was not determined to be statistically significant. It is possible that the slightly reduced performance of the dEM algorithm may be due, in part, to not performing an optimization in the number of reconstruction iterations as was performed for the OSEM method
IEEE Transactions on Nuclear Science 11/2006; · 1.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Ga-67 SPECT studies are especially useful in the pretreatment staging and post-treatment follow-up of patients with Hodgkin's and non-Hodgkin's lymphoma. Ga-67 is however a particularly challenging radionuclide for imaging. Patient specific background variability, which manifests itself as structured noise, can further impact lesion detection accuracy. A number of MCAT phantom studies with simulated lesions and idealized source distributions have been done to study the impact of compensation strategies on lesion detection accuracy. However, to more accurately assess the impact of various correction strategies on lesion detection a study employing actual clinical images with true clinical distributions is of interest. The approach we chose for conducting such an investigation was performing LROC studies employing hybrid images. Hybrid images are normal Ga-67 studies with their projection data modified by the addition of Monte Carlo simulated lesions. Our datasets consist of clinically normal Ga-67 SPECT/CT acquisitions obtained using the GE-VG dual detector SPECT/CT camera. After determining a target image contrast using human observers, we conducted pilot LROC studies to determine the optimal parameters for the reconstruction methods using human observers. Herein we report on the optimization for the forthcoming comparison of attenuation compensation, scatter compensation, and detector resolution compensation strategies used with the RBI reconstruction method and FBP reconstruction.
Nuclear Science Symposium Conference Record, 2005 IEEE; 11/2005
-
[show abstract]
[hide abstract]
ABSTRACT: In pelvic bone SPECT using Tc-99m labelled compounds, the accumulation of activity into the bladder during the data acquisition process often results in data inconsistencies which, when reconstructed with filtered backprojection, results in streak artifacts. If the uptake rate is sufficient, these streaks may be significant enough to impair lesion detection. We have investigated various reconstruction methods in an effort to reduce this artifact. Pelvic SPECT imaging was simulated using the Zubal voxelized phantom, with provisions for a changing activity distribution within the bladder. Reconstructions were performed using filtered backprojection, ordered subset-expectation maximization and dynamic expectation maximization. Each method was first optimized for postreconstruction smoothing parameters using a channelized, nonprewhitening (CNPW) numerical observer model. The numerical observer used is based on human observer LROC methodology whereby both a likely lesion location and a confidence rating is supplied by the observer for each image. Based on the results of the CNPW observer, a human LROC observer study was performed in order to assess the various reconstruction methods in terms of lesion detectability. Three human observers were used in this test. The results of this test indicate that filtered backprojection performs significantly worse than static OSEM iterative reconstruction with attenuation correction when assessed using the area under the LROC curve (A<sub>LROC</sub>=0.47 vs 0.71). Results comparing OSEM with dEM indicate that the dEM algorithm is able to further reduce streak artifacts compared to OSEM, but this improvement was not reflected in improved A<sub>LROC</sub> scores. In fact, detectability actually decreased slightly when using dEM (A<sub>LROC</sub>=0.71 vs 0.66), although this reduction was not seen to be statistically significant. It is possible that the slightly reduced performance of the dEM algorithm may be due, in part, to not performing an optimization in the number of reconstruction iterations as was performed for the OSEM method.
Nuclear Science Symposium Conference Record, 2003 IEEE; 11/2003
-
[show abstract]
[hide abstract]
ABSTRACT: Typical myocardial SPECT imaging using Tc99m labelled sestamibi or PET imaging using N-13 labelled ammonia assumes that the amount of tracer uptake within the myocardium correlates directly with myocardial blood flow. However, this assumption has been seen to only be accurate over a limited range of blood flow rates. The uptake of alternative ligands such as Tc99m labelled Tebororime or O-15 labelled H<sub>2</sub>O have been seen to better correlate with actual blood flow rates. Difficulties arise however in imaging these pharmaceuticals due to the relatively fast washout of tracer from the myocardium, thereby making simultaneous measurements of perfusion and wall motion difficult. An alternative image reconstruction strategy is presented in which both myocardial blood flow and myocardial wall motion can be determined in a single study from very low count projection data. The proposed method uses the dynamic expectation maximization (dEM) algorithm in combination with temporal filtering across cardiac gates in order to improve reconstructed image quality. Data acquisition can be performed from any tomographic acquisition type ranging from a single head, slow rotation SPECT acquisition, to a fixed ring, PET geometry. A computer simulation has been performed in order to evaluate the image quality produced when using this imaging approach with a variety of SPECT acquisition parameters.
Nuclear Science Symposium Conference Record, 2003 IEEE; 11/2003
-
[show abstract]
[hide abstract]
ABSTRACT: Ga-67 citrate SPECT imaging is often used for oncological studies in order to diagnose or stage patient lymphomas. Because the decay of Ga-67 involves multiple emission energies, it is possible that many down-scattered photons will be present in photopeak acquisition data. We have previously shown through human observer LROC studies, that the inclusion of these scattered photons significantly degrades lesion detectability in simulations. We have investigated the use of six different scatter compensation methods representing different strategies. These consist of i) perfect scatter rejection, ii) no scatter compensation, iii) ideal scatter compensation, iv) triple energy window estimation, v) effective scatter source estimation, and vi) post-reconstruction scatter subtraction. Each method has first been optimized using a channelized hotelling numerical observer, then ranked through the use of a human LROC study and by using a newly devised LROC numerical observer. Both human LROC and LROC numerical observer results indicate that both TEW and ESSE scatter compensation methods are able to improve lesion detectability over no compensation, but fail to achieve similar detectability to using perfect scatter rejection. Excellent agreement between the LROC numerical observer and human LROC studies indicate that the LROC observer may be good predictor of human performance in Ga-67 SPECT.
Nuclear Science Symposium Conference Record, 2002 IEEE; 12/2002
-
[show abstract]
[hide abstract]
ABSTRACT: Results from localization ROC (LROC) observer studies evaluating a penalized-likelihood SPECT reconstruction algorithm are presented. Both human and numerical observers were employed for the task of Ga-67 tumor detection in simulated chest images. Projection data from a mathematical phantom were generated with the SIMIND Monte Carlo program. Reconstructions using De Pierro's modified-EM algorithm with the 3D quadratic and Huber penalty functions were Performed. Iterative corrections for nonuniform attenuation, collimator response, and in-patient scatter were included. The LROC studies with these images provided an initial validation of a channelized nonprewhitening (CNPW) numerical observer that explicitly performs the tumor search task, and at the same time demonstrated the usefulness of frequency-selective channels for modeling human detection-task performance. Results drawn from the CNPW observer indicate that for the range of iterations tested, relatively low levels of noise regularization are desirable in the penalized-likelihood images.
Nuclear Science Symposium Conference Record, 2002 IEEE; 12/2002
-
[show abstract]
[hide abstract]
ABSTRACT: We have previously modeled the respiratory motion of solitary pulmonary nodules (SPN) based on the change of location of anatomic structures within the lungs identified on breath-held CT images of volunteers acquired at two different stages of respiration. The goal of this investigation is to make use of this modeling to investigate the impact of respiratory motion on the detection of solitary pulmonary nodules with single photon emission computed tomographic (SPECT) imaging using Tc-99m labeled NeoTect. To do this, end-expiration and respiration-averaged source and attenuation maps were created from the NCAT phantom and input to the SIMIND Monte Carlo package.
Nuclear Science Symposium Conference Record, 2002 IEEE; 12/2002
-
[show abstract]
[hide abstract]
ABSTRACT: In this work, we investigated the potential for improving image quality and quantitative accuracy of dynamic Altropane SPECT scans. We evaluated fully-4D reconstruction techniques as well as conventional frame-by-frame reconstruction approaches and their ability to provide consistent striatal binding ratios for both normal and patients with attention deficit hyperactivity disorder (ADHD). We also used the Zubal brain phantom to compare the different reconstruction strategies. A 3-headed Picker-3000 SPECT system fitted with low-energy ultrahigh resolution fan-beam collimators acquired 10 frames of dynamic SPECT data in 4-min acquisitions over a period of 40 minutes. Each set of 128 × 128 projections were acquired at 120 angles over 360°. The dynamic data sequence was using two conventional frame-by-frame reconstruction methods namely, filtered back-projection (FBP) with multiplicative Chang attenuation correction (AC) and re-scaled block-iterative expectation-maximization (RBI-EM) with uniform AC. We also considered two fully-4D reconstructions methods, the KL-EM that exploits the Karbunen-Loeve transform (or principal component analysis) and a dynamic expectation maximization algorithm or DSPECT that incorporates dynamic inequality constraints in the reconstructed object voxel activity over time. Both 4D methods were implemented with uniform AC. From the results, it is clear that both the 4D methods are more robust in handling the low-count dynamic data. However, derived parameters such as the binding ratio might be more insensitive to noise, although in the patient study FBP binding ratios differed greatly from that of the other reconstruction strategies.
Nuclear Science Symposium Conference Record, 2002 IEEE; 12/2002
-
[show abstract]
[hide abstract]
ABSTRACT: In this paper, we investigate the use of a numerical observer to optimize ordered-subset expectation maximization (OSEM) reconstructions for the detection of coronary artery disease (CAD). The parameters optimized were the iteration number and the full-width at half-maximum of three-dimensional Gaussian postfiltering. The numerical observer employed in the optimization was the channelized Hotelling observer (CHO). The CHO had been used previously to rank tumor detection accuracy for different reconstruction strategies in Ga-67 images, showing good agreement with the rankings of human observers. The intent of this paper was to determine if this CHO could also be employed for the detection of CAD. Results indicate that when grayscale (quantized) images are used, the CHO optimization results correlate well with human observers. On the other hand, when the CHO was used with floating-point images, it provided very good detection performance even when the images were excessively filtered. This result was at odds with the human-observer performance which showed a decrease in detection accuracy with highly smoothed images. This reflects the need to better model the detection task of the human observers who usually view and rank grayscale images and by appropriately modeling the image noise that quantization introduces, we show that the CHO can better match human-observer detection performance.
IEEE Transactions on Nuclear Science 11/2002; · 1.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: During <sup>67</sup>Ga citrate SPECT imaging, photon downscatter will occur from higher energy photons into lower energy acquisition windows thus possibly adversely affecting reconstructed image quality. With these additional scattered photons present in projection data, the effect of using more complex reconstruction strategies such as three-dimensional detector response compensation (3-D DRC) and attenuation correction (AC) is unclear. Using a combination of numerical channelized hotelling observers (CHO) and human localization receiver operating characteristics (LROC) studies it has been found that maximum lesion detectability occurs when projection data is reconstructed using two iterations of the rescaled block iterative (RBI) algorithm with both 3-D DRC and patient specific AC, followed by a postreconstruction low-pass 3-D Gaussian filtering with a FWHM of ≈1 cm. These parameters deviate from optimal reconstruction parameters for primary photon-only projection data which finds that maximum lesion detection occurs after 4 RBI iterations with the same 3-D DRC, AC, and postfiltering. As expected, it is also observed that decreased tumor detectability results when scattered photons are present in <sup>67</sup>Ga projection data compared to primary photon only reconstructions (reducing maximum A<sub>L</sub> from 0.79 to 0.58). This decrease has been found to be statistically significant in both human LROC and numerical observer studies, suggesting the need for scatter compensation during <sup>67</sup>Ga citrate imaging.
IEEE Transactions on Nuclear Science 11/2002; · 1.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Scatter can be a confounding problem in the detection of perfusion defects in myocardial perfusion imaging. The goal of this investigation was to compare practical methods of scatter compensation representing different classes of compensation strategies. The methods investigated were: (1) modification of the energy window; (2) use of effective values of attenuation coefficients; (3) estimating the scatter distribution from energy spectrum information; and (4) estimating the scatter distribution in the spatial domain using the attenuation maps and scatter kernels. Monte Carlo simulated projections from two MCAT source distributions were used to investigate the impact of the various methods on residual scatter and myocardial uniformity. The investigation showed that each of the methods reduce the scatter significantly; however, the spatial domain method gave the least variation with the different source distributions and the best uniformity.
IEEE Transactions on Nuclear Science 11/2002; · 1.45 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: We have modeled the respiratory motion of solitary pulmonary nodules (SPN) based on the change of location of anatomic structures in the lungs determined from breath-held CT images of volunteers acquired at two different stages of respiration. The resulting 3D data was visualized in summed coronal, sagittal, and transaxial projections by drawing an arrow showing the direction and magnitude of motion during inspiration. The findings reveal that there is a three-dimensional component of the motion of structures within the lung and that the magnitude and direction vary based on the specific anatomic region of the lung. This information on respiratory motion within the lungs was combined with the NCAT phantom to allow the creation of source and attenuation maps for investigating the impact of respiratory motion in single photon emission computed tomographic (SPECT) imaging for SPN with Tc-99m labeled NeoTect. With the source and attenuation distribution thus defined, the SIMIND Monte Carlo program is used to produce SPECT projection images for the normal background Tc-99m NeoTect distribution and each of the tumors separately. These projections can be combined according to the desired tumor contrast and employed to investigate the impact of respiratory motion of SPN detection using human-observer localization receiver operating characteristics (LROC) studies.
Nuclear Science Symposium Conference Record, 2001 IEEE; 12/2001
-
[show abstract]
[hide abstract]
ABSTRACT: Psychophysical studies were used to investigate the effect of number of projection angles Φ on images reconstructed from simultaneous emission-transmission scans of fixed duration. The specific task was detection and localization of Ga-67 tumors in simulated images of the thorax, and a Tc-99m transmission source was modeled in the data acquisition. Reconstructions with the rescaled block-iterative (RBI) expectation-maximization algorithm included corrections for nonuniform attenuation and collimator response, but not scatter. A combination of channelized Hotelling observer (CHO) ROC and human observer localization ROC (LROC) studies were conducted. The CHO was used to optimize the number of RBI iterations and the post-reconstruction filtering level for reconstruction strategies based on Φ∈ {20, 30, 40, 60, 90, 120} angles. The LROC study compared these optimized strategies. No significant differences in human detection-task performance were found for Φ⩾30. This finding is attributed largely to collimator effects, the presence of uncorrected scatter in the data, and the post-filtering of the reconstructed images
Nuclear Science Symposium Conference Record, 2001 IEEE; 02/2001
-
[show abstract]
[hide abstract]
ABSTRACT: In this work we investigate the use of a numerical observer to
optimize ordered-subset expectation maximization (OSEM) reconstructions
for the detection of coronary artery disease (CAD). The parameters
optimized are the iteration number and the amount of 3-D Gaussian
postfiltering. The numerical observer employed in the optimization was
the channelized Hotelling observer (CHO). We used channels as previously
determined to rank tumor detection accuracy of different reconstruction
strategies, that showed good agreement with the rankings of human
observers. The intent of this work was to determine if this CHO could
also be employed for the detection of CAD. Results indicate that when
quantized images are used, the CHO optimization results correlates well
with human observers. However, when using floating-point images, the CHO
maintains detection performance even when images are excessively
filtered, a result that is at odds with our human observer study.
Modeling the additive internal noise that quantization introduces,
requires further investigation
Nuclear Science Symposium Conference Record, 2001 IEEE; 02/2001
