Xi-de Zhao

Quantitative cardiac SPECT reconstruction with reduced image degradation due to patient anatomy

Patient anatomy has complicated effects on cardiac SPECT images. The authors investigated reconstruction methods which substantially reduced these effects for improved image quality. A 3D mathematical cardiac-torso (MCAT) phantom which models the anatomical structures in the thorax region were used in the study. The phantom was modified to simulate variations in patient anatomy including regions of natural thinning along the myocardium, body size, diaphragmatic shape, gender, and size and shape of breasts for female patients. Distributions of attenuation coefficients and Tl-201 uptake in different organs in a normal patient were also simulated. Emission projection data were generated from the phantoms including effects of attenuation and detector response. The authors have observed the attenuation-induced artifacts caused by patient anatomy in the conventional FBP reconstructed images. Accurate attenuation compensation using iterative reconstruction algorithms and attenuation maps substantially reduced the image artifacts and improved quantitative accuracy. The authors conclude that reconstruction methods which accurately compensate for nonuniform attenuation can substantially reduce image degradation caused by variations in patient anatomy in cardiac SPECT. >

Comparison between ML-EM and WLS-CG algorithms for SPECT image reconstruction

The properties of the maximum likelihood with expectation maximization (ML-EM) and the weighted least squares with conjugate gradient (WLS-CG) algorithms for use in compensation for attenuation and detector response in cardiac SPECT imaging were studied. A realistic phantom, derived from a patient X-ray CT study to simulate /sup 201/Tl SPECT data, was used in the investigation. In general, the convergence rate of the WLS-CG algorithm is about ten times that of the ML-EM algorithm. Also, the WLS-CG exhibits a faster increase in image noise at large iteration numbers than the ML-EM algorithm. >

Quantitative single-photon emission computed tomography: Basic and clinical considerations

Although quantitative single-photon emission computed tomography (SPECT) has been the goal of much research effort for a number of years, only recently has it received wide interest, especially for clinical applications. It has been increasingly recognized that the achievement of quantitative SPECT will increase the accuracy of measurements, such as dimensions of specific regions of interest, absolute amount of radioactivity, and dosimetry calculations, and substantially reduce reconstruction image artifacts and distortions, thus, greatly improving clinical diagnosis. This article provides a review of the definition of terms, major factors affecting SPECT quantitation and their degrading effects on SPECT image quality, and methods to compensate for these effects. Compensation methods include those that make certain approximations for ease of implementation and those that provide more accurate compensation by modeling the imaging process more exactly, usually at the cost of increased complexity and computational requirements. Different reconstruction and compensation methods may be compared through the use of phantom cardiac and brain SPECT studies. The clinical efficacy of the methods may be demonstrated by applying them to a clinical thallium-201 myocardial perfusion SPECT study. The results clearly demonstrate that, by modeling the imaging process and/or image degrading factors three-dimensionally, quantitative reconstruction and compensation methods provide the best image quality and quantitative accuracy. Important research efforts and developmental work being conducted currently to bring quantitative SPECT into routine clinical use are also discussed.

Characteristics of Reconstructed Point Response in three-Dimensional Spatially Variant Detector Response Compensation in SPECT

We investigated the characteristics of the reconstructed point response in SPECT images obtained with and without compensation of the spatially variant collimator-detector response in 3D. The 3D compensation was achieved by modeling the 3D collimatordetector response function in the projector/backprojector pair of a WLS-CG iterative reconstruction algorithm. Using two 3D numerical phantoms, we studied the effects of collimation and data sampling. Total resolution recovery can be achieved when the reconstruction voxel size is small compared with that of the object. Recovery for smaller objects is limited by the spatial resolution of the collimator and the voxel size used in image reconstruction and requires more number of iterations than larger objects. The reconstructed 3D point response is asymmetric with the best resolution in the longitudinal direction and worst in the radial direction.

Imaging characteristics of scintimammography using parallel-hole and pinhole collimators

The purpose of the study is to investigate the imaging characteristics of scintimammography (SM) using parallel-hole (PR) and pinhole (PN) collimators in a clinical setting. Experimental data were acquired from a phantom that models the breast with small lesions using a low energy high resolution (LEHR) PR and a PN collimator. At close distances, the PN collimator provides better spatial resolution and higher detection efficiency than the PR collimator, at the expense of a smaller field-of-view (FOV). Detection of small breast lesions can be further enhanced by noise smoothing, field uniformity correction, scatter subtraction and resolution recovery filtering. Monte Carlo (MC) simulation data were generated from a realistic 3D MCAT phantom that models the Tc-99m sestamibi uptake and attenuation distributions in an average female patient. The scatter to primary ratio (S/P) decreases from the base to the tip of the breast. For PR collimation, S/P is higher in the left than right breast due to scatter of photons from the heart. For PN collimation, S/P is highest at the base and lowest at the tip of the breast as compared to PR collimation. It is higher in the right than left breast due to contributions from organ uptakes in the body. Results from the study add to understanding of the imaging characteristics of SM using PR and PN collimators and assist in the design of data acquisition and image processing methods to enhance the detection of breast lesions using SM.

Reconstruction methods for quantitative brain SPECT

The reconstruction methods currently available for use in quantitative brain SPECT (single photon emission computed tomography) imaging were studied. The conventional reconstruction methods are based on the filtered backprojection algorithm. Assuming uniform attenuation, the Chang algorithm is effective in attenuation compensation. The collimator response can be compensated for by a restoration filter with an average collimator response function. Scatter compensation can be achieved by the dual window subtraction method or by incorporating an average scatter response function in the restoration filter. The dual window subtraction method is an approximate but efficient means of providing projection data which are compensated for scatter. The iteration reconstruction method can be applied to these data for addition compensation for attenuation and the spatially variant collimator response. The latter is accomplished by incorporating the exact model of attenuation and detector response function in the projector/backprojector of the iterative algorithm. A 3-D brain phantom was used in the evaluation study. The images obtained from various reconstruction methods were compared for quantitative accuracy with respect to the known 3-D phantom activity distribution.<<ETX>>

Compensation for the response function of medium energy collimator in /sup 67/Ga planar and SPECT imaging

This preliminary study is to improve /sup 67/Ga planar and SPECT images by compensating for the response function of a medium energy (ME) collimator. The point response functions (PRFs) of a GE ME collimator for the 93, 185 and 300 keV photons of /sup 67/Ga at 5, 10, 15 and 20 cm from the collimator face were experimentally determined. For small pixel sizes, the PRFs for all distances showed hole pattern effects. Images from the higher energy photopeaks showed increased penetration fractions. For a specified source distance, a Butterworth filter can be designed to eliminate the collimator hole pattern with minimal degradation of the spatial resolution. Compensation for the distance-dependent collimator-detector response was accomplished using iterative reconstruction methods. To evaluate the compensation methods, planar images and SPECT projection data were acquired from a phantom consisting of 3 hot spheres with diameters of 1, 1.3 and 1.6 cm inside a cylindrical phantom. The iterative reconstruction-based compensation method provided improved resolution and fewer artifacts than in images reconstructed with filtered backprojection. The authors conclude that degradation caused by ME collimator in /sup 67/Ga imaging can be effectively compensated for using these techniques.

Practical iterative reconstruction methods for quantitative cardiac SPECT image reconstruction

Summary form only given. The authors investigate the practical implementation of iterative reconstruction methods to improve the quality and quantitative accuracy of cardiac single photon emission computed tomography (SPECT) images over the filtered backpropagation (FB) algorithm. A cardiac-chest phantom based on a patient CT scan and Tl-201 uptake data was used in the study. Projection data were generated which included the effects of nonuniform attenuation detector response, scatter and noise. Use was made of the iterative Chang algorithm with nonuniform attenuation compensation only, the iterative maximum likelihood-expectation maximization (ML-EM) and WLS-CG methods, and variations of the WLS-CG methods, including the use of the FB or Chang reconstructed image as initial estimate and filtering to suppress noise amplifications at high iteration number. The mean-square-error over the region of the heart and the normalized standard deviation over a background region were used to evaluate the different reconstruction methods. It was found that, using the Chang reconstructed image as the initial estimate in the WLS-CG method together with filtering between iterations, cardiac SPECT images with good quality and quantitative accuracy can be obtained in a few iterations. >

Evaluation of collimator-detector response compensation in tumor SPECT using medium- and high-energy collimators

The goal of the study is to evaluate collimator-detector response (CDR) compensation methods that apply to tumor SPECT imaging using medium-energy (ME) and high-energy (HE) collimators. The compensation method involves accurate models of the geometric, penetration and scatter components of ME and HE collimators based on Monte Carlo simulations verified by experimental measurements. The models were used in iterative OS-EM reconstruction methods for CDR compensation. In experimental studies, a cylindrical phantom consisting of spheres with different sizes and filled with Ga-67, In-111 or I-131 was used. Projection data were acquired using ME and HE collimators designed for use with the radionuclides. Patient data included those from Ga-67 citrate, In-111 octreotide, and I-131 tumor studies. The FBP without compensation and the iterative OS-EM with different models of the CDR were used in image reconstruction. Results from phantom studies showed asymptotic decrease of the reconstructed sphere sizes as a function of iteration number. The full CDR model that included the geometric, penetration and scatter components provided the best results. Drastic improvements in clinical image quality were found using the full CDR model. It is concluded that full CDR compensation provides substantial improvements in image quality and quantitative accuracy in tumor SPECT.