Grant Gullberg

Kinetic parameter estimation from dynamic cardiac patient SPECT projection measurements

The estimation of kinetic parameters directly from projection data is potentially useful for clinical dynamic cardiac SPECT studies, particularly those using a single detector system or body contouring orbits with a multi-detector system. A dynamic image sequence reconstructed from the inconsistent projections acquired by a relatively slowly rotating gantry can contain artifacts that lead to biases in kinetic parameters estimated from time-activity curves generated by overlaying regions of interest on the images. Using simulated data we have shown that unbiased kinetic parameter estimates can be obtained directly from the projections. Here we present results of a /sup 99m/Tc-teboroxime patient study where the regions of the left ventricular myocardium, blood pool, liver, and background tissue were determined by automatically segmenting a dynamic image sequence reconstructed from the inconsistent projection data. A spatial model for the projections was then created and one compartment kinetic model parameters for the myocardium and liver were estimated directly from the projections.

Tracking cardiac twist in gated PET imagery

Traditional clinical techniques for imaging heart function, like gated SPECT and PET, have shown that the motion of the left ventricle during the cardiac cycle can be described by relatively global parameters such as ejection fraction, longitudinal shortening, radial contraction and wall thickening. Despite the poor spatial and temporal resolution of these datasets, a number of techniques are now available which can accurately measure these global parameters. Recently, phase contrast and tagged MRI techniques have shown that in addition to the global components, the left ventricle also exhibits considerable local twist during the cardiac cycle. We present a validation study to investigate whether this torsion can be detected and tracked in gated cardiac PET imagery. Data were acquired with 40 msec gating intervals from six normal subjects using retrospectively gated list mode PET and tagged MRI. The MRI data were analyzed using the Findtags software package and custom software to determine the ground truth motion fields. A four-dimensional deformable motion algorithm based on optical flow was then used to estimate the motion from the PET dataset. Results indicate that though the estimation algorithm can quite accurately determine the component of motion normal to the ventricular surfaces, the torsion component is considerably more difficult to track.

Estimation of mechanical properties from gated SPECT and cine MRI data using a finite-element mechanical model of the left ventricle

A significant challenge in diagnosing cardiac disease is determining the viability of myocardial tissue when evaluating the prognosis of vascular bypass surgery. A finite element mechanical model of the left ventricular myocardium was developed to evaluate myocardial deformation, which is an important indicator of viable myocardial tissue. The model of the heart muscle mechanics was derived from the passive and active behavior of skeletal muscle, which is considered to be a quasi-incompressible transversely isotropic hyperelastic material of a specified helical fiber structure configuration. Contraction of the myocardium was replicated by simulating active contractions along the helical fibers, then solving (quasi-statically) for the associated boundary valued problem at a sequence of time steps between end-diastole and end-systole of the cardiac cycle. At each time step the finite element software package ABAQUS was used to determine the deformation of the left ventricle, which was loaded by intra-ventricular pressure. A cylindrical model of the left ventricle was developed under both passive loading and active contraction. Some parameters that describe the material properties of the myocardium were estimated by fitting the motion of the cylindrical model to gated SPECT and cine MRI data. Results from the finite element analysis were compared to those from a mathematical cylindrical model. In the future, more realistic meshes derived from imaging data will be used to perform the finite element analysis. The deformation of the mechanical model will be fitted to a complete strain map from tagged MRI or image warping. Finally, it is proposed to introduce electrical propagation into the finite element model.

The effect of truncation on very small cardiac SPECT camerasystems

Background: The limited transaxial field-of-view (FOV) of avery small cardiac SPECT camera system causes view-dependent truncationof the projection of structures exterior to, but near the heart. Basictomographic principles suggest that the reconstruction of non-attenuatedtruncated data gives a distortion-free image in the interior of thetruncated region, but the DC term of the Fourier spectrum of thereconstructed image is incorrect, meaning that the intensity scale of thereconstruction is inaccurate. The purpose of this study was tocharacterize the reconstructed image artifacts from truncated data, andto quantify their effects on the measurement of tracer uptake in themyocardial. Particular attention was given to instances where the heartwall is close to hot structures (structures of high activity uptake).Methods: The MCAT phantom was used to simulate a 2D slice of the heartregion. Truncated and non-truncated projections were formed both with andwithout attenuation. The reconstructions were analyzed for artifacts inthe myocardium caused by truncation, and for the effect that attenuationhas relative to increasing those artifacts. Results: The inaccuracy dueto truncation is primarily caused by an incorrect DC component. Forvisualizing theleft ventricular wall, this error is not worse than theeffect of attenuation. The addition of a small hot bowel-like structurenear the left ventricle causes few changes in counts onmorexa0» the wall. Largerartifacts due to the truncation are located at the boundary of thetruncation and can be eliminated by sinogram interpolation. Finally,algebraic reconstruction methods are shown to give better reconstructionresults than an analytical filtered back-projection reconstructionalgorithm. Conclusion: Small inaccuracies in reconstructed images fromsmall FOV camera systems should have little effect on clinicalinterpretation. However, changes in the degree of inaccuracy in countsfrom slice toslice are due to changes in the truncated structures. Thesecan result in a visual 3-dimensional distortion. As with conventionallarge FOV systems attenuation effects have a much more significant effecton image accuracy.«xa0less

SPECT imaging of teboroxime during myocardial blood flow changes

Kinetic parameters and static images from dynamic SPECT imaging of (99m)Tc-teboroxime have been shown to reflect blood flow in dogs and in humans at rest and during adenosine stress. When compartment modeling is used, steady-state physiological conditions are assumed. With standard adenosine stress protocols, imaging of teboroxime would likely involve significant changes in flow, even if performed only for five minutes. These flow changes may significantly bias the kinetic parameter estimates. On the other hand, when static imaging is performed, large flow changes during acquisition may improve contrast between normal and occluded regions. Computer simulations were performed to determine the effect of changing flows on kinetic parameter estimation and on static (average tissue uptake) images. Two canine studies were also performed in which adenosine was given with a standard protocol, and then imaging was repeated with adenosine infusion held constant. The simulations predicted biases on the order of 7% for kinetic washin parameter estimation and 18% for the washout parameter. Contrast for static studies was found to depend critically on the time-activity behavior of the distribution as well as on the stress protocol. The differences in washin contrast from the standard and continous adenosine dog studies was slightly larger than predicted from the simulations. Optimal imaging of teboroxime with adenosine using compartment modeling will require non-standard adenosine stress protocols, although sub-optimal imaging may still be useful clinically.

Direct least squares estimation of spatiotemporal distributions from dynamic cardiac SPECT projections

Artifacts can result when reconstructing a dynamic image sequence from inconsistent, as well as insufficient and truncated, cone beam SPECT projection data acquired by a slowly rotating gantry. The artifacts can lead to biases in kinetic model parameters estimated from time-activity curves generated by overlaying volumes of interest on the images. However, the biases in time-activity curve estimates and subsequent kinetic parameter estimates can be reduced significantly by first modeling the spatial and temporal distribution of the radiopharmaceutical throughout the projected field of view, and then estimating the time-activity curves directly from the projections. This approach is potentially useful for clinical SPECT studies involving slowly rotating gantries, particularly those using a single-detector system or body contouring orbits with a multi-detector system. We have implemented computationally efficient methods for fully 4-D direct estimation of spatiotemporal distributions from dynamic cone beam SPECT projection data. Temporal splines were used to model the time-activity curves for the blood pool and tissue volumes in a simulated cardiac data acquisition. Least squares estimates of time-activity curves were obtained quickly and accurately using a workstation. From these curves, kinetic parameters were estimated accurately for noiseless data and with some bias for noisy data.

Simulation of the Beating Heart Based on Physically Modeling aDeformable Balloon

The motion of the beating heart is complex and createsartifacts in SPECT and x-ray CT images. Phantoms such as the JaszczakDynamic Cardiac Phantom are used to simulate cardiac motion forevaluationof acquisition and data processing protocols used for cardiacimaging. Two concentric elastic membranes filled with water are connectedto tubing and pump apparatus for creating fluid flow in and out of theinner volume to simulate motion of the heart. In the present report, themovement of two concentric balloons is solved numerically in order tocreate a computer simulation of the motion of the moving membranes in theJaszczak Dynamic Cardiac Phantom. A system of differential equations,based on the physical properties, determine the motion. Two methods aretested for solving the system of differential equations. The results ofboth methods are similar providing a final shape that does not convergeto a trivial circular profile. Finally,a tomographic imaging simulationis performed by acquiring static projections of the moving shape andreconstructing the result to observe motion artifacts. Two cases aretaken into account: in one case each projection angle is sampled for ashort time interval and the other case is sampled for a longer timeinterval. The longer sampling acquisition shows a clear improvement indecreasing the tomographic streaking artifacts.

Visualization of Fiber Structurein the Left and Right Ventricleof a Human Heart

The human heart is composed of a helical network of musclefibers. Anisotropic least squares filtering followed by fiber trackingtechniques were applied to Diffusion Tensor Magnetic Resonance Imaging(DTMRI) data of the excised human heart. The fiber configuration wasvisualized by using thin tubes to increase 3-dimensional visualperception of the complex structure. All visualizations were performedusing the high-quality ray-tracing software POV-Ray. The fibers are shownwithin the left and right ventricles. Both ventricles exhibit similarfiber architecture and some bundles of fibers are shown linking right andleft ventricles on the posterior region of the heart.