Biomedical subjects

Michel Defrise

Publications and source records attributed to Michel Defrise.

16 recordsLinked to original sources

PubMed2006

Statistical reconstruction for x-ray CT systems with non-continuous detectors.

We analyse the performance of statistical reconstruction (SR) methods when applied to non-continuous x-ray detectors. Robustness to projection gaps is required in x-ray CT systems with multiple detector modules or with defective detector pixels. In such situations, the advantage of statistical reconstruction is that it is able to ignore missing or faulty pixels and that it makes optimal use of the remaining line integrals. This potentially obviates the need to fill the sinogram discontinuities by interpolation or any other approximative pre-processing techniques. In this paper, we apply SR to cone beam projections of (i) a hypothetical modular detector micro-CT scanner and of (ii) a system with randomly located defective detector elements. For the modular-detector system, SR produces reconstruction volumes free of noticeable gap-induced artefacts as long as the location of detector gaps and selection of the scanning range provide complete object sampling in the central imaging plane. When applied to randomly located faulty detector elements, SR produces images free of substantial ring artefacts even for cases where defective pixels cover as much as 3% of the detector area.

Algorithms↗

PubMed2006

Resolution recovery in pinhole SPECT based on multi-ray projections: a phantom study.

PURPOSE: Low sensitivity can become a major problem when very small pinholes are used in SPECT imaging. Although a larger pinhole aperture will improve the sensitivity, this will be at the cost of the spatial resolution. With a view to improving the resolution-sensitivity trade-off, this paper explores an iterative reconstruction algorithm that models the pinhole aperture based on multi-ray projections. METHODS: This new implementation was validated using simulated data and phantom experiments. Two approaches were investigated. Firstly, the pinhole aperture was modelled in both the forward and the back projector. Secondly, the dual matrix implementation was investigated by modelling the pinhole aperture only in the forward projector. The systematic error, the full-width at half-maximum (FWHM) and the statistical error were quantified using the simulated data. Experimental phantom data were acquired for visual comparison with the reconstructions obtained from the simulated data. RESULTS: For a predefined number of iterations, the systematic error, the FWHM and the statistical error could be decreased when the pinhole aperture was modelled during iterative reconstruction. For a fixed, predefined statistical error of +/-10%, smaller systematic errors and smaller FWHM were obtained when modelling the pinhole opening. When the dual matrix implementation was used, equivalent results could be obtained as when modelling the pinhole opening in both the forward and the back projector. CONCLUSION: The multi-ray method to accomplish resolution recovery during the reconstruction of pinhole SPECT projection images offers a better trade-off between spatial resolution and noise compared with a reconstruction which does not model the pinhole aperture.

Algorithms↗

PubMed2006

Image reconstruction.

We give an overview of the role of Physics in Medicine and Biology in the development of tomographic reconstruction algorithms. We focus on imaging modalities involving ionizing radiation, CT, PET and SPECT, and cover a wide spectrum of reconstruction problems, starting with classical 2D tomography in the 1970s up to 4D and 5D problems involving dynamic imaging of moving organs.

Algorithms↗

PubMed2005

Fourier rebinning of time-of-flight PET data.

This paper explores fast reconstruction strategies for 3D time-of-flight (TOF) positron emission tomography (PET), based on 2D data rebinning. Starting from pre-corrected 3D TOF data, a rebinning algorithm estimates for each transaxial slice the 2D TOF sinogram that would have been acquired by a single-ring scanner. The rebinned sinograms can then be reconstructed using any algorithm for 2D TOF reconstruction. We introduce TOF-FORE, an approximate rebinning algorithm obtained by extending the Fourier rebinning method for non-TOF data. In addition, we identify two partial differential equations that must be satisfied by consistent 3D TOF data, and use them to derive exact rebinning algorithms and to characterize the degree of the approximation in TOF-FORE. Numerical simulations demonstrate that TOF-FORE is more accurate than two different TOF extensions of the single-slice rebinning method, and suggest that TOF-FORE will be a valuable tool for practical TOF PET in the range of axial apertures and time resolutions typical of current scanners.

Algorithms↗

PubMed2004

Reproducibility of left ventricular volume and ejection fraction measurements in rat using pinhole gated SPECT.

PURPOSE: The aim of this study was to investigate the intra-individual reproducibility of left ventricular volume and ejection fraction measurements in living rat using pinhole gated single-photon emission computed tomography (SPECT). METHODS: Eight normal male Wistar rats underwent four pinhole gated SPECT acquisitions over a 1-month period. Two pinhole gated myocardial perfusion SPECT studies were acquired at a 1-week interval after injecting the animals with 439+/-52 MBq of (99m)Tc-sestamibi. Subsequently, 1 week after the perfusion studies, two pinhole gated blood pool SPECT studies were acquired at a 1-week interval after in vivo labelling of the red blood cells using 520+/-49 MBq of (99m)Tc-pertechnetate. Pinhole gated SPECT acquisitions were done on a single-head gamma camera equipped with a pinhole collimator with a 3-mm opening and 165-mm focal length. Parameters of acquisition were as follows: 44 mm radius of rotation, 360 degrees rotation using a circular orbit, 64 projections, 64x64 matrix, gating using 16 time frames and 22-min acquisition time. The projection data were reconstructed with a modified version of OSEM taking into account the pinhole geometry and incorporating a prior assumption about the temporal properties of gated SPECT studies to reduce noise. Left ventricular volumes and ejection fraction were measured using automatic quantification algorithms. Inter-study, inter-observer and intra-observer reproducibility was investigated. RESULTS: Pinhole gated myocardial perfusion and pinhole gated blood pool images were of high quality in all animals. No significant differences were observed between the repeated measurements. The pinhole gated myocardial perfusion SPECT studies indicated that differences between repeated measurements larger than 41 microl for end-diastolic volume, 17 microl for end-systolic volume and 3% for ejection fraction were significant. The pinhole gated blood pool SPECT studies indicated that differences between repeated measurements larger than 42 microl for end-diastolic volume, 38 mul for end-systolic volume and 5% for ejection fraction were significant. In addition to the reproducibility measures, the accuracy of volume measurements in pinhole gated blood pool SPECT was confirmed by a phantom study. Excellent correlations were observed between the measured volumes and the actual phantom volumes. CONCLUSION: Pinhole gated SPECT is an accurate and reproducible technique for cardiac studies of small animals. Because this technique is non-invasive, the same animal can be imaged repetitively, allowing follow-up studies.

Animals↗

PubMed2004

Exact and approximate algorithms for helical cone-beam CT.

This paper concerns image reconstruction for helical x-ray transmission tomography (CT) with multi-row detectors. We introduce two approximate cone-beam (CB) filtered-backprojection (FBP) algorithms of the Feldkamp type, obtained by extending to three dimensions (3D) two recently proposed exact FBP algorithms for 2D fan-beam reconstruction. The new algorithms are similar to the standard Feldkamp-type FBP for helical CT. In particular, they can reconstruct each transaxial slice from data acquired along an arbitrary segment of helix, thereby efficiently exploiting the available data. In contrast to the standard Feldkamp-type algorithm, however, the redundancy weight is applied after filtering, allowing a more efficient numerical implementation. To partially alleviate the CB artefacts, which increase with increasing values of the helical pitch, a frequency-mixing method is proposed. This method reconstructs the high frequency components of the image using the longest possible segment of helix, whereas the low frequencies are reconstructed using a minimal, short-scan, segment of helix to minimize CB artefacts. The performance of the algorithms is illustrated using simulated data.

Algorithms↗

PubMed2004

MRI diffusion tensor reconstruction with PROPELLER data acquisition.

MRI diffusion imaging is effective in measuring the diffusion tensor in brain, cardiac, liver, and spinal tissue. Diffusion tensor tomography MRI (DTT MRI) method is based on reconstructing the diffusion tensor field from measurements of projections of the tensor field. Projections are obtained by appropriate application of rotated diffusion gradients. In the present paper, the potential of a novel data acquisition scheme, PROPELLER (Periodically Rotated Overlapping ParallEL Lines with Enhanced Reconstruction), is examined in combination with DTT MRI for its capability and sufficiency for diffusion imaging. An iterative reconstruction algorithm is used to reconstruct the diffusion tensor field from rotated diffusion weighted blades by appropriate rotated diffusion gradients. DTT MRI with PROPELLER data acquisition shows significant potential to reduce the number of weighted measurements, avoid ambiguity in reconstructing diffusion tensor parameters, increase signal-to-noise ratio, and decrease the influence of signal distortion.

Algorithms↗

PubMed2004

General reconstruction theory for multislice X-ray computed tomography with a gantry tilt.

This paper discusses image reconstruction with a tilted gantry in multislice computed tomography (CT) with helical (spiral) data acquisition. The reconstruction problem with gantry tilt is shown to be transformable into the problem of reconstructing a virtual object from multislice CT data with no gantry tilt, for which various algorithms exist in the literature. The virtual object is related to the real object by a simple affine transformation that transforms the tilted helical trajectory of the X-ray source into a nontilted helix, and the real object can be computed from the virtual object using one-dimensional interpolation. However, the interpolation may be skipped since the reconstruction of the virtual object on a Cartesian grid provides directly nondistorted images of the real object on slices parallel to the tilted plane of the gantry. The theory is first presented without any specification of the detector geometry, then applied to the curved detector geometry of third-generation CT scanners with the use of Katsevich's formula for example. Results from computer-simulated data of the FORBILD thorax phantom are given in support of the theory.

Algorithms↗

PubMed2004

The cone-beam algorithm of Feldkamp, Davis, and Kress preserves oblique line integrals.

The algorithm of Feldkamp, Davis, and Kress [J. Opt. Soc. Am. A 1, 612-619 (1984)] is a widely used filtered-backprojection algorithm for three-dimensional image reconstruction from cone-beam (CB) projections measured with a circular orbit of the x-ray source. A well-known property of this approximate algorithm is that the integral of the reconstructed image along any axial line orthogonal to the plane of the orbit is exact when the cone-beam projections are not truncated. We generalize this result to oblique line integrals, thus providing an efficient method to compute synthetic radiographs from cone-beam projections. Our generalized result is obtained by showing that the FDK algorithm is invariant under transformations that map oblique lines onto axial lines.

Algorithms↗

PubMed2003

Comparison of 180 degrees and 360 degrees data acquisition for determination of left ventricular function from gated myocardial perfusion tomography and gated blood pool tomography.

The purpose of this study was to address the issue of 180 degrees versus 360 degrees data collection for left ventricular ejection fraction (LVEF) and left ventricular volume measurements using gated myocardial perfusion tomography (gMPT) and gated blood pool tomography (gBPT). Thirty patients with known coronary artery disease were injected in a random sequence with 925 MBq of technetium-99m tetrofosmin and, within 2 days, with 740 MBq of (99m)Tc-labelled human serum albumin. gMPT and gBPT were acquired using 360 degrees data collection and reconstructed by filtered backprojection using all the acquired projection images and separately using only projection images acquired from 45 degrees LPO to 45 degrees RAO. In order to have the same global count densities, the counts in the 360 degrees data set were redistributed using binomial deviates just before reconstruction. After reorientation along the left ventricular long axis, LVEF and left ventricular volumes were calculated using fully automatic algorithms. Twenty-eight patients also underwent planar radionuclide angiocardiography (PRNA) on the same day as the gBPT. For the gMPT studies, the correlation between 180 degrees data collection and 360 degrees data collection was excellent ( r>0.98). Bland-Altman analysis revealed small systematic and random differences (<6%) between 180 degrees and 360 degrees. For the gBPT studies, the correlation between 180 degrees data collection and 360 degrees data collection was very good ( r>0.93). However, Bland-Altman analysis revealed systematic differences of 26% and random differences of 17%. When PRNA was used as a reference, the best results were obtained with gMPT acquired using 180 degrees data, while the worst results were obtained with gBPT acquired using 180 degrees data. In conclusion, when evaluating LVEF and left ventricular volumes from gMPT, either 180 degrees or 360 degrees orbits can be used. However, 360 degrees data acquisition is recommended when evaluating LVEF and left ventricular volumes from gBPT.

Aged↗

PubMed2003

Body contour 180 degrees pinhole SPET with or without tilted detector: a phantom study.

This study investigated the feasibility of ordered subsets expectation maximisation (OS-EM) reconstruction of 180 degrees pinhole single-photon emission tomography (SPET) acquired in body contour mode (variable distance between the detector and the axis of rotation for each projection) with or without a tilted detector head. Four non-circular orbits were designed bearing in mind the rotation radius and tilt angle values of previous pinhole SPET acquisitions in patients with circular orbits. The reconstructions were performed using a dedicated OS-EM algorithm. Reconstructed images of line and uniformity phantoms showed that the spatial and uniformity characteristics of the radioactive objects were preserved. In comparison with the circular orbits, the non-circular orbits allowed only a moderate gain (maximum 10%) in resolution. However, body contour pinhole SPET would significantly facilitate the camera set-up and in this way should decrease the camera set-up time, which is an important parameter in patient studies.

Algorithms↗

PubMed2003

An analytic study of the effects of attenuation on tumor detection in whole-body PET oncology imaging.

UNLABELLED: There has been considerable debate about the desirability of attenuation correction in whole-body PET oncology imaging. The advantages of attenuation correction are quantitative accuracy, whereas the perceived disadvantages are loss of contrast, noise amplification, and increased scanning time. In this work, we explain contrast changes between images reconstructed with and without attenuation correction. METHODS: To analytically explain both well-known and surprising phenomena in images reconstructed without attenuation correction, we performed a series of simulation studies, a phantom experiment, and a patient experiment. RESULTS: We showed that it is possible to calculate a priori the appearance of images reconstructed without attenuation correction. Compared with attenuation-corrected images, images without attenuation correction may have locally enhanced contrast in the abdomen or other regions of uniform attenuation, although the amount of enhancement varies with position in a complex manner. In regions of nonuniform attenuation, such as the thorax, it is possible that foci of increased tracer uptake disappear in images reconstructed without attenuation correction. The critical tracer concentration for this zero-contrast effect depends on the size, location, and density of the foci. Above the critical value, foci are visible in images with and without attenuation correction, whereas below the critical value, foci are visible in attenuation-corrected images but appear as photopenic regions in images without attenuation correction. CONCLUSION: Even though images without attenuation correction may be desired, these results suggest that all studies should at least be reconstructed with attenuation correction to avoid missing regions of elevated tracer uptake.

Female↗

PubMed2002

Diffusion tensor MR imaging of principal directions: a tensor tomography approach.

A novel approach to reconstructing the principal directions of a diffusion tensor field directly from magnetic resonance imaging (MRI) data using a tensor tomography data acquisition approach was developed. If tensor eigenvalues are assumed to be known, the reconstruction of principal directions requires fewer measurements than the reconstruction of the full tensor field. The tensor tomography data acquisition method (rotating diffusion gradients) leads to a unique reconstruction of principal directions, whereas the conventional MRI acquisition technique (stationary diffusion gradients) leads to an ambiguous reconstruction of principal directions when the same number of measurements are used. A computer-generated phantom was used to simulate the diffusion tensor field in the mid-ventricular region of the myocardium. The principal directions of the diffusion tensor field were assumed to align with the fibre structure of the myocardium. An iterative algorithm was used to reconstruct the principal directions. Computer simulations verify that the proposed method provides accurate reconstruction of the principal directions of a diffusion tensor field.

Algorithms↗

PubMed2002

Image reconstruction from fan-beam projections on less than a short scan.

This work is concerned with 2D image reconstruction from fan-beam projections. It is shown that exact and stable reconstruction of a given region-of-interest in the object does not require all lines passing through the object to be measured. Complete (non-truncated) fan-beam projections provide sufficient information for reconstruction when 'every line passing through the region-of-interest intersects the vertex path in a non-tangential way'. The practical implications of this condition are discussed and a new filtered-backprojection algorithm is derived for reconstruction. Experiments with computer-simulated data are performed to support the mathematical results.

Algorithms↗

PubMed2002

Reconstruction of gated myocardial perfusion SPET incorporating temporal information during iterative reconstruction.

Reconstruction of gated single-photon emission tomography (gSPET) is intrinsically a four-dimensional (4D) problem. In practice, the time frames are reconstructed independently as a sequence of frame-by-frame reconstructions. This approach is not optimal since the strong signal correlations among the individual time frames are not exploited. In this study we propose a simple but efficient algorithm to improve the image quality of myocardial perfusion gSPET by incorporating the cyclic temporal information within the reconstruction using Fourier filtering. The gSPET images were reconstructed using the Ordered Subsets Expectation Maximisation (OSEM) algorithm employing six iterations with eight subsets. Temporal filtering was applied either before (PreOSEM) or after image reconstruction (PostOSEM) or was incorporated within the OSEM algorithm (OSEM4D). The effect of temporal filtering was compared with conventional frame-by-frame OSEM using clinical data. Image quality was evaluated by estimating the systematic and statistical error. The results indicated that temporal filtering introduces a small (<1%) systematic error, while the statistical error was reduced from 15.0%+/-3.1% when conventional frame-by-frame OSEM was applied to 12.6%+/-2.7%, 12.0%+/-2.5% and 9.3%+/-2.4% when PreOSEM, PostOSEM and OSEM4D were used, respectively. It is concluded that temporal filtering incorporated within OSEM reconstruction dramatically reduces noise in gated SPET myocardial images.

Adult↗

PubMed2002

Clinically feasible reconstruction of 3D whole-body PET/CT data using blurred anatomical labels.

We present the results of utilizing aligned anatomical information from CT images to locally adjust image smoothness during the reconstruction of three-dimensional (3D) whole-body positron emission tomography (PET) data. The ability of whole-body PET imaging to detect malignant neoplasms is becoming widely recognized. Potentially useful, however, is the role of whole-body PET in quantitative estimation of tracer uptake. The utility of PET in oncology is often limited by the high level of statistical noise in the images. Reduction in noise can be obtained by incorporating a priori image smoothness information from correlated anatomical information during the reconstruction of PET data. A combined PET/CT scanner allows the acquisition of accurately aligned PET and x-ray CT whole-body data. We use the Fourier rebinning algorithm (FORE) to accurately convert the 3D PET data to two-dimensional (2D) data to accelerate the image reconstruction process. The 2D datasets are reconstructed with successive over-relaxation of a penalized weighted least squares (PWLS) objective function to model the statistics of the acquisition, data corrections, and rebinning. A 3D voxel label model is presented that incorporates the anatomical information via the penalty weights of the PWLS objective function. This combination of FORE + PWLS + labels was developed as it allows for both reconstruction of 3D whole-body data sets in clinically feasible times and also the inclusion of anatomical information in such a way that convergence can be guaranteed. Since mismatches between anatomical (CT) and functional (PET) data are unavoidable in practice, the labels are 'blurred' to reflect the uncertainty associated with the anatomical information. Simulated and experimental results show the potential advantage of incorporating anatomical information by using blurred labels to calculate the penalty weights. We conclude that while the effect of this method on detection tasks is complicated and unclear, there is an improvement on the estimation task.

Computer Simulation↗