Search

The common imaging practice of each detector on a multidetector SPECT imaging system to acquire only a fraction of the total number of projections in a 360/spl deg/ projection set introduces the possibility of imbalances in system response which can cause artifacts and changes in image quality. Here we evaluate the effect of variations in detector sensitivity caused by differences in the position of the photopeak window, window width, and different collimators on reconstructed image quality. Analytical simulations of cardiac phantom geometries and experimental measurements with Data Spectrum's Cardiac and Deluxe SPECT phantoms on a three detector SPECT imaging system were used in the evaluation. The relative efficiency of one of three detectors was changed over a series of values between 100 and 0%, while the efficiency of the remaining two detectors was maintained at 100%. The results show that decreased activity and geometric distortions occur along directions perpendicular to the projections with altered sensitivity. The degree of distortion and decrease of activity depends on the geometry of the activity distribution and the reduction in sensitivity. The relative count ratios in the regions with decreased activity varied over a range of 1.0 to 0.59 as the detector sensitivity is decreased from 100 to 0% of the full energy window. Variations in detector sensitivity can degrade spatial resolution and uniformity and create artifacts in reconstructed tomographic slices.

In SPECT imaging, a cylindrical detector geometry offers many advantages. Because the detector completely surrounds the patient, the cylindrical system geometry should provide more detector coverage than the rotating camera SPECT systems. This increased coverage means increased sensitivity, or improved resolution, and better image quality. Because the cylindrical detector system is stationary and the projection data is acquired through collimator rotation, this geometry simplifies system design and allows very stable imaging performance. The authors propose a cylindrical system geometry that features a large number of projections for a small FOV located at the center of the gantry. This geometry has potential for cardiac SPECT imaging, provided that the patient's body can be offset properly and reliably to place his heart in the specified FOV. General design issues and potential configurations for such a cardiac SPECT system are presented.

The authors have developed a new imaging geometry-the asymmetric-fan-beam (AsF)-to expand the imaging field of view (FOV) for transmission imaging on current SPECT systems. The AsF geometry samples slightly more than one-half of the FOV in each projection and yields half-truncated projection data. Although each projection profile is not complete, the combined acquired data set meets the minimum sampling requirement because the other half of the FOV is sampled in the opposite projections after a full 360/spl deg/ rotation of the detector system. To take advantage of the simple convolution-backprojection algorithm for reconstruction, the key is in the handling of the projection profile for convolution. The authors have investigated such a technique to process the truncated projection profile for reconstruction without the side effects caused by the truncation. This technique entails filling in the truncated portion of each profile with interpolated data derived from other projections. After convolution, only the corresponding half of the original projection profile is backprojected in reconstruction. This is done to minimize propagation of the errors of interpolation in the reconstruction. Reconstructed images from phantoms and human subjects demonstrate that this processing technique yields good quality transmission images.

For proper attenuation correction of SPECT images, a set of 3D attenuation maps specific to the imaging slices is needed. Among the many different approaches for deriving the attenuation maps, fan beam transmission CT (FBTCT), performed on the same SPECT system as emission imaging, has many promising and clinically practical features. The major problem of FBTCT is that the current SPECT systems do not have a large enough field of view (FOV) to cover the typical cross-sectional size of patients. To address this problem, the authors have developed a novel asymmetric fan (AsF) sampling scheme to extend the FOV to practical sizes for clinical TCT imaging on existing SPECT systems. This AsF scheme samples only half of the intended FOV in each projection; the other half would be sampled in an opposing projection after detector rotation. The authors have implemented the AsF sampling on a three-head SPECT system through a specially designed source-collimator assembly. The authors have modified the conventional convolution backprojection algorithm to facilitate simple and fast image reconstruction. The feasibility of the approach is confirmed by the quality of the derived TCT images of various phantoms and human subjects. The AsF sampling scheme could also have applications in other general transmission CT systems.

The authors describe a new technique that uses simultaneous transmission and scatter SPECT imaging for determination of segmented attenuation /spl mu/-maps. These /spl mu/-maps could provide the basis for attenuation correction and quantification of radiopharmaceutical uptake in cardiac SPECT. For transmission imaging, a line source and a fan-beam collimator on one head of a triple-head SPECT system are used. For scatter imaging, two external ring sources and parallel collimators on one or two of the other heads are used. The fan-beam transmission data are truncated, while the parallel-beam data are not. The proposed technique uses both sets of tomographic data to complement each other and to obtain a composite image. A /spl mu/-map with three segmented regions of distinct /spl mu/-values could be derived from the composite image. This final /spl mu/-map can then be used as the basis for attenuation correction. >

A study was conducted to evaluate the feasibility of simultaneous dual radionuclide brainimaging with 123I and 99m Tc using photopeak image subtraction techniques or offset photopeak image acquisition. The contribution of the photons from one radionuclide to a second radionuclide’s photopeak energy window (crosstalk) was evaluated for SPECT and planar imaging of a brain phantom containing 123I and 99m Tc for a range of activity levels and distribution properties approximating those in rCBF images of the adult human brain. Crosstalk was evaluated for 10% symmetrical energy windows centered on the 123I and 99m Tc photopeaks and for 10% energy windows asymmetrically placed to the left and right of the center of the respective photopeaks. Major observations include: (1) in the centered photopeak windows, 99m Tc crosstalk in the 123I window is 8.9% of the 99m Tc seen in the 99m Tc window and ranges from 37.5% to 75.0% of the 123I in the 123I window. 123I crosstalk is 37.8% of the 123I seen in the 123I window and ranges from 4.4% to 8.9% of the 99m Tc seen in the 99m Tc window; (2) the spatial distribution of a radionuclide’s crosstalk photons differs from that observed in the radionuclide’s photopeak window; (3) a 99m Tc photopeak window offset to the left does not decrease 123I crosstalk, and the percentage of 99m Tc scatteredphotons is significantly increased in the window. Offsetting the 123I window to the right decreases 99m Tc crosstalk to 9.0% to 17.9% of the 123I counts, but decreases 123I sensitivity by 39.9%; and (4) offsetting both photopeak windows to the right decreases the 99m Tc scatteredphotons in the 99m Tc window, but increases 123I crosstalk to 17.0% to 33.8% of the 99m Tc counts. The findings show that image quality, spatial resolution, and quantitative accuracy are degraded to unacceptable levels with the combinations of energy windows tested for dual radionuclide imaging of 99m Tc and 123I. This indicates that dual radionuclide imaging must be thoroughly tested and validated before use in clinical studies.

Analytical expressions that describe the dependence of slopes and amplitudes of the scatter distribution functions (SDF) on source depth and media density are used to estimate a scatter component in SPECT projection data. Since the ratio of detected scattered to total photons (S/T), SDF amplitude and slope depend strongly on line source length (SL) used to obtain SDFs, we compared estimated scattered components using SDFs, obtained for lengths of 2-21 cm. At 10 cm source depth, S/T changes from 0.19 to 0.36 when SL changes from 2 to 21 cm. Scatter amplitude's dependence on source depth (d) in water was described by 6.38e/sup -0.186d/ for a 2 cm and 16.15e/sup -0.129d/ for a 21 cm SL. Slope was described by 0.292d/sup -0.601/ for a 21 cm SL, and by 0.396d/sup -0.82/ for a 21 cm SL. The estimated scatter components are compared with simulated SPECT projection data obtained with Monte Carlo modeling of six hot spheres placed in a cylindrical water filled phantom. The comparison of estimated with simulated total counts/projection shows very good agreement when approaching SDF for point source (the % difference varied from 2 to 13% for 2 cm SL). Significant overestimate is seen when source length increases.

Since many different configurations of brain SPECT systems are available for clinical imaging recently, it is informative to investigate if there are much differences in their performance potential. However, the variation of system configurations and performance characteristics make direct comparisons difficult. To clarify the confusion, the authors have developed a method to analyze the relative merits of different basic system geometries and evaluating the potential for different system configurations. This method is based on the calculated system geometric efficiency (SGE) of each configuration under the condition that the reconstructed spatial resolution are normalized. In the normalization process, the design parameters and performances of the collimators were iteratively adjusted to match the reconstructed resolution of an appropriately defined reference specifications. After the collimator was normalized, the geometric efficiency (GE) of the collimator was calculated by integrating the point source geometric efficiency (PSGE) over a brain source model with narrow-beam attenuation effect imposed. Based on the calculated GE for each collimator, the SGE is derived for each system configuration. This analysis was performed for many existing geometries of brain SPECT systems with different detector-collimator configurations of commercial interest. >

The authors have recently developed a new sampling geometry-asymmetric fan (AsF) beam-which completes sampling of the imaging field of view (FOV), and thereby avoids the problems caused by projection truncation in transmission imaging with SPECT systems. The AsF geometry for a particular angular view covers only a half of the FOV. The problem in the raw projection data is that a sharp cut-off that causes spurious oscillations within the central region of a reconstructed image. But, since the 360/spl deg/ acquisition provides complete sampling, each individual projection can be completed by the proper use of the opposing views data. Several methods were investigated to prepare the data for filtering and backprojection. The first approach was the simplest one, using just the mirror reflection of original data; 'fooling' the Fourier transform. The second one was the use of the single opposing projection, the one separated by the 180/spl deg/, without any processing. The last one, yielding the best results, provides missing data by resorting and interpolation. It uses four projection bins, two from each of the two neighboring angular views, that do not exactly oppose the original projection. >

Three-dimensional (3-D) imaging in medicine includes visualization, manipulation and analysis of structural information captured in 3-D (and higher dimensional) digital images. This activity has become an established discipline in medical imaging. A method for surface representation of 3-D images (volumes) is presented. The procedure performs triangulation of the surface of the 3-D object and represents the result using a 3-D scene description language known as the virtual reality modeling language (VRML). The iterative procedure is implemented in C language on a SUN workstation.

Conventional partial orbit SPECT acquisition (90/spl deg/, 120/spl deg/ or 180/spl deg/ depending on number of camera heads) on multihead cameras enhance the problems caused by patient motion or change in organ position ("cardiac creep"), since the misplaced data are present at several intervals in the projection set. The authors investigated the effects of source motion and developed new image motion correction methods based on 360/spl deg/ acquisition of each detector on a three camera SPECT system. Experimental measurements with point sources and cardiac phantom were used to evaluate the influence of source motion on SPECT images. A point source phantom was rotated 5-30/spl deg/ CCW and moved axially 0.5-4.0 cm. A cardiac phantom was moved in laterally (0.5-4.0 cm) and in the axial direction (0.5-4.0 cm). Cross-correlation of images from different heads at the same angular position is used to detect and correct for motion. Two correction methods were investigated: (1) summation of complete 360/spl deg/ projection sets without correction, (2) time normalization of projection frames without motion to replace frames affected by motion. The first method corrects for small displacement of the source (

Recent studies show Tl-201 and Tc-99m MIBI to have high sensitivity for detecting and differentiating tumor from benign lesions in the breast. A study was conducted to optimize the design of a multi-pinhole collimator for high spatial resolution SPECT imaging of the breast to detect small tumors (0.5/spl times/0.5 cm). High sensitivity is achieved by using multiple pinholes, and rotating the camera 90-180 degrees around the breast. Monte Carlo simulation studies and analytical methods are applied to determine the number of holes, hole position, hole aperture size, conical field of view, and focal length of the collimator to optimize spatial resolution and sensitivity for breast imaging. Shielding is optimized to accommodate an energy range of 50-511 keV. An image reconstruction algorithm for cone beam imaging geometry for rotating single pinhole SPECT will be modified to accommodate simultaneous imaging from multiple pinholes with limited angular sampling. Special emphasis is placed on evaluation of imaging with small pinhole apertures (1-2 mm diameter), that are required to achieve the resolution of 3-7 mm at the distance of 10-15 cm with radionuclides having energies /spl les/140 keV. Complete 360/spl deg/ acquisition of the breast tumor model was simulated to study the artifacts in reconstructed images due to limited angle acquisition and to select an optimal reconstruction algorithm.

A new algorithm for blind restoration of space-invariant image degradations is presented. Once the singular value decomposition (SVD) of the degraded image is computed, degradation system identification and image restoration are performed through operations on singular vectors and singular values of the degraded image. The unknown point-spread function (PSF) is estimated from the singular vectors, while noise variance is estimated from the smallest singular values. It has been shown how to estimate the phase function for blurs that have zeros on the unit bicircle. Image restoration is performed through Wiener filtering of singular vectors. Knowledge about statistical properties of singular vectors and singular values is built into estimation and restoration procedures.

A web-based application for computed tomography (CT) image analysis and visualization of the bronchial airways is presented in the paper. The technique used in this application provides a noninvasive way to examine the interior of the bronchial tubes. Many different properties of the tubes such as presence of the foreign objects or structure abnormalities of the tubes can be detected this way. This method can also be used for other similar structures, such as blood vessels. The application uses the CT images obtained by spiral computed tomography as input. Amplitude thresholding is used for segmentation of the airways. The medial axis of the airways has been used as the path for animation. The procedure produces a fly-through animation in the MPEG format, which is obtained by volume rendering of the original CT images along the extracted path. The user can choose the appropriate path using the Virtual Reality Modeling Language (VRML) model, which is sufficient for visualization of basic structures. The algorithms and the software have been developed on the SUN Ultra 1 workstation.

The objective of this study was to determine the feasibility of using a fast (short-duration) transmission computed tomogram (TCT), acquired immediately before or after the emission CT, to correct for photon attenuation in cardiac SPECT.The asymmetric fanbeam geometry with a 99mTc line source was used to acquire TCTs after conventional cardiac emission CT imaging on a triple-head SPECT system. The TCTs were reconstructed to generate patient-specific attenuation maps, which were used with an iterative maximum likelihood algorithm to reconstruct attenuation-corrected cardiac SPECT studies. The results of attenuation correction based on TCTs as short as 1 min were compared with long-duration transmission imaging for a phantom and several human studies.Attenuation correction based on asymmetric fanbeam TCT significantly improves the uniformity of images of a uniform tracer distribution in a cardiac-thorax phantom configured to simulate a large patient. By using a high-activity line source and a rapid camera rotation, a suitable attenuation map for this phantom can be obtained from a 4-min TCT. A similar result is obtained for patients with thorax widths of40 cm.A sequential imaging protocol for acquiring a fast TCT can be used for attenuation correction of cardiac SPECT imaging. The sequential TCT can be acquired without significantly extending the duration of the imaging study. This method provides a way to perform attenuation correction on existing triple-head SPECT systems without extensively modifying the system.

Using circular 180 degrees and 360 degrees SPECT acquisition modes the transaxial resolution of line sources in air and water were measured at different positions in the field of view. With the 180 degrees acquisition mode, all line sources in air located off the axis of camera rotation (AoR) showed an oval distortion. This distortion was systematically related to the starting point of the rotating detector. On axis line sources in air were undistorted, regardless of the 180 degrees acquisition starting angle. The 360 degrees acquisition images of the line sources in air showed a similar effect but in a very mild form. In water, transaxial reconstructions of line sources (off axis) showed an enhancement of the oval distortion for both the 180 degrees and 360 degrees acquisitions. Computer simulations of the line source measurements were performed and correlated well with the experimental data. The line source results are explainable by the inherent depth dependent response of the scintillation camera. In clinical SPECT studies, distortions of this nature will be most appreciable with 180 degrees imaging of small organs that are located off the AoR.

The knowledge of the albedo of an exoplanet can be a very good source of information about its physical and chemical characteristics; however, it is normally very difficult to separate the planet albedo from the star spectrum, as the amplitude of the planet spectrum is much lower than that of its star, and the two spectra are acquired together, as it is not possible to spatially separate them. Here we present an algorithm for the estimation of the planet albedo, which exploits the amplitude modulation of the planet spectrum during its orbit around the star and works also if no eclipses are present. Results are presented using simulated spectra of the CoRoT-1 b system.

The Lucy-Richardson algorithm is a very well-known method for non-blind image deconvolution. It can also deal with space-variant problems, but it is seldom used in these cases because of its iterative nature and complexity of realization. In this paper we show that exploiting the sparse structure of the deconvolution matrix, and utilizing a specifically devised architecture, the restoration can be performed almost in real-time on VGA-size images.

We systematically analyse the design of the low-level vision components of a real-time system able to help a blind person in his/her social interactions. We focus on the acquisition and processing of the video sequences that are acquired by a wearable sensor (a smartphone camera or a Webcam) for the detection of faces in the scene. We review some classical and some very recent techniques that seem appropriate to the requirements of our goal.

The problem of detecting cell nuclei in fluorescence images may be faced by means of a segmentation step, to get the neighbourhood of candidate nuclei, followed by a binary classification step. Important for the latter step is the choice of the descriptors (features) to be extracted from the neighbourhood and used by the classifier. In the present paper, based on a large set of manually labelled samples, we evaluate several of such descriptors combined with some common type of support vector machines. We show that equipping the detection algorithm with the best combination of features/classifier leads to a performance comparable to human labelling by experts.

In this paper we are presenting a method for achieving strain analysis of cardiac tissues from 3D ultrasound images. First of all, the left ventricle was segmented by an iterative snake algorithm which exploits the gradient vector flow field estimated from the 3D images. Once the ventricle was segmented, a set of points located on the external surface of the ventricle was selected. The movement of all these points along the entire cardiac cycle was calculated by using speckle tracking techniques based on the minimization of a Sum of Absolute Differences. The strain of all the segments connecting two of the considered points was computed, together with the rotation of the points along the long axis of the ventricle. The method was tested on images acquired from 10 patients: 3 with heart pathologies, 5 healthy subjects, and 2 patients with mild hypertrophy. The segmentation of the left ventricle with iterative methods demonstrated to be able to delineate the external surface of the ventricle in a consistent way. Preliminary results showed the feasibility of characterizing healthy subjects and patients with well-defined heart pathologies by using the outcomes achieved with the strain analysis.

International audience; The Logarithmic Image Processing (LIP) approach is a mathematical framework developed for the representation and processing of images valued in a bounded intensity range. The LIP theory is physically and psychophysically well justified since it is consistent with several laws of human brightness perception and with the multiplicative image formation model. In this paper, the so-called Color Logarithmic Image Processing (CoLIP) framework is introduced. This novel framework expands the LIP theory to color images in the context of the human color visual perception. Color images are represented by their color tone functions that can be combined by means of basic operations, addition, scalar multiplication and subtraction, opening new pathways for color image processing. In order to highlight the CoLIP relevance with color constancy, a color correction method based on the subtraction is proposed and tested on CoLIP approach and Logarithmic hUe eXtension (LUX) approach, also based on the LIP theory, on differently illuminated images: underwater images with a blue illuminant, and indoor images with yellow illuminant.