Showing papers on "Iterative reconstruction published in 1978"
TL;DR: The method for attenuation correction in RCT is applied a simple, effective two-step procedure to the uncorrected image, and the filtered back-projection algorithm is used for its fast speed.
Abstract: The development of algorithms for Radionuclide Computed Tomography (RCT) is complicated by the presence of attenuation of gamma-rays inside the body. Some of the existing RCT reconstruction algorithms apply approximation formulas to the projection data for attenuation correction, while others take attenuation into account through some iterative procedures. The drawbacks of these algorithms are that the approximation formulas commonly used are generally inadequate and the iterative procedures are usually very time-consuming. The method for attenuation correction in RCT, which we propose, applied a simple, effective two-step procedure to the uncorrected image. In this procedure the filtered back-projection algorithm is used for its fast speed. A simple mathematical basis and description of the procedure together with some illustrative computer results are given in this paper.
1,304 citations
TL;DR: In this article, a technique for image reconstruction by a maximum entropy method is presented, which is sufficiently fast to be useful for large and complicated images and is applicable in spectroscopy, electron microscopy, X-ray crystallography, geophysics and virtually any type of optical image processing.
Abstract: Results are presented of a powerful technique for image reconstruction by a maximum entropy method, which is sufficiently fast to be useful for large and complicated images. Although our examples are taken from the fields of radio and X-ray astronomy, the technique is immediately applicable in spectroscopy, electron microscopy, X-ray crystallography, geophysics and virtually any type of optical image processing. Applied to radioastronomical data, the algorithm reveals details not seen by conventional analysis, but which are known to exist.
969 citations
TL;DR: A method that corrects for artifacts in X-ray transmission scanners created by the alteration of the energy spectrum by bone is described and the nonlinear nature of the spectral artifacts is shown to produce streaks in certain phantoms.
Abstract: A method that corrects for artifacts in X-ray transmission scanners created by the alteration of the energy spectrum by bone is described. The method involves two reconstructions of the pictures: the first establishes the approximate distribution of bone and the second enables the artifacts to be eliminated. The first reconstruction is needed to establish the total amount of bone along each ray. Assuming that the X-ray energy spectrum is known, it is shown how the line integrals can be corrected. The method was tested on several mathematical phantoms as well as on human head data. The nonlinear nature of the spectral artifacts is shown to produce streaks in certain phantoms.
300 citations
01 Jun 1978
TL;DR: The basic physics and mathematics underlying the production of reconstructed tomographic images are introduced and the evolution and philosophies of different possible data collection and reconstruction schemes are discussed.
Abstract: In recent years, Computer Aided Tomography (CAT) has had a major impact on the medical fields of radiology and neurology and nuclear medicine. This paper introduces the basic physics and mathematics underlying the production of reconstructed tomographic images. It discusses the evolution and philosophies of different possible data collection and reconstruction schemes.
186 citations
01 Dec 1978
TL;DR: In this paper, a combined source-channel coding approach is described for the encoding, transmission and remote reconstruction of image data, where the source encoder employs two-dimensional (2-D) differential pulse code modulation (DPCM).
Abstract: A combined source-channel coding approach is described for the encoding, transmission and remote reconstruction of image data. The source encoder employs two-dimensional (2-D) differential pulse code modulation (DPCM). This is a relatively efficient encoding scheme in the absence of channel errors. In the presence of channel errors, however, the performance degrades rapidly. By providing error control protection to those encoded bits which contribute most significantly to image reconstruction, it is possible to minimize this degradation without sacrificing transmission bandwidth. The result is a relatively robust design which is reasonably insensitive to channel errors and yet provides performance approaching the rate-distortion bound. Analytical results are provided for assumed 2-D autoregressive image models while simulation results are described for real-world images.
167 citations
108 citations
TL;DR: An approach, based on the theory of the Radon transform is used here to show that this problem of radar identification of perfectly conducting targets is equivalent to the problem of image reconstruction from projections, which has been solved in many diverse disciplines.
Abstract: In radar identification of perfectly conducting targets, a need arises to determine a convex body from its cross-sectional areas normal to the lines of sight for a number of view angles. An approach, based on the theory of the Radon transform is used here to show that this problem is equivalent to the problem of image reconstruction from projections, which has been solved in many diverse disciplines. Two of the inversion algorithms developed in these fields are used to estimate object shapes from their cross-sectional areas, obtainable from the objects' far-field, backscattered ramp responses.
89 citations
01 Jan 1978
83 citations
79 citations
TL;DR: This work focuses on X-ray scanners presently in production that use a method of image reconstruction called filtered back-projection, which has two limitations in data sampling which require that approximations be used during the back- projection operation, and hence result in error in the reconstructed image.
Abstract: Most X-ray scanners presently in production use a method of image reconstruction called filtered back-projection. There are two limitations in data sampling which require that approximations be used during the back-projection operation, and hence result in error in the reconstructed image. I
75 citations
TL;DR: The results indicate that the ART-reconstruction from twin-cone beam projections can be achieved without blurring artefacts.
Abstract: The twin-cone beam geometry has been used to reconstruct a three-dimensional object from projections. The reconstruction has been performed by means of the additive algebraic reconstruction technique (ART) extended to three dimensions. Several simulation experiments have been undertaken to explore the effect of pseudo-projections, analytic projections, iteration-dependent damping and interpolation schemes on the performance of the reconstruction. The results indicate that the ART-reconstruction from twin-cone beam projections can be achieved without blurring artefacts. The measured reconstruction error is of the same order of magnitude as the error of a ART-multislice reconstruction from coaxial projections.
TL;DR: A new imaging system for incoherent objects is constructed and the coherence function of the diffracted field is derived from the signals of three scanning intensity detectors by using computational manipulations.
Abstract: A new imaging system for incoherent objects is constructed. In this system, the coherence function of the diffracted field is derived from the signals of three scanning intensity detectors by using computational manipulations. The concrete optical and electronic systems, the details of the signal processings for the derivation of the coherence function, and calculations for image reconstruction are shown. The reconstructed images of asymmetric objects show the usefulness of the system.
01 Nov 1978
TL;DR: In this article, a true 3D reconstruction (TTR) algorithm for a cone beam with 4π detection geometry is introduced, which is applicable to position and X-ray computerized tomographic image reconstruction.
Abstract: A true three-dimensional reconstruction (TTR) algorithm which is applicable to a cone beam with 4π detection geometry is introduced. The TTR differs from conventional slice by slice 3-D reconstruction methods. Promising areas of application for the TTR algorithm are position and X-ray computerized tomographic image reconstruction.
TL;DR: This paper describes a method for the computer reconstruction of surfaces from a sequence of electron micrographs, and a data structuring approach to the problem of representing and analyzing objects of physiological importance.
Abstract: This paper describes a method for the computer reconstruction of surfaces from a sequence of electron micrographs, and a data structuring approach to the problem of representing and analyzing objects of physiological importance. The reconstruction technique involves the following stages: 1) object outlines are traced from each section, 2) the computer chain encodes these outlines, 3) the chain codes are reduced to the minimum number of boundary points which satisfactorily define the boundary, 4) polygons are mapped onto the boundary points between sections to approximate the surface, and 4) color coded, shaded surface views are computed of any subset of objects viewed and illuminated from arbitrary locations.
TL;DR: This chapter presents a series of images reconstructed from different combinations of diffraction spots, in which the interpretation is illustrated.
Abstract: Publisher Summary This chapter provides information on the image reconstruction in electron microscopy. Electron microscopy can routinely provide images of negatively stained biological specimens at a resolution of 1.5–3 nm, and of embedded and sectioned specimens at a resolution of 3–5 nm. This resolution, in particular with negative stain, is sufficient for the visualization of individual protein molecules and to show some details of subunit arrangement and shape. One approach to understanding the results of image reconstruction is to ask what each diffraction spot contributes to the image. The answer to this question follows directly from the theory of Fourier analysis and physical optics, but without going into the theory, the chapter presents a series of images reconstructed from different combinations of diffraction spots, in which the interpretation is illustrated. In addition, the optical diffraction pattern provides a direct criterion for distinguishing between the desired periodic features and the granular noise: all the information related to the periodic structures is contained in the discrete diffraction spots, whereas the noise produces a background of weak spots spread over the whole diffraction plane.
TL;DR: In this paper, real image reconstruction at the Stimulated Raman Scattering (SRS) of the object, illuminated by laser light, has been observed experimentally for the first time in the large wavelength diapason.
TL;DR: A detectability phantom is described which will provide a means of investigation of the effect on human detection capability of the peculiar correlations present in the noise present in CT scanner images.
Abstract: The EMI 5005 scanner produces images with noise characteristics similar to those in simulated CT reconstructions. A detectability phantom is described which will provide a means of investigation of the effect on human detection capability of the peculiar correlations present in the noise present in CT scanner images.
TL;DR: A technique is described which allows the removal of seeing distortions from a single frame of speckle-type imagery, provided that this frame is obtained using an aperture consisting of a nonredundant array of subapertures, each smaller than the seeing correlation length.
Abstract: A technique is described which allows the removal of seeing distortions from a single frame of speckle-type imagery, provided that this frame is obtained using an aperture consisting of a nonredundant array of subapertures, each smaller than the seeing correlation length. Although performed a posteriori, the method is related to those already proposed for use with active optical systems. Computer simulations are decribed which verify the basic features of this technique. The simulations indicate that reconstructed images of diffraction-limited quality should be obtainable for starlike objects as dim as eighth magnitude. For more extended objects, the limiting magnitude depends somewhat on the object structure. The technique described is immediately applicable to any large telescope, and because the processing is done after the fact, a frame containing many isoplanatic patches may be processed piecewise, allowing the reconstruction of large areas.
TL;DR: In this article, an iterative approach was proposed to reduce the structured background existing in the images from the contribution of offfocal plane activity, resulting in low contrast, poorly resolved images.
Abstract: Positron cameras using two opposed area detectors for the coincidence detection of 511 keV annihilation photons have been available for some time. Image reconstruction from the collected annihilation data in this camera configuration is usually based on focal plane tomography. However, this method suffers from structured background existing in the images from the contribution of offfocal plane activity, resulting in low contrast, poorly resolved images. This paper presents a method which reduces this off-focal plane background through an iterative approach which weights individual picture elements according to the activity computed by the previous iteration.
TL;DR: Using a small x-rays source designed for use in dental clinics and an x-ray grid matched with the lenticular sheet, 3-D x-Ray images having a 0.001-rad resolvable angle and a 0-02-rad flipping angle are directly recognized with clear parallax.
Abstract: Theoretical and experimental investigations of a method for direct recording and reconstruction of 3-D x-ray images are described. When recording, an x-ray source is placed at discrete positions, and an x-ray grid is fixed on the x-ray plate. When reconstructing, a lenticular sheet matched with the grid is overlayed on the developed plate, and a parallax–panoramagram type 3-D image is seen with the naked eye. Using a small x-ray source designed for use in dental clinics and an x-ray grid matched with the lenticular sheet, 3-D x-ray images having a 0.001-rad resolvable angle and a 0.02-rad flipping angle are directly recognized with clear parallax.
TL;DR: In this paper, a computational method for obtaining three-dimensional reconstructions of positron emitting radionuclei distributions using a Planar Positron camera is described, which involves the use of a Filtered Fourier Deconvolution Method Construction of Generalized Tomograms capable of emphasizing the large angle events.
Abstract: A computational method for obtaining Three-Dimensional reconstructions of positron emitting radionuclei distributions using a Planar Positron camera is described The method involves the use of a Filtered Fourier Deconvolution Method Construction of Generalized Tomograms capable of emphasizing the large angle events is introduced The finite size of the Positron Camera detector and its effect on the reconstruction are discussed
TL;DR: Techniques for digital reconstruction of 3-D imagery from 2-D pictures are discussed and digital implementations of serioscopy, tomosynthesis, and pulsed-source coded-scan tomography are illustrated.
Abstract: Techniques for digital reconstruction of 3-D imagery from 2-D pictures are discussed. Digital implementations of serioscopy, tomosynthesis, and pulsed-source coded-scan tomography are illustrated.
TL;DR: In this paper, the authors propose to use a uniformly redundant array (URA) as the coded aperture in conjunction with a special correlation decoding method, which results in a delta function with deterministically zero sidelobes.
Abstract: Coded aperture imaging of gamma ray sources has long promised an improvement in the sensitivity of various detector systems. The promise has remained largely unfulfilled, however, for either one of two reasons. First, the encoding/decoding method produces artifacts, which even in the absence of quantum noise, restrict the quality of the reconstructed image. This is true of most correlation-type methods. Second, if the decoding procedure is of the deconvolution variety, small terms in the transfer function of the aperture can lead to excessive noise in the reconstructed image. We propose to circumvent both of these problems by use of a uniformly redundant array (URA) as the coded aperture in conjunction with a special correlation decoding method. The correlation of the decoding array with the aperture results in a delta function with deterministically zero sidelobes. The properties of the encoding/decoding method are similar to those of the nonredundant pinhole array (NRA), however, the URA can be composed of thousands of holes whereas the NRA contains less than 40. In short, the URA offers the transmission advantage of the random array or Fresnel zone plate without introducing the artifacts typically seen when those apertures and others are used. It is shown that the reconstructed image in the URA system contains virtually uniform noise regardless of the structure in the original source. Therefore, the improvement over a single pinhole camera will be relatively larger for the brighter points in the source than for the low intensity points.
TL;DR: The basic principles of a quantitative tomographic technique for bone mineral determination are considered and the bone mineral content could be evaluated from a single reconstruction using projection data from two different energies.
Abstract: The basic principles of a quantitative tomographic technique for bone mineral determination are considered. The bone mineral content could be evaluated from a single reconstruction using projection data from two different energies. Beam hardening effects and photon counting statistics are studied by computer simulation and found to be the limiting factors in precision and accuracy.
TL;DR: In this article, the dependence of the perturbations upon the system parameters is studied and criteria are derived for limiting them, where the sampling operation is done using a receiving transducer array to convert the acoustic signals into electrical ones, which may be amplified before being applied to another reemitting transducers array: an acoustical field identical to the original one is so reconstructed.
Abstract: It is often difficult and unpractical, in the acoustic imaging area, to insert a focusing lens between the observed object and the ultrasonic receiver. Furthermore, the focal length must, in the B scan mode, be varied very fast in order to keep the line sources lying at different depths well focused. A spatial‐sampling scheme may be used for circumventing these problems by taking off the possibly complex processing unit from the element contacting the observed object, without additional complexity or degradation in the flexibility and usefulness of the apparatus. The sampling is done using a receiving transducer array to convert the acoustic signals into electrical ones, which may be amplified before being applied to another reemitting transducer array: an acoustical field identical to the original one is so reconstructed. The sampling operation will, however, alter the image quality by introducing parasitic images. The dependence of the perturbations upon the system parameters is studied in this paper and criteria are derived for limiting them.
Patent•
13 Nov 1978TL;DR: In this article, a method and apparatus for computerized axial tomography (CAT) image reconstruction applicable to X-ray scanning of the human body is presented, where a series of calculations determine the values of a characteristic in defined eegmented areas of an examination plane.
Abstract: A method and apparatus for computerized axial tomography (CAT) image reconstruction applicable to X-ray scanning of the human body. Successive series of calculations determine the values of a characteristic in defined eegmented areas of an examination plane. A reduction in the time and equipment requirements for reconstruction calculations may thereby be effected. Images representative of the difference between the value of the characteristic at a reconstruction point and the average value of the characteristic in adjacent regions (Δµ) may be calculated and displayed from measurements taken within a localized contiguous region of the examination plane. Radiation dose to patients and computation time in X-ray computerized axial tomography scanning systems is thus reduced. Differential displays of the type described may be adjusted to image boundaries, in which case they do not suffer from gray scale resolution problems which are typical of prior art displays. Image artifacts attributable to interpolation errors may be reduced with negligable effect on resolution by incorporating a weighting function, preferentially a Gaussian function, in the image reconstruction.
TL;DR: It is demonstrated that matrix inversion generates artifact free tomographic information and this method is perfectly general and has been applied to zone plate coded aperture imaging.
Abstract: Longitudinal tomoplanes characterizing a three-dimensional object distribution suffer from the problem of interference between planes and their information content can be misleading for diagnostic purposes. Techniques for unscrambling linked tomograms are examined and it is demonstrated that matrix inversion generates artifact free tomographic information. This method is perfectly general and has been applied to zone plate coded aperture imaging. For this system the quantum noise limitation has been examined and is relatively independent of the spatial frequency content of the object. Provided sufficient information can be collected, processed reconstructions faithfully represent both high frequency and uniform object distributions.
TL;DR: In this paper, a new analogue recording system that uses a fan-shaped X-ray beam and a built-in photographic summation technique for image reconstruction is described, which is illustrated with different test samples.
Abstract: Transaxial tomography is a radiographic technique that provides an image of a selected two-dimensional section of a three-dimensional object. The present paper describes a new analogue recording system that uses a fan-shaped X-ray beam and a built-in photographic summation technique for image reconstruction. The performance of the camera is illustrated with different test samples.
TL;DR: Three cases are presented to illustrate this new supplementary modality in the pediatric patient and the technique, diagnostic implications and value of orthogonal plane image reconstruction are discussed.
Abstract: Recent technical advances in computed tomography (CT) allow sagittal and coronal image reconstruction by computer manipulation of data from thin-section transverse scans. Three cases are presented to illustrate this new supplementary modality in the pediatric patient. The technique, diagnostic implications and value of orthogonal plane image reconstruction are discussed.