A method is presented for producing model limited angle artifact in phantom images. Iucr limited angle tomography for transmission xray. Both the art algorithm and an orthogonal function algorithm were investigated. Limitedangle optical computed tomography algorithms. Artifacts and visible singularities in limited data xray tomography eric todd quinto received. The otherwise attractive convolutionbackprojection algorithm is not suited for ct image reconstruction from measurements comprising an incomplete set of projection data.
Based on these characterizations, we will explain how to mitigate those. Characterization and reduction of artifacts in limited angle tomography 3 characterizations of the artifacts in limited angle fbp and reconstructions. Projection data replenishment algorithm for limited angle. A reconstruction algorithm, based on a hybrid xray excitationdiffuse optical emission model, was tested in a numerical and experimental phantom that had dimensions similar to the human. Image reconstruction from such limited angle data suffers from artifacts because of missing data. One group of deep learning reconstruction algorithms apply postprocessing neural networks to achieve imagetoimage reconstruction, where input images are reconstructed by conventional reconstruction methods. Algorithms have been developed by researchers 6, 14, 17, 30. It is applicable to limited angle tomography as well since typical imaged objects in computed tomography have a limited spatial extent, which means that the fourier transforms of the objects can be considered band limited signals. We will develop a hybrid reconstruction framework that fuses model. It is caused by missing data over a certain angle interval, which make an inverse radon transform impossible. Artifacts and visible singularities in limited data xray. The artifact produced by the different methods was very similar.
Optimal selection of initial image would influence the iterative reconstruction performance but has not been studied deeply yet. Captures more photons, complex inverse problem limited angle tomography. Photoninefficient but simple algorithms coded aperture imaging. Pdf backprojection filtering method for limited angle. The algorithm is based on calculating artificial projections in those directions where projection data are unavailable. Many specialized iterative algorithms exist which clearly improve the image quality 1. Exterior and limitedangle tomography in nondestructive. We will use fdtd to generate the signals the radar collects at a spill site. As the projection data are highly incomplete, standard tomographic reconstruction algorithms, such as the wellknown fbp algorithm, will not afford a reliable.
Those characterizations will precisely explain where and why artifacts are created by these algorithms. Pdf an improved statistical iterative algorithm for sparseview and. An iterative algorithm for computed tomography image reconstruction from limitedangle projections. Tests of the algorithm on two highquality data sets provided by perceptics, inc. An algorithm specifically designed for limited angle tomography. We describe a backprojection filtering method for limited angle tomography. Introduction as a nondestructive imaging technology, computed tomography ct has been widely.
Scalespace anisotropic total variation for limited angle tomography yixing huang, oliver taubmann, xiaolin huang, viktor haase, guenter lauritsch, and andreas maier abstractthis paper addresses streak reduction in limited angle tomography. A low radiation dose and high inplane resolution make lat an attractive alternative to computed tomography. Iterative image reconstruction for limitedangle ct using. Evaluation of a limited angle scanned electron beam xray. Pdf computed tomography ct is commonly used in clinical. Limited angle computed tomography ct has great impact in some clinical applications. Various existing algorithms are available to solve the illposed regression problem by different solvers with different regularization terms louis and tornig and natterer solve it with the svd. To better preserve the edges and suppress the artifacts to limited angle ct image reconstruction, we developed an alternating iterative reconstruction algorithm for limited angle ct based on. Classical incomplete data problems have been studied from the beginning of tomography, including limited angle tomography, where the data can be collected only from certain view angles 24,30. For limited angle tomography, algorithms of adaptive steepest descent.
Pet is accurate image reconstruction algorithms which can image the object of interest faithfully. Limited angle ct reconstruction is a very challenging illposed issue and of great interest in several clinical applications, such as digital breast tomosynthesis 1, dental tomography 2, short exposure time 3, etc. Total variation iterative constraint algorithm for limited. By extending our prior work, here we propose a novel multiscale waveletdomain residual learning network for limited angle ct reconstruction. In this paper, we improve a statistical iterative algorithm based on the. Accurate image reconstruction from fewview and limited. A new scale space total variation algorithm for limited angle tomography yixing huang, oliver taubmannz, xiaolin huang, viktor haasex, guenter lauritschy and andreas maierz pattern recognition lab, friedrichalexanderuniversity erlangennuremberg, erlangen, germany. Improving reconstructed image quality in a limitedangle positron. A new iterative algorithm of tomographic reconstruction of objects on the basis of projection data available in a limited range of angles only is proposed. Although our algorithm is proposed for the limited angle tct, since the sart method and deep learning technique will not be influenced by the geometrical structure of the scanning mode, our algorithm can be extended to the generic limited angle tomography, such as carm conebeam ct. Limited angle tomography lat is an xray imaging technique for generating a stack of crosssectional images based on a limited number of twodimensional projections. In particular, the network is designed in a directional wavelet transform domain to exploit the directional property of the limited angle artifacts.
Pdf deep learning based image reconstruction algorithm. An iterative algorithm for computed tomography image reconstruction from limited angle projections. Improved backpropagation algorithms by exploiting data. Reconstructions in limited angle xray tomography mediatum. A deep learning architecture for limitedangle computed. The developed limited angle tomography technique was demonstrated to be able to detect practically significant manufacturing defects voids and structural damage. It has been studied for a variety of clinical applications, including vascular imaging, dental imaging, orthopedic imaging, mammographic imaging, musculoskeletal imaging, and chest imaging. This misplacement, however, could be minimized easily by routinely calibrating transducer positions. Pdf as a lowend computed tomography ct system, translational ct tct is in urgent demand in developing countries.
Another well known approach to limited angle reconstruction are regularized iterative algorithms. However, in the limited angle case the algorithm converges to a solution with typical limited angle artefacts. Computed tomography ct is commonly used in clinical procedures and limited angle tomography reconstruction has important applications in diagnostic ct, breast tomography, dental tomography, etc. Image prediction for limited angle tomography via deep learning with convolutional neural network hanming zhang 1, liang li2,3, 1kai qiao 1, linyuan wang, bin yan1, lei li, guoen hu 1 national digital switching system engineering and technological research center, zhengzhou, 450002, peoples republic of china 2 department ofengineering physics, tsinghua university, beijing, 84. The limited angle problem is a wellknown problem in computed tomography. Deep learning computed tomography pattern recognition lab. Characterization and reduction of artifacts in limited. By means of numerical simulations, it is verified that the algorithm developed ensured high quality of reconstruction. In application of tomography imaging, limited angle. Abstractthe papoulisgerchberg pg algorithm is widely used for extrapolation of band limited signals.
With this code we will be able to model any type of spill located at in any type of background material. Deep learning based image reconstruction algorithm for limited. Filters are investigated for reducing the artifact, and it. This thesis investigates the limited angle tomography problem where axial. The optimal scan acquisition procedure paired with the improved reconstruction algorithm facilitated higher inplane and depth resolution compared to the clinical application. Consequently, the use of the fbp algorithm is theoretically not justified in limited angle tomography. Classical reconstruction methods, such as the conventional filtered backprojection fbp algorithm and the algebraic reconstruction method arm 47, have well. In a limited angle ct scan, the projection data can be obtained in less than 180 angular range, and the data insuf. A related algorithm has been developed by the author and tested on electron microscope data from the karolinska institute and sidec technologies unpublished. Accurate computed tomography ct reconstruction from incomplete projections is an important research topic. Restoration of missing data in limited angle tomography. Unfortunately, viewing access in many tomographic experiments, such as in plasma physics, is extremely limited, which leads to a highly undetermined inversion problem. A deep learning architecture for limitedangle computed tomography reconstruction kerstin hammernik1, tobias wur 2, thomas pock1.
In this paper, we present comparative studies on exposure conditions in dbt, including narrow and wide angle scan using an invariant tube voltage of xray source, and wide angle scan with the voltage modulation technique. In daily routine this problem can arise for example in tomosynthesis, carm ct or dental ct. An iterative algorithm for computed tomography image reconstruction from limited angle projections yuli sun, jinxu tao, conggui liu department of electronic engineering and information science, university of science and technology of china, hefei 230027, peoples republic of china abstract. Limited angle tomography uct digital image processing.
Deep learning based image reconstruction algorithm for. Limited angle xray luminescence tomography has been developed as a means to image molecular deeper than is available with fluorescence molecular tomography. Cardiac xray computed tomography ct has been limited due to scanning times which are considerably longer 1 s than required to resolve the beating heart 0. Limited angle ultrasound tomography of the compressed breast. Optimization for limited angle tomography in medical image processing xiaoqiang lua, yi suna, yuan yuanb,n a school of electronic and information engineering, dalian university of technology, china b center for optical imagery analysis and learning optimal, state key laboratory of transient optics and photonics, xian institute of optics and precision mechanics. Limited angle ct reconstruction is an illposed inverse problem 3. Next, this dissertation aims to advance speed of sound, attenuation, and photoacoustic image reconstruction algorithms for the limited angle tomography approach. Sparse sampling and limited angle sampling are two effective ways to reduce the xray. Scalespace anisotropic total variation for limited angle. Wang 9 proposed a limitedangle ct image reconstruction algorithm based on the wavelet frame, and the reconstructed images show that it. An iterative algorithm for computed tomography image. The reconstruction problem is formulated as an optimization problem in a hilbert space setting and solved using landweber iteration.
We present a method for obtaining accurate image reconstruction from highly sparse data in diffraction tomography dt. Limitedangle computed tomography suffers from missing data in the. A new scale space total variation algorithm for limited. However, ct images reconstructed from limited angle acquisitions suffer from severe artifacts due to incomplete sinogram data. Tomosynthesis, also digital tomosynthesis dts, is a method for performing highresolution limited angle tomography at radiation dose levels comparable with projectional radiography. Improved backpropagation algorithms by exploiting data redundancy in limited angle di.
The first category of algorithms are to restore missing data based on. The resulting algorithm is implemented numerically in two spatial dimensions and tested on simulated data. Indeed, although in our work we limit the number of line projections, but do not limit the angles, the fundamental understanding of the relationship between a functions edge singularities and its radon representation, as explained in 10, is at the core of our algorithm see figure 3. Limited view tomography ct with limited angle data and few views irr algorithm iterative reconstructionreprojection irr. The illposedness of the limited angle tomography problem has been well investigated 1, 2. Standard tomographic algorithms applied to optical limited angle tomography result in the reconstructions that have highly anisotropic resolution and thus special algorithms are developed. In this paper, we propose an endtoend limited angle tomography. In this thesis, we consider both sparse and limited angle tomography situations. In the following section we outline the basic theory of diffraction tomography. One of the data sets is limited angle exterior data det with a range of 5.
A practical need exists for reconstruction from fewview and limited angle data, as this can greatly reduce required scan times in dt. A new approach to limited angle tomography using the. In the article iii, a new sinogram estimation algorithm called sinogram inter polation. Artifact reduction using the unet in limited angle tomography is such an example application. In this article, we develop a model based iterative reconstruction mbir algorithm for haadfstem tomography. A deeplearning method for limited angle tomography in synchrotron radiation transmission xray microscopies and a demonstration of its application in 3d visualization of a chlorella cell. Multiple limitedangles computed tomography reconstruction. Optical computed tomography oct is often used for measuring thermophysical parameters. Multiscale wavelet domain residual learning for limited. The principle depends on the geometry of the data set, and it applies. Limited angle computed tomography was studied in a project to develop algorithms for a limited angle scanner. Optimization for limited angle tomography in medical image. Unsupervised learnable sinogram inpainting network sin. In the sparse angle tomography cases we used only 10.
Iterative reconstruction algorithms can incorporate prior. A limited angle lambda tomography algorithm was given in 12. In the course of this thesis, various algorithms are proposed to deal with data insufficiency in limited angle tomography. In addition, we compared the conventional reconstruction methods with recently proposed idir algorithms. Ct, reconstructions from sparse views or limited angle tomography are. Existing iterative reconstruction algorithms could not reconstruct highquality images, leading to severe artifacts nearby edges. Papoulisgerchberg algorithms for limited angle tomography. In this paper, we will focus on the severely illposed inverse problem of limited angle computed tomography, in which entire boundary sections are not captured in the measurements. State of the art approaches utilize the total variation tv minimization technique. Image prediction for limitedangle tomography via deep.
Xdisystem multispectral source transmission detector. Image processing methods for limited angle tomography and. Since haadfstem tomography is a limited angle tomography modality with low signal to noise ratio, these methods can result in signi. An algorithm for limited data cardiaccomputed tomography by m.
888 19 801 568 628 127 829 449 399 222 815 109 1154 575 263 1351 519 1295 164 964 47 150 357 62 1462 541 469 1262 191 481 827 1146 922 1135 1116 99 469 1386