|
Article information
2017 , Volume 22, ¹ 1, p.25-36
Likhachev A.V.
Investigation of projection data recursive filtration for ROI-tomography problem
Purpose. The purpose of this paper is to improve the quality of images reconstructed from truncated data for ROI-tomography. Methodology. The proposed approach is to apply a new method of filteringdimensional projections, based on the imaging area of interest. The opportunity of replacing the ramp-filter digital recursive filter of the first order is investigated. Its frequency response is calculated on the basis of the Z-transform. Each truncated projection is presented as the production of full projection and the rectangular pulse of corresponding width. Its Fourier-image is determined according to the convolution theorem. The algorithm for the estimation of the digital filter coefficients is developed from the equality of frequency characteristics at certain points. The algorithm accounts the size of the area of the interest. The filter properties are studied by means of computational experiments. Findings. The impulse response of the recursive filter was found to be much narrower compared to that of a well-known approximation of the ramp-filter proposed by Shepp and Logan. It therefore provides less influence of unknown projections areas on the resulting image. Normalized root mean square error of reconstruction for it is 40-70 %. At the same time for the Shepp-Logan algorithm it reaches some hundreds percent when the size of the area of interest is small. The use of model data, containing additive Gaussian white noise, shows a high stability of the method. Originality. New adoptive algorithm for ROI-tomography is developed. The greatest effect from its usage is obtained when the area of interest is small relative to the investigated object. Under these conditions, the algorithm may significantly improve the reconstruction quality compared with standard tomographic algorithms.
[full text]
Keywords: two-dimensional tomography, region of interest, digital recursive filtering of projections
Author(s):
Likhachev Alexey Valerievich
Dr.
Position: Senior Research Scientist
Office: Institute of Automation and Electrometry Siberian Branch of Russian Academy of Science
Address: 630090, Russia, Novosibirsk, Academician Koptug ave. 1
Phone Office: (383) 330 82 43
E-mail: ipm1@iae.nsk.su
SPIN-code: 4283-4592 References:
[1] Sahine, B., Yagle, A.E. Region-of-interest tomography using exponential radial sampling. IEEE Transactions on Image Processing. 1995; 4(8):1120–1127.
[2] Sourbelle, K., Lauritsch, G., Tam, K.C., Noo, F., Kalender, W.A. Performance evaluation of local ROI algorithms for exact ROI reconstruction in spiral cone-beam computed tomography. Pt II. IEEE Transactions on Nuclear Science. 2001; 48(3):697–702.
[3] Lauritsch, G., Sourbelle, K., Tam, K.C. Optimization of derivative kernels for exact conebeam ROI reconstruction in spiral computed tomography. Pt I. IEEE Transactions on Nuclear Science. 2002; 49(3):728–732.
[4] Das, P.C., Sastry, Ch.S. Region-of-interest tomography using a composite Fourier-wavelet algorithm. Numerical Functional Analysis and Optimization. 2002; 23(7-8):757–777.
[5] Zeng, G.L., Gagnon, D., Natterer, F., Wenli, W., Wrinkler, M., Hawkins, W. Local tomography property of residual minimization reconstruction with planar integral data. IEEE Transactions on Nuclear Science. 2003; 50(5):1590–1594.
[6] Likhachov, A.V. ROI-tomography from the data distorted with variable dispersion noise. Computional Technologies. 2014; 19(2):62–75. (In Russ.)
[7] Likhachov, A.V. Use of the Cavalieri conditions in region-of-interest tomography. Optoelectronics, Instrumentation and Data Processing. 2015; 51(4):364-371.
[8] Herman, G.T. Image reconstruction from projections: The fundamentals of computerized tomography. New York: Academic Press; 1980: 316.
[9] Natterer, F. The mathematics of computerized tomography. Stuttgart: Wiley; 1986: 222.
[10] Shepp, L.A., Logan, B.F. The Fourier reconstruction of a head section. IEEE Transactions on Nuclear Science. 1974; 21(3):21–43.
[11] Lewitt, R.M. Reconstruction algorithms: transform methods. Proceedings of the IEEE. 1983; 71(3):390–408.
[12] Lavrent’ev, M.M., Zerkal, S.M., Trofimov, O.E. Computer modelling in tomography and ill-posed problems. Utrecht: VSP; 2001: 129.
[13] Likhachov, A.V. Investigation of 1/z2 filtration in tomography algorithms. Avtometriya. 2007; 43(3):57–64. (In Russ.)
[14] Likhachov, A.V. Double-filtration algorithm for two-dimensional tomography. Mathematical Models and Computer Simulations, 2010; 2(2):172–179
[15] Hamming, R.W. Digital Filters. Englewood Cliffs, NJ: Prentice-Hall; 1977: 197.
Bibliography link:
Likhachev A.V. Investigation of projection data recursive filtration for ROI-tomography problem // Computational technologies. 2017. V. 22. ¹ 1. P. 25-36
|
