X-ray or gamma ray systems or devices – Specific application – Absorption
Patent
1998-10-15
2000-08-15
Bruce, David V.
X-ray or gamma ray systems or devices
Specific application
Absorption
378 87, 378 98, G01N 2304
Patent
active
061047775
DESCRIPTION:
BRIEF SUMMARY
TECHNICAL FIELD
The present invention relates to a process for correcting scatter energy in digital X-ray images.
STATE OF THE PRIOR ART
The different known processes of the prior art for the correction of scatter energy are; approximation with an average equivalent plane.
Convolution Methods
The first scatter correction methods using image processing are convolution methods. They are based on the following empirical model observed transmission image;
This model is not derived from physical considerations, but only from the finding that scatter appears as a "blurring" phenomenon in the image. This model was postulated by Shaw et al in 1982 with a fixed kernel and a constant weighting factor w, for the entirety of the scatter image. This model was used and improved over the following years: variable factor w in relation to the thickness of the object, based on the remark that a "thicker" object produces more scatter (see document reference [1] at the end of this disclosure). variable per region, in an attempt to give consideration to the non-homogeneity of objects (often made up of several different materials). With this variability it is possible, for example in chest radiography, to have a kernel f.sub.s1 corresponding to the highly scattering central zone (spine, heart . . . ) and a kernel f.sub.s2 for the less scattering lung areas (see document reference [2]). correction method based on this same empirical principle of convolution of observed transmission, but using convolution kernels whose size was dependent upon the thickness of compression at the time of acquisition.
These kernels are estimated by calibration on phantoms of variable thickness. The protocol for mammography imaging allows easy access to this parameter, since the breast is compressed between two parallel plates (see document referenced [3]).
Deconvolution Methods
The second family of scatter correction methods by image processing relates to the so-called "deconvolution" methods, the empirical model being the same as that used in the previous methods. But in this case, primary transmission is not approximated with observed transmission but the following equation is used:
Determination of I.sub.p, and consequently determination of I.sub.s, requires an inversion of the equation, that is to say deconvolution. This is generally made using a Fourier transform. convolution kernel of exp(-b.vertline.r.vertline.) type. The choice of a and b is made by calibration on a chest phantom (see document referenced [4]). referenced [5]) but with a convolution kernel of Gaussian distribution type with the form: ##EQU1## .rho. being the scattered photon fraction, r being the radial distance in polar co-ordinates of predetermined origin, and .sigma. being the extent of Gaussian distribution.
The two parameters .rho. and .sigma. are also obtained by phantom calibration.
The Analytical Method
This method differs largely from the preceding methods. It is the result of equating the physical phenomenon which is the cause of the creation of scatter. By construction this method gives consideration to the different materials in the imaged object, these materials having an effect on the generation and attenuation of scattered photons.
The main obstacle with this type of approach results from lack of knowledge of the three-dimensional structure of the object; all that is available is the image of observed transission which only represents a projection of this structure.
In 1988 Boone and Seibert suggested an approach of analytical type which equated the first Compton scattering, restricted to homogeneous objects of constant thickness. In this particular case, the scatter phenomenon may be modelled by convolution of primary transmission with a kernel derived from analytical calculations (see document reference [8]).
In 1992, C. Burq suggested an approach of this type in which she equated the first Compton scattering using the physical laws governing the scatter phenomenon (see document referenced [6]). To solve the problem of the unknown three-dimension
REFERENCES:
patent: 5440647 (1995-08-01), Floyd, Jr. et al.
patent: 5862199 (1999-01-01), MacKenzie
1991 IEEE Nuclear Science Symposium & Medical Imaging Conference, vol. 3, Nov. 2-9, 1991, Santa Fe, NM USA, "Solid Geometry Based Modelling of Non-Uniform Attenuation and Compton Scattering in Objects for Spect Imaging Systems" Wang, H. et al.
Medical Physics, vol. 22, No. 10, Oct. 1, 1995 "A Transmission-Map-Based Scatter Connection Technique for Spect in Inhomogeneous Media" Welch, A. et al.
IEEE Nuclear Science Symposium, vol. 2, Nov. 2-9, 1996 "A New Method for Modelling the Spatially-Variant, Object-Dependent Scatter Response Function in Spect" Frey .E.C. et al.
Darboux Michel
Dinten Jean-Marc
Bruce David V.
Commissariat a l''Energie Atomique
Lieberstein Eugene
Meller Michael N.
LandOfFree
Process for the correction of scattering in digital X-ray images does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Process for the correction of scattering in digital X-ray images, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Process for the correction of scattering in digital X-ray images will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2015540