X-ray or gamma ray systems or devices – Specific application – Absorption
Reexamination Certificate
2002-02-21
2003-09-09
Bruce, David V. (Department: 2882)
X-ray or gamma ray systems or devices
Specific application
Absorption
C378S007000, C378S098400, C378S901000
Reexamination Certificate
active
06618466
ABSTRACT:
FIELD OF THE INVENTION
The present invention is directed to a system and method for reduction of x-ray scatter in imaging systems and more particularly to such a system and method which use both spatial and temporal interpolation for such reduction. The present invention is usable with several imaging techniques, including fan beam and cone beam CT.
DESCRIPTION OF RELATED ART
Scatter reduction and correction are required for both medical (e.g., clinical, small animal imaging) and nonmedical imagine applications (e.g., explosive detection, nondestructive testing, industrial imagine guided manufacturing applications). Compton scatter interaction always exists in the x-ray energy range useful for imaging (10 keV to a few MeV). In some segments of the x-ray spectrum, X-ray scatter interaction is dominant. For example, within the effective energy range of medical diagnostic CT (60~80 keV), it is recognized that the Compton interaction (>80%) plays a predominant role over photoelectrical interaction (<10%) and coherent interaction (<5%) when x-ray photons pass through water-like soft tissues, although slightly higher percentages for both of them are observed when x-ray photons pass through compact bone. However, scatter intensity detected by a detector usually contributes to noise, not to a useful signal for imaging. Therefore, scatter reduction and correction are required to improve reconstruction accuracy of linear attenuation coefficient (LAC) distribution for both medical (clinical, small animal imaging and nonmedical imagine applications (explosive detection, nondestructive testing, industrial imagine guided manufacturing applications).
Before the mechanism in which x-ray scatter interferes with cone beam volume computed tomography (CBVCT) imaging is described, it is instructive to know how x-ray scattering deteriorates the quality of projection imaging.
FIG. 1A
shows an example in which a disc
102
is embedded in a uniform cylindrical object
104
whose LAC is slightly larger than that of the disc
102
. If the object
104
is disposed in a cone beam C emitted by a cone-beam source
106
, and the cone beam C is then received by a detector
108
, a projection image is acquired as shown in FIG.
1
B. Due to the LAC discrepancy, the number of x-ray photons detected in the area corresponding to the disc is N+&Dgr;N, but that detected in the background area is N, with no x-ray photons scattered. By limiting to an additive-noise-free situation, the ability to visualize such a structure is generally evaluated by contrast, which is defined as
C
=
N
+
Δ
N
-
N
N
=
Δ
N
N
(
1
)
It is known that the flux of x-ray photons observes the Poisson distribution, while the transmittance of x-ray photons through an object observes the binomial distribution. The cascading of a Poisson process and a binomial process is still a Poisson process. Hence, if the mean number of the x-ray photons transmitting the disc and its surrounding area are N+&Dgr;N and N respectively, their corresponding standard deviations are {square root over (N+&Dgr;N)} and {square root over (N)}. From the perspective of system analysis, the scattered x-ray photons behave like additive noise. In digital projection imaging where the display window of an ROI (region of interest) can be adjusted arbitrarily, the ability to visualize such a local structure is more appropriately measured by SNR (signal-to-noise ratio), which is defined as
SNR
=
N
+
Δ
N
-
N
N
=
Δ
N
N
=
C
N
(
2
)
On the other hand, letting the mean number of the scattered x-ray photons be N
s
, the scatter-to-primary ratio (SPR) is defined as
SPR
=
N
s
N
(
3
)
and the scatter degradation factor (SDF) is defined as
SDF
=
N
N
+
N
s
=
1
1
+
N
s
/
N
=
1
1
+
SPR
(
4
)
Consequently, (1) and (2) respectively degrade to
C
s
=
N
+
N
s
+
Δ
N
-
(
N
+
N
s
)
N
+
N
s
=
Δ
N
N
+
N
s
=
C
1
+
N
s
/
N
=
SDF
·
C
(
5
)
SNR
s
=
N
+
N
s
+
Δ
N
-
(
N
+
N
s
)
N
+
N
s
=
Δ
N
N
+
N
s
=
SNR
1
+
N
s
/
N
=
SDF
·
SNR
(
6
)
in the presence of x-ray scatter. In other words, due to scattered x-ray photons, the local contrast and SNR of the structure are deteriorated by the factors of SDF and {square root over (SDF)} in a projection image, respectively.
Suppose that I
p
(i,j)(i&egr;I,j&egr;J) refers to the image formed by the primary x-ray photons, and I
s
(i,j)(i&egr;I,j&egr;J) the image formed by the scattered photons, where I and J are the vertical and horizontal dimension of a projection image. As a result, a 2D SPR distribution is defined as
SPR
(
i
,
j
)
=
I
s
(
i
,
j
)
I
p
(
i
,
j
)
(
i
∈
I
,
j
∈
J
)
(
7
)
and subsequently
SDF
(
i
,
j
)
=
1
1
+
SPR
(
i
,
j
)
(
i
∈
I
,
j
∈
J
)
(
8
)
It has been found that, given an object, the distribution of I
s
(i,j) is dependent on its structure, thickness and field of view (FOV) through which I
p
(i,j) is formed. However, regardless of how the distribution of I
p
(i,j) fluctuates, the variation of I
s
(i,j) is so smooth that it can be approximately treated as a 2D spatial low-pass filtering of I
p
(i,J), and several low-pass filtering models associated with different filter kernels have been proposed. That means that a strong spatial correlation exists between neighboring pixels. Intuitively, scatter distribution in each projection image can be recovered from its spatial samples using either interpolation methods, such as cubic spline interpolation or bi-linear interpolation methods, or a convolution operation with a selected convolution kernel (a low pass filtering in the frequency domain).
On the other hand, the SPR(i,j) or SDF(i,j) usually fluctuates very much, especially where I
p
(i,j) is relatively low. Hence, the SPR(i,j) or SDF(i,j) is usually taken into account to reflect the severity of the x-ray scatter in projection imaging.
A CBVCT (cone beam volume computed tomography) image is reconstructed from a set of consecutive 2D projection images that are sequentially acquired. Intrinsically, the x-ray scatter interferes with the X-ray transform non-linearly. Once the X-ray transform is acquired, the artifact caused by the x-ray scatter in a tomographic image is reconstruction-algorithm-dependent. Thus, the experimental investigation of the x-ray scatter artifact in tomographic imaging is preferable in practice. In conventional CT (computed tomography), due to the adoption of a slit collimator, the severity of the longitudinally scattered x-ray photons is reduced to a secondary order in comparison to that of the transversely scattered ones. To further reduce the transversely scattered x-ray, other measures, such as the bow-tie x-ray attenuator, and the post-patient out-of-slice collimator used in a 3
rd
-generation CT or the reference detector used in a 4
th
-generation CT, are usually taken into account.
Unfortunately, for CBVCT, the slit collimator has to be removed to fully use the generated cone shaped x-ray beam. Hence, much more severe x-ray scatter interference is expected in CBVCT, although the air gap technique and beam-shaping (bow-tie) attenuator are still useful for reducing scatter. Supposing I
p
(x,y) refers to the projection image formed by primary x-ray photons, and I
s
(x,y) that formed by scattered x-ray photons, we have
I
t
(
x,y
)=
I
p
(
x,y
)+
I
s
(
x,y
) (9)
As a result, a 2D SPR distribution is defined as
SPR
(
x
,
y
)
=
I
s
(
x
,
y
)
I
p
(
x
,
y
)
(
10
)
By substituting (10) into (9), one gets
I
t
(
x,y
)=
I
p
(
x,y
)[1.0
+SPR
(
x,y
)] (11)
Furthermore, by taking the incident x-ray intensity distribution I
0
(x,y) into account and applying logarithm, we get the X-ray transform data for CB reconstruction
P
(
x
,
y
)
=
ln
I
0
(
x
,
y
)
I
i
(
x
,
y
)
=
ln
I
0
(
x
,
y
)
I
p
(
x
,
y
)
[
1.0
+
SPR
&af
Blank Rome LLP
Bruce David V.
University of Rochester
LandOfFree
Apparatus and method for x-ray scatter reduction and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for x-ray scatter reduction and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for x-ray scatter reduction and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3050906