U.S. patent application number 14/471989 was filed with the patent office on 2016-03-03 for method and apparatus for estimating scatter in a positron emission tomography scan at multiple bed positions.
This patent application is currently assigned to Kabushiki Kaisha Toshiba. The applicant listed for this patent is Kabushiki Kaisha Toshiba, Toshiba Medical Systems Corporation. Invention is credited to Xiaofeng NIU, Wenli WANG, Hongwei YE.
Application Number | 20160063741 14/471989 |
Document ID | / |
Family ID | 55403090 |
Filed Date | 2016-03-03 |
United States Patent
Application |
20160063741 |
Kind Code |
A1 |
YE; Hongwei ; et
al. |
March 3, 2016 |
Method and Apparatus for Estimating Scatter in a Positron Emission
Tomography Scan at Multiple Bed Positions
Abstract
A method is provided for estimating scatter in a positron
emission tomography (PET) scan at multiple bed positions, the
method comprising calculating a first scatter sinogram based on
scatter data obtained at a first bed position, calculating a second
scatter sinogram based on scatter data obtained at a second bed
position, and deriving a third scatter sinogram for a third bed
position between the first bed position and the second bed
position, wherein the third scatter sinogram is derived from the
first scatter sinogram according to a first percentage of overlap
of the first bed position with the third bed position, and from the
second scatter sinogram according to a second percentage of overlap
of the second bed position with the third bed position
Inventors: |
YE; Hongwei; (Kenosha,
WI) ; NIU; Xiaofeng; (Mundelein, IL) ; WANG;
Wenli; (Briarcliff Manor, NY) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Kabushiki Kaisha Toshiba
Toshiba Medical Systems Corporation |
Minato-ku
Otawara-shi |
|
JP
JP |
|
|
Assignee: |
Kabushiki Kaisha Toshiba
Minato-ku
JP
Toshiba Medical System Corporation
Otawara-shi
JP
|
Family ID: |
55403090 |
Appl. No.: |
14/471989 |
Filed: |
August 28, 2014 |
Current U.S.
Class: |
382/131 |
Current CPC
Class: |
A61B 6/037 20130101;
G01T 1/2985 20130101; A61B 6/5282 20130101; G06T 11/005
20130101 |
International
Class: |
G06T 11/00 20060101
G06T011/00; G01T 1/29 20060101 G01T001/29 |
Claims
1. A method for estimating scatter in a positron emission
tomography (PET) scan at multiple bed positions, the method
comprising: calculating a first scatter sinogram based on scatter
data obtained at a first bed position; calculating a second scatter
sinogram based on scatter data obtained at a second bed position;
and deriving a third scatter sinogram for a third bed position
between the first bed position and the second bed position, wherein
the third scatter sinogram is derived from the first scatter
sinogram according to a first percentage of overlap of the first
bed position with the third bed position, and from the second
scatter sinogram according to a second percentage of overlap of the
second bed position with the third bed position.
2. The method according to claim 1, wherein the step of deriving
the third scatter sinogram comprises: determining a first portion
to copy, the first portion being equal to the first percentage of
the first scatter sinogram; determining a second portion to copy,
the second portion being equal to the second percentage of the
second scatter sinogram; and copying the first portion and the
second portion to the third scatter sinogram.
3. The method according to claim 2, wherein the first scatter
sinogram, the second scatter sinogram and the third scatter
sinogram have the same dimensions as one another, and when a sum of
the first percentage of overlap and the second percentage of
overlap is less than 100%, the step of deriving the third scatter
sinogram further comprises: determining a remaining portion of the
third scatter sinogram, the remaining portion having an area that
is equal to a difference between an area of the third scatter
sinogram and a sum of an area of the first portion and an area of
the second portion; interpolating the remaining portion of the
third scatter sinogram to create an interpolated portion; and
copying the interpolated portion to the third scatter sinogram,
wherein the third scatter sinogram includes the first portion, the
second portion, and the interpolated portion.
4. The method according to claim 2, wherein the first scatter
sinogram, the second scatter sinogram and the third scatter
sinogram have the same dimensions as one another, and when a sum of
the first percentage of overlap and the second percentage of
overlap is greater than 100%, the step of deriving the third
scatter sinogram further comprises: determining an overlapping
portion of the third scatter sinogram, the overlapping portion
including a first part of the first portion that overlaps with a
second part of the second portion; averaging the first part and the
second part of the overlapping portion to create an averaged
portion; and copying the averaged portion to the third scatter
sinogram, wherein the third scatter sinogram includes the averaged
portion, a first remaining part of the first portion that does not
include the first part, and a second remaining part of the second
portion that does not include the second part.
5. The method according to claim 2, wherein the step of copying the
first portion and the second portion to the third scatter sinogram
comprises: determining a first position for the first portion in
the third scatter sinogram; determining a second position for the
second portion in the third scatter sinogram; and copying the first
portion to the first position and the second portion to the second
position, wherein a size of the third scatter sinogram is equal to
a size of the first scatter sinogram and equal to a size of the
second scatter sinogram.
6. The method according to claim 3, wherein the step of copying the
interpolated portion to the third scatter sinogram comprises:
determining a position for the interpolated portion in the third
scatter sinogram, the position of the interpolated portion being
determined based on positions of the first portion and the second
portion; and copying the interpolated portion to the position in
the third scatter sinogram.
7. The method according to claim 4, wherein the step of copying the
averaged portion to the third scatter sinogram comprises:
determining a position for the averaged portion in the third
scatter sinogram, the position of the averaged portion being
determined based on positions of the first portion and the second
portion; and copying the averaged portion to the position in the
third scatter sinogram.
8. The method according to claim 1, wherein the first scatter
sinogram and the second scatter sinogram are calculated using a
model-based scatter estimation (MBSE), a background subtraction
method or a convolution subtraction method.
9. The method according to claim 1, wherein the first scatter
sinogram and the second scatter sinogram are calculated using Monte
Carlo-based scatter estimation.
10. The method according to claim 2, wherein the step of deriving
the third scatter sinogram further includes rescaling the third
scatter sinogram for the third bed position.
11. The method according to claim 1, wherein the PET scan is
performed with continuous bed movement.
12. The method according to claim 1, wherein the PET scan is
performed using step-and-shoot bed movement.
13. The method according to claim 1, further comprising:
determining each bed position within the multiple bed positions of
the PET scan.
14. The method according to claim 1, further comprising:
determining the first percentage of overlap; and determining the
second percentage of overlap.
15. A non-transitory computer readable medium storing computer
executable instructions that, when executed by a processor, cause
the processor to perform a method for estimating scatter in a
positron emission tomography (PET) scan at multiple bed positions,
the method comprising, the method comprising: calculating a first
scatter sinogram based on scatter data obtained at a first bed
position; calculating a second scatter sinogram based on scatter
data obtained at a second bed position; and deriving a third
scatter sinogram for a third bed position between the first bed
position and the second bed position, wherein the third scatter
sinogram is derived from the first scatter sinogram according to a
first percentage of overlap of the first bed position with the
third bed position, and from the second scatter sinogram according
to a second percentage of overlap of the second bed position with
the third bed position.
16. An apparatus for estimating scatter in a positron emission
tomography (PET) scan at multiple bed positions, the apparatus
comprising: processing circuitry configured to calculate a first
scatter sinogram based on scatter data obtained at a first bed
position; calculate a second scatter sinogram based on scatter data
obtained at a second bed position; and derive a third scatter
sinogram for a third bed position between the first bed position
and the second bed position, wherein the processing circuitry
derives the third scatter sinogram from the first scatter sinogram
according to a first percentage of overlap of the first bed
position with the third bed position, and from the second scatter
sinogram according to a second percentage of overlap of the second
bed position with the third bed position.
17. The apparatus according to claim 16, wherein the processing
circuitry is further configured to determine a first portion to
copy, the first portion being from the first scatter sinogram;
determine a second portion to copy, the second portion being from
the second scatter sinogram; and copy the first portion and the
second portion to the third scatter sinogram.
18. The apparatus according to claim 17, wherein the first scatter
sinogram, the second scatter sinogram and the third scatter
sinogram have the same dimensions as one another, and when a sum of
the first percentage of overlap and the second percentage of
overlap is less than 100%, the processing circuitry is further
configured to determine a remaining portion of the third scatter
sinogram, the remaining portion having an area that is equal to a
difference between an area of the third scatter sinogram and a sum
of an area of the first portion and an area of the second portion;
interpolate the remaining portion of the third scatter sinogram to
create an interpolated portion; and copy the interpolated portion
to the third scatter sinogram, wherein the third scatter sinogram
includes the first portion, the second portion, and the
interpolated portion.
19. The apparatus according to claim 17, wherein the first scatter
sinogram, the second scatter sinogram and the third scatter
sinogram have the same dimensions as one another, and when a sum of
the first percentage of overlap and the second percentage of
overlap is greater than 100%, the processing circuitry is further
configured to determine an overlapping portion of the third scatter
sinogram, the overlapping portion including a first part of the
first portion that overlaps with a second part of the second
portion; average the first part and the second part of the
overlapping portion to create an averaged portion; and copy the
averaged portion to the third scatter sinogram, wherein the third
scatter sinogram includes the averaged portion, a first remaining
part of the first portion that does not include the first part, and
a second remaining part of the second portion that does not include
the second part.
20. The apparatus according to claim 17, wherein the processing
circuitry is further configured to determine a first position for
the first portion in the third scatter sinogram; determine a second
position for the second portion in the third scatter sinogram; and
copy the first portion at the first position and the second portion
at the second position, wherein a size of the third scatter
sinogram is equal to a size of the first scatter sinogram and equal
to a size of the second scatter sinogram.
Description
FIELD
[0001] The disclosed embodiments relate generally to scatter
estimation in a positron emission tomography (PET) scan, and in
particular, to a method for estimating scatter in a PET scan at
multiple bed positions.
BACKGROUND
[0002] In three-dimensional PET scans, scatter is one of the most
significant physical effects relating to the degradation of image
quality. In typical PET systems, scatter events can be as much as
30%.about.50% of the total detected events in a PET scan. Image
quality can be improved by correcting scatter events before or
during image reconstruction.
[0003] There are several approaches for the correction of scatter
events. Such approaches include a background subtraction or
tail-fitting method, a convolution subtraction method, a Monte
Carlo-based method, or a model-based scatter estimation (MBSE)
including single scatter simulation (SSS). MBSE is a popular method
used in modern PET systems and provides good scatter
correction.
[0004] For a PET scan in a multi-bed position scanning system, bed
positions that are adjacent to one another typically have at least
20% area overlap so as to achieve a more uniform axial sensitivity.
To achieve good scatter estimation, MBSE requires the collection of
scan data from bed positions that are adjacent to one another in
order to estimate scatter that is out of the axial field of view
(FOV).
[0005] However, extensive calculations are required when estimating
scatter using the MBSE method. Even when a PET scan system has
extremely high processing power with high optimization, scatter
estimation suing the MBSE method can still take a long period of
time because of the extensive calculations required. For example, a
PET scan system for estimating scatter for typical patient data
with 8 bed positions using a single 3.3 GHz CPU can take
approximately 2500.about.4300 seconds/bed position, or 450
minutes.
[0006] As a result, it can be beneficial to reduce the processing
time necessary to acquire reliable scatter estimation for the
correction of scatter data and the reconstruction of PET data to
improve image quality.
BRIEF DESCRIPTION OF THE DRAWINGS
[0007] A more complete appreciation of the invention and many of
the attendant advantages thereof will be readily obtained as the
same becomes better understood by reference to the following
detailed description when considered in connection with the
accompanying drawings, wherein:
[0008] FIG. 1(A) illustrates an exemplary PET scanning system;
[0009] FIG. 1(B) illustrates exemplary detection of single scatter
events by PET ring 20;
[0010] FIG. 2(A) illustrates an axial view of PET scanning system
10 including a plurality of PET rings 20 and a relationship between
axial sensitivity of PET scanning system 10 and an axial distance
of PET ring 20;
[0011] FIG. 2(B) illustrates bed positions having an overlap of 50%
as bed 30 moves in a bed movement direction through PET scanning
system 10 and a relationship between overlapped sensitivity of PET
scanning system 10 and an axial distance of PET ring 20;
[0012] FIG. 3 illustrates method 100 for estimating scatter in
accordance with a plurality of bed positions;
[0013] FIG. 4 illustrates bed positions of bed 30 as moves through
PET scanning system 10 in the bed movement direction;
[0014] FIG. 5 provides details of steps performed in Step 150 of
method 100;
[0015] FIG. 6(A) illustrates an interpolated sinogram;
[0016] FIG. 6(B) illustrates the coordinate system of an
interpolated sinogram in a transaxial view (left) and an axial view
(right);
[0017] FIG. 7(A) illustrates lines of response (LORs) of an
interpolated sinogram of 700-mm-FOV and an amount of overlap of 50%
with varying B's ranging from -15.3 to 15.3 degrees;
[0018] FIG. 7(B) illustrates bed positions with 4 different
.theta.'s with an amount of overlap of 50%;
[0019] FIG. 7(C) illustrates LORs of a bed position with
.theta.=-10.2;
[0020] FIG. 8 illustrates a calculation percentage of copied LORs
in an overlapping bed position;
[0021] FIG. 9(A) illustrates adjacent bed positions when an amount
of overlap is greater than 50% (left) and when an amount of overlap
is less than 50% (right);
[0022] FIG. 9(B) illustrates a weight function w(z) for weighting
same LORs from different bed positions;
[0023] FIG. 10 illustrates an exemplary component configuration of
PET scanning system 10;
[0024] FIG. 11(A) illustrates a traversal view of a reconstructed
image with scatter estimation by MBSE;
[0025] FIG. 11(B) illustrates a traversal view of a reconstructed
image with scatter estimation by MBSE utilizing according to the
proposed method of FIG. 3;
[0026] FIG. 12(A) illustrates a sagittal view of a reconstructed
image with scatter estimation by MBSE; and
[0027] FIG. 12(B) illustrates a sagittal view of a reconstructed
image with scatter estimation by MBSE utilizing according to the
proposed method of FIG. 3.
DETAILED DESCRIPTION
[0028] According to one embodiment, there is provided a method for
estimating scatter in a positron emission tomography (PET) scan at
multiple bed positions, the method comprising calculating a first
scatter sinogram based on scatter data obtained at a first bed
position, calculating a second scatter sinogram based on scatter
data obtained at a second bed position, and deriving a third
scatter sinogram for a third bed position between the first bed
position and the second bed position, wherein the third scatter
sinogram is derived from the first scatter sinogram according to a
first percentage of overlap of the first bed position with the
third bed position, and from the second scatter sinogram according
to a second percentage of overlap of the second bed position with
the third bed position.
[0029] In another embodiment, the step of deriving the third
scatter sinogram comprises determining a first portion to copy, the
first portion being equal to the first percentage of the first
scatter sinogram, determining a second portion to copy, the second
portion being equal to the second percentage of the second scatter
sinogram, and copying the first portion and the second portion to
the third scatter sinogram.
[0030] In another embodiment, the first scatter sinogram, the
second scatter sinogram and the third scatter sinogram have the
same dimensions as one another, and when a sum of the first
percentage of overlap and the second percentage of overlap is less
than 100%, the step of deriving the third scatter sinogram further
comprises determining a remaining portion of the third scatter
sinogram, the remaining portion having an area that is equal to a
difference between an area of the third scatter sinogram and a sum
of an area of the first portion and an area of the second portion,
interpolating the remaining portion of the third scatter sinogram
to create an interpolated portion, and copying the interpolated
portion to the third scatter sinogram, wherein the third scatter
sinogram includes the first portion, the second portion, and the
interpolated portion.
[0031] In another embodiment, the first scatter sinogram, the
second scatter sinogram and the third scatter sinogram have the
same dimensions as one another, and when a sum of the first
percentage of overlap and the second percentage of overlap is
greater than 100%, the step of deriving the third scatter sinogram
further comprises determining an overlapping portion of the third
scatter sinogram, the overlapping portion including a first part of
the first portion that overlaps with a second part of the second
portion, averaging the first part and the second part of the
overlapping portion to create an averaged portion, and copying the
averaged portion to the third scatter sinogram, wherein the third
scatter sinogram includes the averaged portion, a first remaining
part of the first portion that does not include the first part, and
a second remaining part of the second portion that does not include
the second part.
[0032] In another embodiment, the step of copying the first portion
and the second portion to the third scatter sinogram comprises
determining a first position for the first portion in the third
scatter sinogram, determining a second position for the second
portion in the third scatter sinogram, and copying the first
portion to the first position and the second portion to the second
position, wherein a size of the third scatter sinogram is equal to
a size of the first scatter sinogram and equal to a size of the
second scatter sinogram.
[0033] In another embodiment, the step of copying the interpolated
portion to the third scatter sinogram comprises determining a
position for the interpolated portion in the third scatter
sinogram, the position of the interpolated portion being determined
based on positions of the first portion and the second portion, and
copying the interpolated portion to the position in the third
scatter sinogram.
[0034] In another embodiment, the step of copying the averaged
portion to the third scatter sinogram comprises determining a
position for the averaged portion in the third scatter sinogram,
the position of the averaged portion being determined based on
positions of the first portion and the second portion, and copying
the averaged portion to the position in the third scatter
sinogram.
[0035] In another embodiment, the first scatter sinogram and the
second scatter sinogram are calculated using model-based scatter
estimation (MBSE).
[0036] In another embodiment, the first scatter sinogram and the
second scatter sinogram are calculated using Monte Carlo-based
scatter estimation.
[0037] In another embodiment, the step of deriving the third
scatter sinogram further resealing the third scatter sinogram for
the third bed position.
[0038] In another embodiment, the PET scan is performed with
continuous bed movement.
[0039] In another embodiment, the PET scan is performed using
step-and-shoot bed movement.
[0040] In another embodiment, the method further comprises
determining each bed position within the multiple bed positions of
the PET scan.
[0041] In another embodiment, the method further comprises
determining the first percentage of overlap, and determining the
second percentage of overlap.
[0042] Referring now to the drawings, wherein like reference
numerals designate identical or corresponding parts throughout the
several views, FIG. 1(A) illustrates an exemplary PET scanning
system. FIG. 1(B) illustrates exemplary detection of single scatter
events in a PET ring.
[0043] As illustrated in FIG. 1(A), PET scanning system 10 includes
PET ring 20 and bed 30. A scanning target, typically a patient,
lies upon bed 30. Bed 30 moves through PET ring 20 in a bed
movement direction. While FIG. 1(A) illustrates PET scanning system
10 that includes a single PET ring 20, PET scanning system 10 can
include a plurality of PET ring 20.
[0044] As bed 30 moves through PET ring 20, position emission
events are measured by PET ring 20 by detecting photons that are
emitted when positrons and electrons collide and annihilate one
another. Specifically, detectors 22A and 22B in PET ring 20 detect
and measure the emission of photons as bed 30 moves through PET
ring 20 of PET scanning system 10. Although FIG. 1(B) illustrates
detectors 22A and 22B, each PET ring 20 can include more than two
detectors 22.
[0045] In exemplary use of PET scanning system 10, bed 30 moves
through PET ring 20 with a continuous movement motion, i.e.,
maintaining a constant speed. In such continuous motion, scan data
is continuously collected by PET ring 20. Alternatively, bed 30
moves through PET scanning system 10 with a step-and-shoot motion,
i.e., bed 30 stops at specific bed positions, stopping at each bed
position for a predetermined period of time. With such a
step-and-shoot motion, PET ring 20 collects scan data at each
specific bed position. In other words, PET scanning system 10
collects scan data using either step-and-shoot or continuous bed
movement.
[0046] Further, PET scanning system 10 can estimate scatter events
using a time-of-flight (TOF) estimation method or a non-TOF
estimation method.
[0047] All emission events detected by PET scanning system 10 are
collected as bed 30 moves through PET scanning system 10. Once the
PET scan is complete, PET scanning system 10 processes the
collected scan data to reconstructs an image from the collected
scan data by correcting the scatter data.
[0048] To correct the scatter data, PET scanning system 10
estimates scatter by using the MBSE method. Alternatively, PET
scanning system 10 can estimate scatter by using methods such as
the Monte Carlo scatter estimation method, background subtraction
method or convolution subtraction method.
[0049] The MBSE method considers and models the physics of Compton
scattering in a system by a mathematical model to calculate single
scatter events, and a typical example is SSS. In one embodiment,
PET scanning system 10 calculates single scatter events utilizing
physical effects such as inter-crystal scattering and photo
penetration in crystals, positron ranges, non-colinearity, multiple
scatter events, accurate photon attenuation, etc. However, an
exemplary embodiment of PET scanning system 10 reduces
computational time by solely calculating scatter events. To
compensate for the difference between model and real systems, a
tail-fitting or other method is applied.
[0050] In an exemplary embodiment, PET scanning system 10 utilizes
known SSS formula to calculate single scatter events.
[0051] To accurately collect scan data, bed positions can
theoretically overlap anywhere from 0% to 100%. In other words, bed
positions do not overlap at a value of 0%, and bed positions
completely overlap at a value of 100%. Practically, PET scanning
system 10 benefits from an existence of at least some bed position
overlap, i.e. an overlap of more than 0%, by an increase in uniform
axial sensitivity.
[0052] FIG. 2(A) illustrates an axial view of PET scanning system
10 including a plurality of PET ring 20 and a relationship between
axial sensitivity of PET scanning system 10 and an axial distance
of PET ring 20.
[0053] As illustrated in FIG. 2(A), axial sensitivity of PET
scanning system 10 is greater towards the middle of the plurality
of PET ring 20 and axial sensitivity linearly decreases away from
the middle and towards either end of the plurality of PET ring 20.
As a result, uniform axial sensitivity of scan data can be obtained
by overlapping scans taken at adjacent bed positions, as
illustrated in FIG. 2(B).
[0054] FIG. 2(B) illustrates bed positions having an overlap of 50%
as bed 30 moves in the bed movement direction through PET scanning
system 10. Scan data is collected at each bed position by PET
scanning system 10 of the scan target, a patient, upon bed 30 as
bed 30 moves through PET scanning system 10. Because the bed
positions overlap, sensitivity of PET scanning system 10 is
increased. FIG. 2(B) illustrates a relationship between an
overlapped sensitivity of PET scanning system 10 and an axial
distance of PET ring 20. In FIG. 2(B), axial sensitivity is
increased at positions closer to an edge of the plurality of PET
ring 20 by acquiring scan data at overlapping adjacent bed
positions. While FIG. 2(B) illustrates bed positions having an
overlap of 50%, PET scanning system 10 can operate with an overlap
of bed positions with a value anywhere between 0% and 100%.
[0055] While sensitivity is increased as the amount of bed position
overlap increases, scan speed decreases as the amount of bed
position overlap increases. Alternatively, scan speed is increased
when an amount of bed position overlap decreases.
[0056] After scan data is collected by PET scanning system 10, PET
scanning system 10 processes the collected scan data to identify
scatter events. Because bed positions overlap, much of the scan
data collected by PET scanning system 10 can be duplicative. In
other words, calculation of scatter data at each bed position can
be unnecessary because the scan data collected from a first bed
position is the same as scan data collected at a second bed
position that overlaps with the first bed position. As a result,
PET scanning system 10 performs method 100 illustrated in FIG. 3 to
decrease the computational time in the scatter estimation, but keep
the similar image quality by utilizing the overlaps between
adjacent bed positions.
[0057] FIGS. 3 and 4 will be referenced to describe method 100
performed by PET scanning system 10. In particular, FIG. 3
illustrates method 100 for estimating scatter in accordance with a
plurality of bed positions. FIG. 4 illustrates bed positions of bed
30 as bed 30 moves through PET scanning system 10 in the bed
movement direction.
[0058] Bed 30 moves through PET scanning system 10 in the bed
movement direction illustrated in FIG. 1(A). Prior to performance
of method 100, scan data is collected by PET ring 20 as bed 30
moves through PET scanning system 10. Alternatively, method 100 is
performed by PET scanning system 10 on scan data stored in a
memory, e.g., data that has been previously collected.
[0059] In method 100, PET scanning system 10 performs method 100 by
beginning at Step 110.
[0060] In Step 110, PET scanning system 10 determines a number of
bed positions. In exemplary implementations, the number of bed
positions is a preset amount. Alternatively, the number of bed
positions is adjusted according to a desired sensitivity or
scanning speed. Further, the number of bed positions can be input
or set by a user. After determining the number of bed positions,
PET scanning system 10 proceeds to Step 120.
[0061] In Step 120, PET scanning system 10 determines an amount of
overlap between each bed position. In exemplary implementations,
the amount of overlap between adjacent bed positions is a preset
amount. Alternatively, the amount of overlap between adjacent bed
positions can depend upon any of the number of bed positions, a
length of bed 30, and a size of a scanning area of each bed
position. Further, the amount of overlap can be input or set by a
user. After determining the amount of overlap, PET scanning system
proceeds to Step 130.
[0062] In Step 130, PET scanning system 10 calculates a first
scatter sinogram. In an exemplary implementation, PET scanning
system 10 calculates the first scatter sinogram by processing the
scan data collected at the first bed position illustrated in FIG.
4. Then in Step 140, PET scanning system 10 calculates a second
scatter sinogram by processing the scan data collected at the
second bed position illustrated in FIG. 4. The scan data collected
within the first bed position is used to calculate the first
scatter sinogram and the scan data collected within the second bed
position is used to calculate the second scatter sinogram.
[0063] In Step 150, PET scanning system 10 derives a third scatter
sinogram. In Step 150, PET scanning system 10 does not derive the
third scatter sinogram by processing the scan data collected at a
third bed position. Instead, PET scanning system 10 derives a third
scatter sinogram for the third position based on the first scatter
sinogram and the second scatter sinogram.
[0064] In the exemplary implementation as illustrated in FIG. 4,
the first bed position has a first amount of overlap of 50% with
the third bed position and the second bed position has a second
amount of overlap of 50% with the third bed position. Because of
the respective overlaps of the first, second and third bed
positions, the scatter data collected in the third bed position can
be substantially the same as scatter data collected in a portion of
the first bed position and in a portion of the second bed
position.
[0065] As illustrated in FIG. 4, the third scatter sinogram is
derived by PET scanning system 10 in Step 150 by copying portion A
of the first scatter sinogram calculated from scatter data
collected at the first bed position, and copying portion B of the
second scatter sinogram calculated from scatter data collected at
the second bed position. Thus, PET scanning system 10 derives the
third scatter sinogram in Step 150 by copying portion A of the
first scatter sinogram and copying portion B of the second scatter
sinogram.
[0066] Note that error in the derived scatter sinograms increases
with a decrease of an amount of overlap between adjacent bed
positions. Further detail of the derivation of the third scatter
sinogram in Step 150 is described below.
[0067] Returning to the method illustrated in FIG. 3, PET scanning
system 10 calculates scatter sinograms for scan data collected at
ends of bed 30. Such calculations can be necessary because bed
positions at ends of bed 30 are only adjacent to one other bed
position. In other words, scatter sinograms are calculated for
scatter data at an end bed position because there is not enough
overlapping data for the derivation of a scatter sinogram. For
example, in a PET scan acquisition with 8 bed positions and an
amount of overlap of 50%, sinograms at bed positions 1, 3, 5, 7 and
8 is calculated. For bed positions 2, 4 and 6, however, scatter
sinograms is derived from the scatter sinograms calculated at
neighboring bed positions. For example, the scatter sinogram for
the scatter data collected at bed position 2 is derived from the
scatter sinograms calculated for scatter data collected at bed
positions 1 and 3.
[0068] By deriving scatter sinograms for scan data at bed positions
that are overlapped by adjacent bed positions, i.e., skipping
calculations of scatter sinograms, computation time can be saved as
shown below:
Time saved = ( 1 - n N s + 1 + m n ) .times. 100 % ##EQU00001## N s
= { 1 x < 0.5 Round ( x 1 - x ) x .gtoreq. 0.5 m = { 1 , n % ( N
s + 1 ) .ltoreq. 1 2 , n % ( N s + 1 ) > 1 ##EQU00001.2##
[0069] where n is the total number of bed positions, x is the
overlap region in the axial direction, and Ns is the number of bed
positions to skip. For example, an amount of overlap of 50% can
yield a 1-bed position skip, an amount of overlap of 66.7% can
yield a 2-bed position skip, and an amount of overlap of 75% can
yield a 3-bed position skip.
[0070] Note that PET scanning system 10 rescales the scatter
sinograms from the previous bed position and the next bed position
when an overlap amount is greater than 50%. Further, PET scanning
system 10 can perform a tail-fitting process on derived scatter
sinograms to rescale them to the current bed position.
Additionally, PET scanning system 10 interpolates scatter data for
missing slices in scatter sinograms when an amount of overlap
between adjacent bed positions is less than 50%. Further detail of
these processes is provided below.
[0071] Alternatively, method 100 as illustrated in FIG. 3 is
performed by processing circuitry. In other words, method 100 is
performed by processing circuitry upon data stored in a memory.
Further, computer executable instructions for the steps described
in method 100 are stored in a memory and executed by the processing
circuitry. For example, a non-transitory computer readable medium
stores computer executable instructions that, when executed by the
processing circuitry, causes the processing circuitry to perform
the method as illustrated in FIG. 3.
[0072] FIG. 5 provides details of steps performed in Step 150 of
method 100.
[0073] In Step 151, PET scanning system 10 determines a first
portion of the first scatter sinogram. In the exemplary
implementation illustrated in FIG. 4, the first portion of the
first scatter sinogram is labeled portion A. PET scanning system 10
determines the first portion of the first scatter sinogram
according to (i) the amount of overlap of the bed positions, as
determined in Step 120, (ii) the portion of the first bed position
that overlaps with the third bed position, and (iii) an angle
.theta. of LORs within the first bed position. Angle .theta. of
LORs will be discussed below with reference to FIGS. 7(A)-7(C).
[0074] For example, when the amount of overlap is determined in
Step 120 to be 50%, and .theta.=0, then in Step 151, PET scanning
system 10 determines that the first portion is equal to the 50% of
the first scatter sinogram that overlaps with the third bed
position.
[0075] In Step 152, PET scanning system 10 determines a second
portion of the second scatter sinogram. In the exemplary
implementation illustrated in FIG. 4, the second portion of the
second scatter sinogram is labeled portion B. PET scanning system
10 determines the second portion of the second scatter sinogram
according to (i) the amount of overlap of the bed positions, as
determined in Step 120, (ii) the portion of the second bed position
that overlaps with the third bed position, and (iii) angle .theta.
of LORs within the second bed position.
[0076] For example, when the amount of overlap is determined in
Step 120 to be 40% and .theta.=0, then in Step 152, PET scanning
system 10 determines that the second portion is equal to the 40% of
the second scatter sinogram that overlaps with the third bed
position.
[0077] In Step 153, PET scanning system 10 copies the first portion
and second portion to create a third scatter sinogram. In
particular, PET scanning system 10 copies the first portion to the
third scatter sinogram in which the third position overlaps with
the first position, from which the first portion was created. PET
scanning system 10 copies the second portion to the third scatter
sinogram in which the third position overlaps with the second
position, from which the second portion was created. In other
words, PET scanning system 10 creates the third scatter sinogram
consistent the overlapping adjacent bed positions.
[0078] In Step 154, PET scanning system 10 determines whether
interpolation is necessary to complete the third sinogram.
Interpolation may be necessary, for example, when
.theta..noteq.0.degree. or when an amount of overlap is less than
50%. For example, when an amount of overlap of bed positions is 50%
and .theta..noteq.0.degree., interpolation may be necessary to
compute a remaining portion of the third sinogram. If PET scanning
system 10 determines that interpolation is necessary, PET scanning
system 10 proceeds to Step 155. Should PET scanning system 10
determine that interpolation is not necessary, PET scanning system
10 proceeds to Step 158.
[0079] In Step 155, PET scanning system 10 determines a remaining
portion that is necessary to complete the third scatter sinogram.
For example, when the amount of overlap is less than 50%, each of
the first portion and the second portion copied from the first and
second scatter sinograms are less than 50%, thus leaving a third
sinogram less than 100% complete. Should .theta..noteq.0.degree.,
interpolation may also be necessary. Further discussion of .theta.
and interpolation will be provided below with reference to FIGS.
7(A)-7(C).
[0080] In other words, PET scanning system 10 determines a
remaining portion of the third scatter sinogram based on a
difference between 100% and a combination of the first portion and
the second portion.
[0081] Once the remaining portion is determined, PET scanning
system 10 proceeds to Step 156. In Step 156, PET scanning system 10
interpolates the remaining portion to create an interpolated
portion. The interpolation process performed by PET scanning system
10 is described in detail below.
[0082] In Step 157, PET scanning system 10 copies the interpolated
portion into the third scatter sinogram. In particular, PET
scanning system 10 copies the interpolated portion in between the
first portion and the second portion. In other words, the
interpolated portion represents a calculated transition between the
first portion and the second portion. PET scanning system 10 then
proceeds to Step 158.
[0083] After completion of Step 157, or when PET scanning system 10
determines that interpolation is not necessary in Step 154, PET
scanning system 10 proceeds to Step 158. In Step 158, PET scanning
system 10 determines whether data averaging is necessary. Data
averaging may be necessary, for example, when an amount of overlap
is greater than 50%, depending upon a value of .theta.. Should PET
scanning system 10 determine that data averaging is necessary, PET
scanning system 10 proceeds to Step 159. However, should PET
scanning system 10 determine that data averaging is not necessary,
method Step 150 is complete.
[0084] In Step 159, PET scanning system 10 determines an
overlapping portion of the first portion and the second portion. In
other words, because the amount of overlap is greater than 50%, the
overlapping portion is a segment in which the first portion and the
second portion overlap. In an exemplary implementation, the
overlapping portion includes a first part of the first portion that
overlaps with a second part of the second portion.
[0085] As a result, PET scanning system 10 determines an
overlapping portion of the third scatter sinogram based on a
difference between 100% and a combination of the amount of overlap
of the first scatter sinogram and the amount of overlap of the
second scatter sinogram. In an exemplary embodiment, should PET
scanning system 10 determine an interpolation portion in Step 157,
the overlapping portion determined by PET scanning system 10 in
Step 159 could coincide with at least a part of the interpolation
portion.
[0086] Once the overlapping portion is determined, PET scanning
system 10 proceeds to Step 160. In Step 160, PET scanning system 10
rescales the overlapping portion of the first portion and the
second portion by averaging or weighting the overlapping portion to
create an averaged portion. The resealing process performed by PET
scanning system 10 is described in detail below.
[0087] Finally, in Step 161, PET scanning system 10 copies the
averaged portion into the third scatter sinogram to replace the
overlapping portions of the first portion and the second portion.
In particular, PET scanning system 10 copies the averaged portion
between the remainder of the first portion and the second
portion.
[0088] The copying and/or interpolating portions of scatter
sinograms within method 100 is achieved by PET scanning system 10
by the copying and/or interpolating LORs of captured scan data of
the respective adjacent bed position. Further discussion of copying
and interpolating LORs is provided below.
[0089] Discussion will now transition to estimating scatter using
the MBSE method.
[0090] Statistically, LORs in different bed positions include a
same number of scatter events because they pass through a same
object at a same position. In other words, a same scatter event is
detected by detectors 22 in different PET ring 20. When detected by
detectors 22 in different PET ring 20, the scatter event is
detected in different LORs.
[0091] Duplicate computation for scatter estimation occurs in a bed
position that is covered by multiple bed position data acquisition.
In other words, duplicate computation for scatter estimation occurs
when bed positions overlap. As a result, scatter estimation for
overlapping bed positions are accurately estimated from adjacent
bed positions for LORs with small axial tilted angles.
[0092] The MBSE method can be performed in the raw sinogram domain
or in the interpolated sinogram domain. A raw sinogram includes all
possible LORs in a corresponding FOV. For example, there can be
numRadial.times.numPhi.times.numRing.times.numRing LORs in a
specific FOV raw sinogram. Those LORs can be used directly in the
final reconstruction.
[0093] However, it can be difficult to calculate such a huge number
of LORs using the MBSE method even with interpolation processes.
Thus, it is beneficial to estimate scatter sinograms in the
interpolated sinogram domain instead of in the raw sinogram domain.
This is because the interpolated sinogram domain has much fewer
LORs than the raw sinogram domain. For example, there can be only
N.sub.s.times.N.sub..phi..times.N.sub.A*N.sub..theta. LORs in a
specific FOV interpolated sinogram, where the total number of LORs
is much fewer than that in a raw sinogram (N.sub.s, N.sub..phi.,
N.sub.z and N.sub..theta. are the number of s, .phi., z and
.theta.). After calculating scatter interpolated sinograms, a
back-interpolation process is needed to transfer scatter events
back to the raw sinograms for the final reconstruction.
[0094] FIGS. 6(A) and 6(B) illustrate an interpolated sinogram and
its coordinate system in transaxial and axial views.
[0095] In an interpolated sinogram domain, any LOR is represented
by a set of parameters (s, .phi., z, .theta.) and a typical
interpolated sinogram is arranged with s varying fastest and
.theta. varying slowest, as illustrated in FIG. 6(A). In
particular, z and .theta. are calculated as follows:
z = ( z a + z b ) / 2 ##EQU00002## .theta. = tan - 1 z b - z a ( x
b - x a ) 2 + ( y b - y a ) 2 ##EQU00002.2##
[0096] In an exemplary implementation, the last three parameters
are fixed. However, s can vary with each FOV.
[0097] FIG. 6(B) illustrates the coordinate system of an
interpolated sinogram. In particular, FIG. 6(B) illustrates a set
of LORs with fixed (z, .theta.), but different (s, .phi.) in an
axial view. As illustrated, each LOR is superimposed together at a
same line, but with different ending points. Such a same line
includes two ending points that are symmetric about a middle point
on the Z-axis. As a result, in each of the following figures, any
line in an axial view represents a set of LORs which actually form
one slice with fixed (z, .theta.) in an interpolated sinogram.
[0098] FIGS. 7(A)-7(C) illustrate LORs for different .theta.'s when
an amount of overlap of adjacent bed positions is 50%.
[0099] PET scanning system 10 determines whether a LOR can be
directly copied from an adjacent bed position or whether a LOR
requires interpolation. PET scanning system 10 makes such a
determination by analyzing tilted angle .theta. and the z
coordinate. In particular, PET scanning system 10 utilizes the
following equation that relates a LOR in a current bed position
(third position) to LORs in previous bed position (second position)
and/or next bed position (first position):
LOR.sup.c(s,.phi.,z,.theta.)=LOR.sup.p(s,.phi.,z-z.sub.shift,.theta.)=LO-
R.sup.n(s,.phi.,z+z.sub.shift,.theta.)
Where p represents a previous bed position, c represents a current
bed position and n represents a next bed position. Additionally,
z.sub.shift=overlap %*z.sub.max, and
0.ltoreq.z.+-.z.sub.shift.ltoreq.z.sub.max, where z.sub.max is the
PET scanner' axial field-of-view. The indexing of +z.sub.shift or
-z.sub.shift in previous and next bed depends on the increment
direction of the z-axis.
[0100] In other words, PET scanning system 10 can directly copy a
LOR from an adjacent bed position when LOR.sup.p and/or LOR.sup.n
are available for creation of a scatter sinogram. However, when
both LOR.sup.p and LOR.sup.n are not available, PET scanning system
10 must perform an interpolation process with LOR.sup.p and/or
LOR.sup.n to calculate LOR.sup.c for the creation of the scatter
sinogram.
[0101] In an exemplary implementation with an interpolated sinogram
of 700-mm-FOV and an amount of overlap of 50%, the scatter
interpolated sinogram have 7 .theta.'s ranging from -15.3 to 15.3
degrees and 95 z's covering the whole axial FOV, e.g., 48 rings.
FIG. 7(A) illustrates each .theta., which includes a set of LORs
with different z's are grouped together. FIG. 7(B) illustrates 4
different .theta.'s with an amount of overlap of 50%. In FIG. 7(B),
solid lines stand for a set of LORs which can be directly copied
from either previous or next bed while dashed lines stand for a set
of LORs which can be partially copied or partially interpolated
from either previous or next bed position. Other .theta.'s with
positive values are symmetric with ones with negative values and
therefore are not illustrated in FIGS. 7(A)-7(C). Boundary LORs are
indicated by {circumflex over (1)} and {circumflex over (2)} in
FIGS. 7(B) and 7(C), and any LORs between them can require
interpolations.
[0102] As illustrated in FIG. 7(C), for any LORs between LOR
{circumflex over (1)} and {circumflex over (2)}:
[0103] Crystal a and b can be one of three cases: [0104] (1) a and
b all locate in the previous bed, such as a triangle; [0105] (2) a
and b all locate in the next bed, such as an `x`; and [0106] (3) a
and b locate in the previous and next bed, such as a square.
[0107] Only LORs of case 3 are interpolated from their existing
neighboring LORs.
[0108] FIG. 8 illustrates a calculation percentage of copied LORs
in a dashed line.
[0109] In an exemplary embodiment, PET scanning system 10 uses
known properties of interpolated sinograms to calculate the
percentage of copied LORs: [0110] (1) Two crystals of a LOR have
opposite t coordinates, i.e., t.sub.a+t.sub.b=0 (its two ending
points are symmetric about its middle point); and [0111] (2)
Crystals of a set of LORs uniformly distributed along the line in
an axial view since s and .phi. are uniformly sampled.
[0112] Therefore, the percentage of copied LORs in a dashed line is
calculated by the line segment (2*L) divided by the length of whole
dashed line, as shown in FIG. 8.
[0113] Calculations of copy % are strongly related to the scanning
geometry of PET scanning system 10. However, such calculations of
copy % may be approximated as in the following discussion.
[0114] In an exemplary embodiment, then, PET scanning system 10
derives the formula to calculate this percentage by:
copy dashed - line ( z , .theta. ) % = copied LORs in a dashed line
total LORs in a dashed line .times. 100 % = 2 .times. L length of a
dashed line .times. 100 % = 2 .times. z 0 - z sin .theta. transFOV
cos .theta. .times. 100 % = 2 .times. z 0 - z transFOV tan .theta.
.times. 100 % ##EQU00003##
[0115] In an interpolated sinogram, both transaxial FOV and axial
FOV can be considered. Any LORs out of axial FOV, i.e., z.sub.a
and/or z.sub.b<0 or z.sub.a and/or z.sub.b>z.sub.max, are not
included in the sinogram. For example, any LORs outside of the
dashed boxes in FIG. 7(A) are partially or totally discarded
depending on locations of their ending points. To simplify the
calculation, PET scanning system 10 first calculates copy(.theta.)%
without considering restriction of axial FOV, then introduces an
additional parameter eff(.theta.) to take care of axial FOV.
[0116] Firstly, without considering axial FOV, PET scanning system
10 calculates copy(.theta.)% for a given tilted angle .theta.
as:
copy ( .theta. ) % = copied LORs in .theta. total LORs in .theta.
.times. 100 % = copied LORs from solid lines + copied LORs from
dashed lines total LORs in .theta. .times. 100 % = N s .times. N
.PHI. .times. ( z max - ( z end - z start ) ) + N s .times. N .PHI.
.times. z = z start z = z end copy dashed - line ( z , .theta. ) %
N s .times. N .PHI. .times. N z .times. 100 % = ( z max - ( z end -
z start ) ) + z = z start z = z end copy dashed - line ( z ,
.theta. ) % N z .times. 100 % ##EQU00004## Where z end ( z start =
- z end ) is given by : ##EQU00004.2## z end = transFOV tan .theta.
/ 2 ##EQU00004.3##
[0117] And N.sub.s, N.sub..phi., and N.sub.z are the number of s,
.phi. and z.
[0118] For example, copy(.theta.)% is about 51% when .theta.=10.2
and transFOV=700 mm.
[0119] Then the total copy % is given by:
total copy % = 1 N .theta. .theta. = .theta. min .theta. = .theta.
max copy ( .theta. ) % ##EQU00005##
[0120] In an exemplary implementation, PET scanning system 10
copies about 44% of LORs from the previous or next bed position
with an amount of overlap of 50%.
[0121] After discarding any LORs out of axial FOV, copy % becomes
larger since a lot of LORs are not required in scatter sinograms
especially for large tilted angles. Roughly, PET scanning system 10
calculates copy % without out-of-FOV LORs by:
total copy % = 1 .theta. = .theta. min .theta. = .theta. max eff (
.theta. ) .theta. = .theta. min .theta. = .theta. max eff ( .theta.
) * copy ( .theta. ) % ##EQU00006##
[0122] Where eff(.theta.) is the effective LORs rate (=LORs in
axial FOV/all LORs).
[0123] For a 700-mm-FOV, eff(.theta.) is about 1.4%, 25%, 63%, and
100% for .theta.=.+-.15.3, .+-.10.2, .+-.5.1, and 0 degree. As a
result, PET scanning system 10 achieves a final copy % of
approximately 80% for an amount of overlap of 50%.
[0124] FIG. 9(A) illustrates adjacent bed positions when an amount
of overlap is greater than 50% (left) and when an amount of overlap
is less than 50% (right).
[0125] When an amount of overlap of adjacent bed positions is
greater than 50%, as illustrated on the left side of FIG. 9(A), the
copy % becomes larger and the error of the method performed by PET
scanning system 10 further decreases. On the other hand, when an
amount of overlap of adjacent bed positions is less than 50%, as
illustrated on the right side of FIG. 9(A), the copy % becomes
smaller and the error of this method increases.
[0126] When an amount of overlap is greater than or equal 50%, as
illustrated on the left side of FIG. 9(A), LORs in the current bed
position can be directly copied from the previous or next bed
position should they be available. When an amount overlap is
greater than 50%, there can be more than one available LOR from the
previous or next bed positions for a target LOR in the current bed
position. That is, when the amount of overlap is greater than 50%,
PET scanning system 10 can geometrically average the two or more
LORs to get a target LOR. Alternatively, PET scanning system 10 can
use a weight function w(z) that is proportional to the axial
sensitivity of PET scanning system 10 to calculate the target
LOR:
Target LOR = i = 1 n LOR i from neighboring beds * w ( z i ) i = 1
n w ( z i ) ##EQU00007##
[0127] FIG. 9(B) illustrates a weight function w(z) for weighting
same LORs from different bed positions.
[0128] Alternatively, when an amount of overlap is less than 50%,
as illustrated on the right side of FIG. 9(A), LORs in the current
bed position can be interpolated from scan data in overlapping
adjacent bed positions. When corresponding LOR.sup.p and LOR.sup.n
are not available, then LOR.sup.c can be interpolated from
neighbors of LOR.sup.p and/or LOR.sup.n. PET scanning system 10 can
use any kind of interpolation methods such as nearest neighbors,
linear with different orders in each dimension, cubic and splines
to acquire the desired LOR.sup.c based on the closest neighbors of
LOR.sup.p and/or LOR.sup.n. In particular,
LOR.sup.c(s,.phi.,z,.theta.)=f(LOR.sup.p(s+.DELTA.s,.phi.+.DELTA..phi.,z-
-z.sub.shift+.DELTA.z,.theta.+.DELTA..theta.),LOR.sup.n(s+.DELTA.s,.phi.+.-
DELTA..phi.,z+z.sub.shift+.DELTA.z,.theta.+.DELTA..theta.))
[0129] Where .DELTA.s, .DELTA..phi., .DELTA.z and .DELTA..theta.
stand for changes as small as possible (therefore, closest
neighbors). Depending on the accuracy and computational
requirements, the method of interpolation and the number of
neighbors can be varied.
[0130] The method for estimating scatter performed by PET scanning
system 10 can also be performed using data in the raw sinogram
domain. In such calculations, the parameter set (s, .phi., z,
.theta.) can be replaced with (rad, phi, ringSum, ringDiff).
Corresponding parameters can have similar meanings and intrinsic
correlations. To understand the raw sinograms, PET scanning system
10 can simply use (rad, phi, ringSum, ringDiff) to replace (s,
.phi., z, .theta.) in above discussion. For example:
LOR c ( rad , phi , ringSum , ringDiff ) = ( rad , phi , ringSum -
ringSum shift , ringDiff ) = LOR n ( rad , phi , ringSum + ringSum
shift , ringDiff ) with ringSum shift = overlap % * # rings and 0
.ltoreq. ringSum .+-. ringSum shift .ltoreq. # rings .
##EQU00008##
[0131] FIG. 10 illustrates an exemplary component configuration of
a PET scanning system. In particular, FIG. 10, PET scanning system
300 includes CPU 500, which performs the processes shown in the
flowcharts and described above. Process data and executable
instructions for a method for estimating scatter, as described
relating to FIG. 3, is stored in memory 502. These processes and
instructions can also be stored on a storage medium disk 504 such
as a hard drive (HDD) or portable storage medium or can be stored
remotely. Further, the claimed advancements are not limited by the
form of the computer-readable media on which the instructions of
the inventive process are stored. For example, the instructions can
be stored on CDs, DVDs, in FLASH memory, RAM, ROM, PROM, EPROM,
EEPROM, hard disk or any other information processing device with
which the PET scanning system 10 communicates, such as a server or
computer.
[0132] Further, the claimed advancements can be provided as a
utility application, background daemon, or component of an
operating system, or combination thereof, executing in conjunction
with CPU 500 and an operating system such as, for example,
Microsoft Windows 7, UNIX, Solaris, LINUX, Apple MAC-OS and other
systems known to those skilled in the art.
[0133] CPU 500 can be implemented using discrete logic circuits.
Further, CPU 500 can also be implemented as multiple processors
cooperatively working in parallel to perform the instructions of
the inventive processes described above. In other embodiments, PET
scanning device may include a CPU, a GPU, or both.
[0134] PET scanning device 300 as illustrated in FIG. 10 can also
include a network controller 506 for interfacing with network 400.
Network 400 can be a public network, such as the Internet, or a
private network such as an LAN or WAN network, or any combination
thereof and can also include PSTN or ISDN sub-networks. Network 400
can be wired, such as an Ethernet network, or can be wireless such
as a cellular network including EDGE, 3G and 4G wireless cellular
systems. The wireless network can also be WiFi, Bluetooth, or any
other wireless form of communication that is known.
[0135] In an exemplary implementation, the calculations performed
in a method for estimating scatter in a PET scan at multiple bed
positions can be performed entirely by PET scanning system 300.
Alternatively, the calculations performed in the method can be
subdivided and performed in series or in parallel by devices over
network 400. For example, calculations can be performed by multiple
devices in communication over network 400.
[0136] PET scanning system 300 can further include a general
purpose I/O interface 512 that interfaces with a keyboard and/or
mouse 510 as well a display 508. I/O interface 512 can also
connects to a variety of peripherals 514 such as printers and
scanners.
[0137] The general purpose storage controller 516 connects the
storage medium disk 504 with communication bus 518 for
interconnecting all of the components of the PET scanning system
300.
[0138] FIGS. 11 and 12 illustrate reconstructed images of scanned
data after scatter estimation. In particular, FIG. 11(A)
illustrates a traversal view of a reconstructed image with scatter
estimation by MBSE and FIG. 12(A) illustrates a coronal view of a
reconstructed image with scatter estimation by MBSE. By contrast,
FIG. 11(B) illustrates a traversal view of a reconstructed image
with scatter estimation by MBSE utilizing according to the proposed
method of FIG. 3 and FIG. 12(B) illustrates a coronal view of a
reconstructed image with scatter estimation by MBSE utilizing
according to the proposed method of FIG. 3.
[0139] While certain embodiments have been described, these
embodiments have been presented by way of example only, and are not
intended to limit the scope of the inventions. Indeed the novel
methods and systems described herein may be embodied in a variety
of other forms; furthermore, various omissions, substitutions, and
changes in the form of the methods and systems described herein may
be made without departing from the spirit of the inventions. The
accompanying claims and their equivalents are intended to cover
such forms or modifications as would fall within the scope and
spirit of the inventions.
* * * * *