U.S. patent application number 10/309301 was filed with the patent office on 2004-06-03 for method and system for tomosynthesis image enhancement using transverse filtering.
Invention is credited to Claus, Bernhard Erich Hermann, Eberhard, Jeffrey Wayne.
Application Number | 20040105528 10/309301 |
Document ID | / |
Family ID | 32312237 |
Filed Date | 2004-06-03 |
United States Patent
Application |
20040105528 |
Kind Code |
A1 |
Claus, Bernhard Erich Hermann ;
et al. |
June 3, 2004 |
METHOD AND SYSTEM FOR TOMOSYNTHESIS IMAGE ENHANCEMENT USING
TRANSVERSE FILTERING
Abstract
A technique is provided for reducing contrast variability in
images produced by tomosynthesis. In particular, contrast
variability in images produced by a radiation source which has a
linear or elongated scan path is reduced. The technique filters the
radiographic projections or the reconstructed image slices in a
direction transverse to the scan-direction of the radiation source
to reduce contrast variations related to the orientation of the
structure and the scan geometry.
Inventors: |
Claus, Bernhard Erich Hermann;
(Niskayuna, NY) ; Eberhard, Jeffrey Wayne;
(Albany, NY) |
Correspondence
Address: |
Patrick S. Yoder
Fletcher, Yoder & Van Someren
P.O. Box 692289
Houston
TX
77269-2289
US
|
Family ID: |
32312237 |
Appl. No.: |
10/309301 |
Filed: |
December 3, 2002 |
Current U.S.
Class: |
378/210 |
Current CPC
Class: |
G06T 2211/436 20130101;
Y10S 378/901 20130101; G06T 11/005 20130101 |
Class at
Publication: |
378/210 |
International
Class: |
H05G 001/00 |
Claims
What is claimed is:
1. A method for processing radiographic image data, comprising:
moving a radiation source relative to an imaged object to obtain
two or more radiographic projections of the imaged object at two or
more corresponding view angles; filtering the radiographic
projections in a direction generally transverse to a scan path of
the radiation source; and applying a reconstruction algorithm to
the filtered projections to reconstruct one or more reconstructed
slices.
2. The method as recited in claim 1, wherein the reconstruction
algorithm utilizes a re-projection consistency constraint.
3. The method as recited in claim 1, wherein filtering the
radiographic projections comprises applying a complementary filter
to the radiographic projections.
4. The method as recited in claim 3, wherein the complementary
filter comprises at least one of a linear filter, a non-linear
filter, and a multi-scale method based filter.
5. The method as recited in claim 3, wherein the complementary
filter comprises a high-pass filter.
6. The method as recited in claim 1, wherein filtering the two or
more radiographic projections comprises selecting a filter having a
desired characteristic adapted to an apparent filtering effect
produced by the reconstruction algorithm.
7. The method as recited in claim 1, wherein filtering the two or
more radiographic projections comprises selecting a filter which,
in conjunction with an apparent filtering effect produced by the
reconstruction algorithm, reduces an orientation dependent contrast
variation associated with a reconstructed structure in the one or
more reconstructed slices.
8. The method as recited in claim 1, wherein filtering the two or
more radiographic projections comprises selecting a filter which,
in conjunction with an apparent filtering effect produced by the
reconstruction algorithm, produces a rotationally symmetric
filtering effect in the one or more reconstructed slices.
9. The method as recited in claim 1, further comprising filtering
the one or more reconstructed slices in a direction transverse to
the scan path of the radiation source.
10. The method as recited in claim 9, wherein filtering the one or
more reconstructed slices comprises applying a complementary filter
to the reconstructed slices.
11. A method for processing radiographic image data, comprising:
moving a radiation source relative to an imaged object to obtain
two or more radiographic projections of the imaged object at two or
more corresponding view angles; applying a reconstruction algorithm
to the radiographic projections to reconstruct one or more image
slices; and filtering the image slices in a direction transverse to
a scan path of the radiation source.
12. The method as recited in claim 11, wherein the reconstruction
algorithm utilizes a re-projection consistency constraint.
13. The method as recited in claim 11, wherein filtering the image
slices comprises applying a complementary filter to the image
slices.
14. The method as recited in claim 13 wherein applying the
complementary filter comprises at least one of a linear filter, a
non-linear filter, and a multi-scale method based filter.
15. The method as recited in claim 13, wherein the complementary
filter comprises a high-pass filter.
16. The method as recited in claim 11, wherein filtering the image
slices comprises selecting a filter having a desired characteristic
adapted to an apparent filtering effect produced by the
reconstruction algorithm.
17. The method as recited in claim 11, wherein filtering the image
slices comprises selecting a filter which, in conjunction with an
apparent filtering effect produced by the reconstruction algorithm,
reduces an orientation dependent contrast variation associated with
a reconstructed structure in the filtered image slices.
18. The method as recited in claim 11, wherein filtering the image
slices comprises selecting a filter which, in conjunction with an
apparent filtering effect produced by the reconstruction algorithm,
produces a rotationally symmetric filtering effect in the filtered
image slices.
19. The method as recited in claim 11, further comprising filtering
the two or more radiographic projections in a direction transverse
to the scan path of the radiation source.
20. The method as recited in claim 19, wherein filtering the two or
more radiographic projections comprises applying a complementary
filter to the two or more radiographic projections.
21. A system for processing radiographic image data, comprising: a
radiation source capable of producing a stream of radiation; a
control circuit operably coupled to the radiation source; a
detector capable of detecting the stream of radiation and
generating two or more radiographic projections; a motor controller
configured to displace at least one of the radiation source, the
detector, and a patient platform; a processing circuit operably
coupled to the detector configured to receive the two or more
radiographic projections and to reconstruct the two or more
radiographic projections to form one or more reconstructed slices
representative of one or more structures within an imaged volume,
wherein the processing circuit applies at least one complementary
filter to at least one of the two or more radiographic projections
and the one or more reconstructed slices in a direction generally
transverse to a scan path of the radiation source; and an operator
workstation operably couple to the processing circuit configured to
display the one or more reconstructed slices.
22. The system as recited in claim 21, wherein the processing
circuit applies a reconstruction algorithm utilizing a
re-projection consistency constraint to reconstruct the two or more
radiographic projections.
23. The system as recited in claim 21, wherein the filter comprises
at least one of a linear filter, a non-linear filter, and a
multi-scale method based filter.
24. The system as recited in claim 21, wherein the filter comprises
a high-pass filter.
25. The system as recited in claim 21, wherein the filter has a
characteristic adapted to an apparent filtering effect produced by
a reconstruction algorithm used to reconstruct the two or more
radiographic projections.
26. The system as recited in claim 21, wherein the filter produces
a rotationally symmetric filtering effect in conjunction with an
apparent filtering effect produced by reconstructing the two or
more radiographic projections.
27. The system as recited in claim 21, wherein a contrast variation
of the one or more reconstructed slices attributable to the
orientation of the imaged structures is reduced.
28. A system for processing radiographic image data, comprising: a
radiation source capable of producing a stream of radiation; a
control circuit operably coupled to the radiation source; a
detector capable of detecting the stream of radiation and
generating two or more radiographic projections; a motor controller
configured to displace at least one of the radiation source, the
detector, and a patient platform; a processing circuit operably
coupled to the detector configured to receive the two or more
radiographic projections and to reconstruct the two or more
radiographic projections to form one or more reconstructed slices
representative of one or more structures within an imaged volume,
wherein the processing circuit comprises a means for reducing
contrast asymmetry in the one or more reconstructed slices; and an
operator workstation operably couple to the processing circuit
configured to display the one or more reconstructed slices.
29. A tangible medium for processing radiographic image data,
comprising: a routine for filtering two or more radiographic
projections in a direction generally transverse to a scan path of a
radiation source used to generate the projections; and a routine
for applying a reconstruction algorithm to the filtered projections
to generate one or more reconstructed slices.
30. The tangible medium as recited in claim 29, wherein the
reconstruction algorithm utilizes a re-projection consistency
constraint.
31. The tangible medium as recited in claim 29, wherein the routine
for filtering the two or more radiographic projections applies a
complementary filter to the radiographic projections.
32. The tangible medium as recited in claim 31, wherein the
complementary filter comprises at least one of a linear filter, a
non-linear filter and a multi-scale method based filter.
33. The tangible medium as recited in claim 31, wherein the
complementary filter comprises a high-pass filter.
34. The tangible medium as recited in claim 29, wherein the routine
for filtering the two or more radiographic projections applies a
filter having a desired characteristic adapted to an apparent
filtering effect produced by the reconstruction algorithm.
35. The tangible medium as recited in claim 29, wherein the routine
for filtering the two or more radiographic projections applies a
filter which, in conjunction with an apparent filtering effect
produced by the reconstruction algorithm, reduces orientation
dependent contrast variations associated with one or more
reconstructed structures in the one or more reconstructed
slices.
36. The tangible medium as recited in claim 29, wherein the routine
for filtering the two or more radiographic projections applies a
filter which, in conjunction with an apparent filtering effect
produced by the reconstruction algorithm, produces a rotationally
symmetric filtering effect in the one or more reconstructed
slices.
37. The tangible medium as recited in claim 29, further comprising
a routine for filtering the one or more reconstructed slices in a
direction transverse to the scan path of the radiation source.
38. The tangible medium as recited in claim 37, wherein filtering
the one or more reconstructed slices comprises applying a
complementary filter to the one or more reconstructed slices.
39. A tangible medium for processing radiographic image data,
comprising: a routine for applying a reconstruction algorithm to
two or more radiographic projections to generate one or more
reconstructed slices; and a routine for filtering the one or more
reconstructed slices in a direction generally transverse to a scan
path of the radiation source used to generate the projections.
40. The tangible medium as recited in claim 39, wherein the routine
for applying a reconstruction algorithm utilizes a re-projection
consistency constraint.
41. The tangible medium as recited in claim 39, wherein the routine
for filtering the one or more reconstructed slices applies a
complementary filter to the reconstructed slices.
42. The tangible medium as recited in claim 41, wherein the
complementary filter comprises at least one of a linear filter, a
non-linear filter, and a multi-scale method based filter.
43. The tangible medium as recited in claim 41, wherein the
complementary filter comprises a high-pass filter.
44. The tangible medium as recited in claim 39, wherein the routine
for filtering the image slices applies a filter having a desired
characteristic adapted to an apparent filtering effect produced by
the reconstruction algorithm.
45. The tangible medium as recited in claim 39, wherein the routine
for filtering the reconstructed slices applies a filter which, in
conjunction with an apparent filtering effect produced by the
reconstruction algorithm, reduces orientation dependent contrast
variations associated with one or more reconstructed structures in
the filtered reconstructed slices.
46. The tangible medium as recited in claim 39, wherein the routine
for filtering the reconstructed slices applies a filter which, in
conjunction with an apparent filtering effect produced by the
reconstruction algorithm, produces a rotationally symmetric
filtering effect in the filtered reconstructed slices.
47. The tangible medium as recited in claim 39, further comprising
a routine for filtering the two or more radiographic projections in
a direction transverse to the scan path of the radiation
source.
48. The tangible medium as recited in claim 47, wherein filtering
the two or more radiographic projections comprises applying a
complementary filter to the two or more radiographic
projections.
49. A method for processing radiographic image data, comprising:
moving a radiation source relative to an imaged object to obtain
two or more radiographic projections of the imaged object at two or
more corresponding view angles; applying a reconstruction algorithm
to the radiographic projections to reconstruct one or more
reconstructed slices, wherein applying the reconstruction algorithm
comprises applying a complementary filter to the radiographic
projections in a direction generally transverse to a scan path of
the radiation source.
50. The method as recited in claim 49, wherein the complemenary
filter comprises a high-pass filter.
51. The method as recited in claim 49, wherein applying the
complementary filter to the radiographic projections comprises
selecting the complementary filter which, in conjunction with an
apparent filtering effect produced by the reconstruction algorithm,
reduces an orientation dependent contrast variation associated with
a reconstructed structure in the one or more reconstructed
slices.
52. The method as recited in claim 49, wherein applying the
complementary filter to the radiographic projections comprises
selecting the complementary filter which, in conjunction with an
apparent filtering effect produced by the reconstruction algorithm,
produces a rotationally symmetric filtering effect in the one or
more reconstructed slices.
53. A system for processing radiographic image data, comprising: a
radiation source capable of producing a stream of radiation; a
control circuit operably coupled to the radiation source; a
detector capable of detecting the stream of radiation and
generating two or more radiographic projections; a motor controller
configured to displace at least one of the radiation source, the
detector, and a patient platform; a processing circuit operably
coupled to the detector configured to apply a reconstruction
algorithm to the two or more radiographic projections to
reconstruct one or more reconstructed slices representative of one
or more structures within an imaged volume, wherein the
reconstruction algorithm comprises a complementary filter which
filters the radiographic projections in a direction generally
transverse to a scan path of the radiation source; and an operator
workstation operably couple to the processing circuit configured to
display the one or more reconstructed slices.
54. The system as recited in claim 53, wherein the complementary
filter comprises a high-pass filter.
55. The system as recited in claim 53, wherein the filter produces
a rotationally symmetric filtering effect in conjunction with an
apparent filtering effect produced by reconstructing the two or
more radiographic projections.
56. The system as recited in claim 53, wherein a contrast variation
of the one or more reconstructed slices attributable to the
orientation of the imaged structures is reduced.
57. A tangible medium for processing radiographic image data,
comprising: a routine for applying a reconstruction algorithm to
two or more radiographic projections to reconstruct one or more
reconstructed slices, wherein the two or more radiographic
projections are obtained at different view angles relative to an
imaged object and wherein applying the reconstruction algorithm
comprises applying a complementary filter to the radiographic
projections in a direction generally transverse to a scan path of
the radiation source.
58. The tangible medium as recited in claim 57, wherein the
complementary filter comprises a high-pass filter.
59. The tangible medium as recited in claim 57, wherein the routine
for applying a reconstruction algorithm to two or more radiographic
projections applies a complementary filter which, in conjunction
with an apparent filtering effect produced by the reconstruction
algorithm, reduces orientation dependent contrast variations
associated with one or more reconstructed structures in the one or
more reconstructed slices.
60. The tangible medium as recited in claim 57, wherein the routine
for applying a reconstruction algorithm to two or more radiographic
projections applies a complementary filter which, in conjunction
with an apparent filtering effect produced by the reconstruction
algorithm, produces a rotationally symmetric filtering effect in
the one or more reconstructed slices.
Description
BACKGROUND OF THE INVENTION
[0001] The present invention relates generally to the field of
medical imaging, and more specifically to the field of
tomosynthesis. In particular, the present invention relates to
minimizing contrast variations in tomosynthetic
reconstructions.
[0002] Tomographic imaging technologies are of increasing
importance in medical diagnosis, allowing physicians and
radiologists non-invasively to obtain three-dimensional
representations of selected organs or tissues of a patient.
Tomosynthesis is a variation of conventional planar tomography in
which a limited number of radiographic projections are digitally
acquired at different angles relative to the patient. In
tomosynthesis, an X-ray source produces a fan or cone-shaped X-ray
beam that is collimated and passes through the patient to then be
detected by a set of detector elements. The detector elements
produce a signal based on the attenuation of the X-ray beams. The
signals may be processed to produce a radiographic projection,
comprising generally the line integrals of the attenuation
coefficients of the object along the ray path. The source, the
patient, or the detector are then moved relative to one another for
the next exposure, typically by moving the X-ray source, so that
each projection is acquired at a different angle.
[0003] By using reconstruction techniques, such as filtered
backprojection, the set of acquired projections may then be
reconstructed to produce diagnostically useful three-dimensional
images. Because the three-dimensional information is obtained
digitally during tomosynthesis, the image can be reconstructed in
whatever viewing plane the operator selects. Typically, a set of
slices representative of some volume of interest of the imaged
object is reconstructed, where each slice is a reconstructed image
representative of structures in a plane that is parallel to the
detector plane, and each slice corresponds to a different distance
of the plane from the detector plane.
[0004] In addition, because tomosynthesis reconstructs
three-dimensional data from projections, it provides a fast and
cost-effective technique for removing superimposed anatomic
structures and for enhancing contrast in in-focus planes as
compared to the use of a single X-ray radiograph. Further, because
the tomosynthesis data consists of relatively few projection
radiographs that are acquired quickly, often in a single sweep of
the X-ray source over the patient, the total X-ray dose received by
the patient is comparable to the dose of a single conventional
X-ray exposure and is typically less than the dose received from a
computed tomography (CT) examination. In addition, the resolution
of the detector employed in tomosynthesis is typically greater than
the resolution of detectors used in CT examinations. These
qualities make tomosynthesis useful for such radiological tasks as
detecting pulmonary nodules or other difficult to image
pathologies.
[0005] Though tomosynthesis provides these considerable benefits,
the techniques associated with tomosynthesis also have
disadvantages. In particular, the reconstruction problem is
difficult to solve because only incomplete information is available
due to the nature of the technique. That is, the radiographic
projections may be acquired from only a few angles within a
relatively narrow angular range and are not densely spaced over the
full angular range, limiting the amount of information acquired.
Advanced reconstruction algorithms are employed to solve these
reconstruction problems. A good reconstruction algorithm provides
efficient separation of overlying tissue, minimizes artifacts, and
enhances contrast, particularly of small structures.
[0006] Reconstructed data sets in tomosynthesis often exhibit a
blurring of structures in the direction of the projections that
were used to acquire the tomosynthesis data. These artifacts
associated with an imaged structure will vary depending on the
orientation of the structure with respect to the acquisition
geometry. Therefore, the blurring of structures may create
undesirable image artifacts and inhibit the separation of
structures located at different heights in the reconstruction of
the imaged volume.
[0007] Systems employing advanced reconstruction algorithms
utilizing a re-projection consistency constraint, either directly
or indirectly, to obtain high-quality reconstructions attempt to
recover the contrast of the imaged structures and to minimize the
aforementioned blurring of structures. In algorithms incorporating
a re-projection consistency constraint, the degree of contrast
which can be recovered and the degree of blurring will vary
depending on the algorithm used, the acquisition geometry of the
imaging system, and the geometry, position, and orientation of the
imaged object or structure. Examples of algorithms incorporating
the re-projection consistency constraints include linear/additive
ART, matrix inversion tomosynthesis (MITS), volumetric non-linear
reconstruction, and generalized filtered backprojection. As can be
observed in these algorithms, the contrast recovered and the
remaining blur are interdependent. In particular, the re-projection
consistency constraint has the effect of keeping constant the total
amount of the contrast of the reconstructed structure and the
contrast of the blurring artifacts associated with that structure.
Hence, the better the reconstruction algorithm is at suppressing
blurring, the better the contrast of the reconstructed
structure.
[0008] However, the shape and extent of the remaining blur is
strongly dependent on the shape and orientation of the structure in
relation to the specific system geometry used for image
acquisition, as discussed above. In particular, if the X-ray source
travels along a generally linear trajectory during the imaging
process, a structure that is "long" in a direction generally
parallel to the linear path of the source will produce a widespread
blur, while a structure that is "short" in a direction generally
parallel to the linear path of the source will produce only a
localized blur. For example, an elongated structure will produce a
widespread blur if it is oriented generally parallel to the linear
path of the source, and only a localized blur if it is oriented
generally perpendicular to the source trajectory. Due to the
aforementioned interdependence between contrast of the
reconstructed structure and the remaining blur due to that
structure, the contrast of the reconstruction of that same
elongated structure is higher if it is oriented generally
perpendicular to the source trajectory, and lower if it is oriented
generally parallel to the source trajectory.
[0009] One manner in which this problem has been indirectly
addressed has been to utilize symmetric system geometries, such as
in circular tomosythesis, which acquire projections at a number of
different orientations relative to each structure orientation. For
example, in circular tomosynthesis, the X-ray source is not moved
in a linear trajectory, but is instead moved in a circular
trajectory in a plane substantially parallel to the plane of the
detector. However, in many instances, a less symmetric acquisition
geometry may be preferred, for example for reasons of system
complexity or scanning speed. For instance, in pulmonary
tomosynthesis, a generally linear or an elongated two-dimensional
geometry (such as elliptical) may be preferred. An effective method
of minimizing contrast variability in reconstructed images while
allowing the use of non-symmetric source trajectories, for example
linear or elongated acquisition geometries, is therefore
needed.
BRIEF DESCRIPTION OF THE INVENTION
[0010] The present technique provides a novel approach to
correcting contrast asymmetry in three-dimensional images derived
from radiographic projections. Particularly, the technique applies
a filter which provides contrast symmetry in the reconstructed
image. The technique thereby compensates for contrast variations
attributable to acquisition system geometry and subject
orientation.
[0011] In accordance with one aspect of the technique, a method is
provided for processing radiographic image data. Two or more
radiographic projections of an imaged object are obtained at two or
more corresponding view angles by moving a radiation source
relative to an imaged object. The radiographic projections are
filtered in a direction that is generally transverse to a scan path
of the radiation source. A reconstruction algorithm is applied to
the filtered projections to reconstruct one or more reconstructed
slices.
[0012] In accordance with another aspect of the technique, a method
is provided for processing radiographic image data. Two or more
radiographic projections of an imaged object are obtained at two or
more corresponding view angles by moving a radiation source
relative to an imaged object. A reconstruction algorithm is applied
to the radiographic projections to reconstruct one or more image
slices. The image slices are filtered in a direction transverse to
a scan path of the radiation source.
[0013] In accordance with a further aspect of the technique, a
system is provided for processing radiographic image data. The
system includes a radiation source capable of producing a stream of
radiation and a control circuit operably coupled to the radiation
source. In addition, the system includes a detector capable of
detecting the stream of radiation and generating two or more
radiographic projections and a motor controller configured to
displace at least one of the radiation source, the detector, and a
patient platform. A processing circuit operably coupled to the
detector and configured to receive the two or more radiographic
projections is also included. The processing circuit is further
configured to reconstruct the two or more radiographic projections
to form one or more reconstructed slices representative of the
structures at the corresponding location within an imaged volume.
The processing circuit is further configured to apply at least one
complementary filter to at least one of the two or more
radiographic projections and the one or more reconstructed slices
in a direction that is generally transverse to a scan path of the
radiation source. An operator workstation operably couple to the
processing circuit is configured to display the one or more
reconstructed slices.
[0014] In accordance with another aspect of the technique, a system
is provided for processing radiographic image data. The system
includes a radiation source capable of producing a stream of
radiation and a control circuit operably coupled to the radiation
source. In addition, the system includes a detector capable of
detecting the stream of radiation and generating two or more
radiographic projections and a motor controller configured to
displace at least one of the radiation source, the detector, and a
patient platform. A processing circuit operably coupled to the
detector and configured to receive the two or more radiographic
projections is also included. The processing circuit is further
configured to reconstruct the two or more radiographic projections
to form one or more reconstructed slices representative of one or
more structures within an imaged volume. The processing circuit
includes a means for reducing contrast asymmetry in the one or more
reconstructed slices. An operator workstation operably coupled to
the processing circuit is configured to display the one or more
reconstructed slices.
[0015] In accordance with an additional aspect of the technique, a
tangible medium for processing radiographic image data is provided.
The tangible medium includes a routine for filtering two or more
radiographic projections in a direction that is generally
transverse to a scan path of a radiation source used to generate
the projections. A routine is also included for applying a
reconstruction algorithm to the filtered projections to generate
one or more reconstructed slices.
[0016] In accordance with another aspect of the technique, a
tangible medium for processing radiographic image data is provided.
The tangible medium includes a routine for applying a
reconstruction algorithm to two or more radiographic projections to
generate one or more reconstructed slices. A routine is also
included for filtering the one or more reconstructed slices in a
direction that is generally transverse to a scan path of the
radiation source used to generate the projections.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The foregoing and other advantages and features of the
invention will become apparent upon reading the following detailed
description and upon reference to the drawings in which:
[0018] FIG. 1 is a diagrammatical view of an exemplary imaging
system in the form of a tomosynthesis imaging system for use in
producing processed images in accordance with aspects of the
present technique;
[0019] FIG. 2 is a diagrammatical view of a physical implementation
of the tomosynthesis system of FIG. 1;
[0020] FIG. 3 is another diagrammatical view of a physical
implementation of the tomosynthesis system of FIG. 1, in which the
X-ray source moves along a linear track;
[0021] FIG. 4 is a view of the system of FIG. 2, in which the X-ray
source is seen to move in a rotationally symmetric manner relative
to the imaged anatomy;
[0022] FIG. 5 is a view of a tomosynthesis system of FIG. 3 in
which the X-ray source is seen to obtain exposures at different
locations along the linear track;
[0023] FIG. 6 is a state transition diagram depicting the formation
of a reconstructed image volume from projection data according to
one embodiment of the present technique;
[0024] FIG. 7 is a state transition diagram depicting the formation
of a reconstructed image volume from projection data according to a
another embodiment of the present technique; and
[0025] FIG. 8 is a state transition diagram depicting the formation
of a reconstructed image volume from projection data according to
yet another embodiment of the present technique.
DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS
[0026] FIG. 1 illustrates diagrammatically an imaging system 10
which may be used for acquiring and processing image data. In the
illustrated embodiment, the system 10 is a tomosynthesis system
designed both to acquire original image data, and to process the
image data for display and analysis in accordance with the present
technique. In the embodiment illustrated in FIG. 1, the imaging
system 10 includes a source 12 of X-ray radiation which is freely
movable generally within a plane. In this exemplary embodiment, the
X-ray radiation source 12 typically includes an X-ray tube and
associated support and filtering components.
[0027] A stream of radiation 16 is emitted by the source 12 and
passes into a region in which a subject, such as a human patient 18
is positioned. A portion of the radiation 20 passes through or
around the subject and impacts a detector array, represented
generally at reference numeral 22. Detector elements of the array
produce electrical signals that represent the intensity of the
incident X-ray beam. These signals are acquired and processed to
reconstruct an image of the features within the subject. A
collimator 23 may define the size and shape of the X-ray beam 16
that emerges from the X-ray source 12.
[0028] Source 12 is controlled by a system controller 24 which
furnishes both power and control signals for tomosynthesis
examination sequences, including positioning of the source 12
relative to the patient 18 and the detector 22. Moreover, detector
22 is coupled to the system controller 24, which commands
acquisition of the signals generated in the detector 22. The system
controller 24 may also execute various signal processing and
filtration functions, such as for initial adjustment of dynamic
ranges, interleaving of digital image data, and so forth. In
general, system controller 24 commands operation of the imaging
system to execute examination protocols and to process acquired
data. In the present context, system controller 24 also includes
signal processing circuitry, typically based upon a general purpose
or application-specific digital computer, associated memory
circuitry for storing programs and routines executed by the
computer, as well as configuration parameters and image data,
interface circuits, and so forth.
[0029] In the embodiment illustrated in FIG. 1, system controller
24 is coupled to a positional subsystem 26 which positions the
X-ray source 12 relative to the patient 18 and the detector 22. In
alternative embodiments the positional subsystem 26 may move the
detector 22 or even the patient 18 instead of the source 12. In yet
another embodiment, more than one component may be movable,
controlled by positional subsystem 26. Thus, radiographic
projections may be obtained at various angles through the patient
18 by changing the relative positions of the source 12, the patient
18, and the detector 22 via the positional subsystem 26.
[0030] Additionally, as will be appreciated by those skilled in the
art, the source of radiation may be controlled by an X-ray
controller 30 disposed within the system controller 24.
Particularly, the X-ray controller 30 is configured to provide
power and timing signals to the X-ray source 12. A motor controller
32 may be utilized to control the movement of the positional
subsystem 26.
[0031] Further, the system controller 24 is also illustrated
comprising a data acquisition system 34. In this exemplary
embodiment, the detector 22 is coupled to the system controller 24,
and more particularly to the data acquisition system 34. The data
acquisition system 34 receives data collected by readout
electronics of the detector 22. The data acquisition system 34
typically receives sampled analog signals from the detector 22 and
converts the data to digital signals for subsequent processing by a
computer 36.
[0032] The computer 36 is typically coupled to the system
controller 24. The data collected by the data acquisition system 34
may be transmitted to the computer 36 and moreover, to a memory 38.
It should be understood that any type of memory adapted to store a
large amount of data may be utilized by such an exemplary system
10. Also the computer 36 is configured to receive commands and
scanning parameters from an operator via an operator workstation
40, typically equipped with a keyboard and other input devices. An
operator may control the system 10 via the input devices. Thus, the
operator may observe the reconstructed image and other data
relevant to the system from computer 36, initiate imaging, and so
forth.
[0033] A display 42 coupled to the operator workstation 40 may be
utilized to observe the reconstructed image and to control imaging.
Additionally, the image may also be printed on to a printer 43
which may be coupled to the computer 36 and the operator
workstation 40. Further, the operator workstation 40 may also be
coupled to a picture archiving and communications system (PACS) 44.
It should be noted that PACS 44 may be coupled to a remote system
46, radiology department information system (RIS), hospital
information system (HIS) or to an internal or external network, so
that others at different locations may gain access to the image and
to the image data.
[0034] It should be further noted that the computer 36 and operator
workstation 46 may be coupled to other output devices which may
include standard or special purpose computer monitors and
associated processing circuitry. One or more operator workstations
40 may be further linked in the system for outputting system
parameters, requesting examinations, viewing images, and so forth.
In general, displays, printers, workstations, and similar devices
supplied within the system may be local to the data acquisition
components, or may be remote from these components, such as
elsewhere within an institution or hospital, or in an entirely
different location, linked to the image acquisition system via one
or more configurable networks, such as the Internet, virtual
private networks, and so forth.
[0035] Referring generally to FIG. 2, an exemplary imaging system
utilized in a present embodiment may be a tomosynthesis imaging
system 50. In an arrangement similar to that described above, the
tomosynthesis imaging system 50 is illustrated with a source 12 and
a detector 22 between which a patient 18 may be disposed. The
source of radiation 12 typically includes an X-ray tube which emits
X-ray radiation from a focal point 52. The stream of radiation is
directed towards a particular region of the patient 18. It should
be noted that the particular region of the patient 18 is typically
chosen by an operator so that the most useful scan of a region may
be made.
[0036] In typical operation, X-ray source 12 projects an X-ray beam
from the focal point 52 and toward detector array 22. The detector
22 is generally formed by a plurality of detector elements,
generally corresponding to pixels, which sense the X-rays that pass
through and around a subject of interest, such as particular body
parts, for instance the chest, lungs and so on. In one embodiment,
the detector consists of a 2,048.times.2,048 rectangular array of
elements which correspond to a pixel size of 200 .mu.m.times.200
.mu.m, though other configurations and sizes of both detector 22
and pixel are of course possible. Each detector element produces an
electrical signal that represents the intensity of the X-ray beam
at the position of the element at the time the beam strikes the
detector. Furthermore, the source 12 may be moved generally within
a source plane 54, which is substantially parallel to the plane of
the detector 22, so that a plurality of radiographic views from
different view angles may be collected by the computer 36. In one
embodiment the distance between the source 12 and the detector 22
is approximately 180 cm and the total range of motion of the source
12 is between 31 cm and 131 cm, which translates to .+-.5.degree.
to .+-.20.degree. where 0.degree. is a centered position. In this
embodiment, typically at least 10 projections are acquired,
covering the full angular range.
[0037] The computer 36 is typically used to control the entire
tomosynthesis system 50. The main computer that controls the
operation of the system may be adapted to control features enabled
by the system controller 24. Further, the operator workstation 40
is coupled to the computer 36 as well as to a display, so that the
reconstructed image may be viewed.
[0038] As the X-ray source 12 is moved generally within plane 54,
the detector 22 collects data of the attenuated X-ray beams. Data
collected from the detector 22 then typically undergo
pre-processing and calibration to condition the data to represent
the line integrals of the attenuation coefficients of the scanned
objects. The processed data, commonly called projections, are then
typically backprojected to formulate an image of the scanned area.
In tomosynthesis, a limited number of projections are acquired,
typically twenty or less, each at a different angle relative to the
patient and detector. Because tomosynthesis techniques acquire such
a limited number of projections which are not densely spaced over
the full angular range, the information available for image
formulation is limited and the reconstruction problem is therefore
difficult. The reconstruction algorithms employed to perform the
reconstruction on this limited data efficiently separate tissue
layers, minimize artifacts, and enhance image contrast,
particularly of small structures. To ensure image quality,
reconstruction algorithms may employ a re-projection consistency
constraint that requires that an object corresponding to the
reconstruction, if exposed to X-rays, reproduce the initial
images.
[0039] Once reconstructed, the image produced by the system of
FIGS. 1 and 2 reveals the three-dimensional relationship of
internal features of the patient 18. The image may be displayed to
show these features and their three-dimensional relationships.
Though the reconstructed image may comprise a single reconstructed
slice representative of structures at the corresponding location
within the imaged volume, more than one slice is typical.
[0040] In traditional approaches to diagnosis of medical
conditions, such as disease states, and more generally of medical
events, a radiologist or physician would consider a hard copy of
the image, produced by the printer 43 or on photographic film, to
discern characteristic features of interest. Such features might
include nodules, lesions, sizes and shapes of particular anatomies
or organs, and other features which would be discernable in the
image based upon the skill and knowledge of the individual
practitioner. Other analyses may be based upon soft-copy reading,
volume rendering of the reconstructed three dimensional dataset, or
capabilities of various Computer-Aided Diagnosis or Detection (CAD)
algorithms which offer the potential for identifying, or at least
localizing, certain features of interest, such as anatomical
anomalies. Subsequent processing and data acquisition is then,
typically, at the discretion and based upon the expertise of the
practitioner.
[0041] Localized variations in the contrast of the reconstructed
image, however, may impair the analysis of the image by either the
practitioner or the CAD algorithm. These contrast variations may
arise due to the acquisition geometry, such as non-symmetric source
trajectories, in conjunction with the shape and orientation of
structures within the imaged volume and due to the reconstruction
algorithms employed in reconstruction. In particular, some
reconstruction algorithms, particularly those utilizing
re-projection consistency constraints, generate reconstructed
slices which appear to have been high-pass filtered even in the
absence of such a filtering step. However, the perceived filtering
by the reconstruction algorithm is only in the direction
substantially parallel to the scan path of the source 12. The
perceived filtering results in the variations in contrast in the
final image which are associated with the shape and orientation of
the imaged structures. In particular, reconstructed data sets
exhibit a more widespread blurring and a lower contrast for
structures that are long in a direction generally disposed parallel
to the scan path 56 of the source 12, and a more localized blurring
and a higher contrast for structures that are short in a direction
generally disposed parallel to the scan path 56 of the source
12.
[0042] Because of this perceived filtering caused by the
reconstruction algorithm, the acquisition geometry, i.e., the
motion of the source 12, and the orientation of the internal
structures within the imaging volume, may adversely interact to
produce blurring and other artifacts within the image, and
introduce a variation of the structure contrast in the
reconstruction. The extent of the variation in contrast and the
blurring is proportional to the length of the structure in a
direction which is roughly parallel to the scan path of the source
12. This is depicted in FIG. 3, in which the scan path 56 of the
source 12 is seen to be generally parallel to the long axis 58 of a
first internal structure 60. Conversely, the scan direction 56 is
only generally parallel to the short axis 62 of a second internal
structure 64. The degree of contrast variation and blurring in the
reconstructed final image associated with the long axis 58 and the
short axis 62 will be generally linearly proportional to the their
respective lengths and the actual orientation of the axes 58 and 62
with respect to the scan path 56. As a consequence, the same
structure 60 will exhibit a far lower contrast in the
reconstruction if its long axis is substantially parallel to the
scan path 56 than if it is transverse (e.g., substantially
perpendicular) thereto. That is, the orientation of a structure in
the image volume at least partially determines the contrast
observed for that structure in the reconstructed image. Due to this
relationship, linear or elongated scan paths yield reconstructed
images with contrast and blur artifacts associated with those
internal dimensions in a direction which is substantially parallel
to the scan path 56.
[0043] The use of non-elongated or circular symmetric scan paths,
such as the circular path 66 depicted in FIG. 4, reduces the
contrast variation associated with scan path and structure shape
and orientation. However, the use of non-elongated or circular
symmetric scan paths may not always be desired in terms of system
simplicity and construction or in terms of scan speed. Indeed, a
tomosynthesis system 50 which limits the source 12 to a linear scan
path 56, such as that depicted in FIG. 5, may be particularly
desirable for reasons of system simplicity.
[0044] For purposes of illustration, the linear scan paths 56 have
been discussed and illustrated as occurring within a generalized
source plane 54. However, more general source trajectories can also
be used. For example, a non-planar source trajectory that is
substantially linear relative to an imaged structure, such as an
arc, may also be used. Indeed, any general three-dimensional source
trajectory may be utilized in accordance with the present
technique.
[0045] Referring now to FIG. 6, a method for addressing the
contrast asymmetry due to the perceived filtering effect produced
by the reconstruction algorithms will be discussed in greater
detail. As depicted in FIG. 6, the acquired projection data 70 are
processed by applying a reconstruction algorithm 72. The
reconstruction process produces reconstructed slices 74 which
together constitute a reconstructed image volume 76. To address the
perceived contrast asymmetry introduced by the reconstruction
algorithm, a complementary filter or a combination of filters
constituting a complementary filter may be applied to the
reconstructed slices 74, as indicated at reference numeral 78.
Application of the complementary filter results in filtered slices
80 which do not possess the contrast asymmetry produced by
application of the reconstruction algorithm alone.
[0046] In one embodiment, the complementary filter may be an
asymmetric linear filter which is generally high-pass in character.
In this embodiment, the complementary filter may act in only one
direction, such as transverse to the scan path 56 of the source 12.
Because of the combined effects of the perceived high-pass
filtering introduced by the reconstruction algorithm and of the
complementary filter, the filtered slices 80 appear to be filtered
in two generally orthogonal directions, i.e., parallel to and
transverse to the scan path 56 of the source 12. Custom filters may
be used as the complementary filter, as well as known linear
filters, such as a Butterworth filter. Where custom filters are
employed, they may be designed to complement the apparent filtering
introduced by reconstruction such that the resulting filtered
slices 80 are substantially symmetric in appearance. For example, a
one-dimensional filter that approximates the apparent high-pass
filter characteristics of the reconstructed slice can be applied to
the reconstructed slices 74 in a direction transverse, e.g.
substantially perpendicular, to the scan path 56 to provide the
desired symmetric character to the filtered slices 80.
[0047] Additionally, the complementary filter may be chosen or
designed such that the resulting filtered slices 80 appear to have
been filtered with a rotationally symmetric filter. In particular,
a suitable complementary filter can be designed in the Fourier
domain, which results in a rotationally symmetric filtering
characteristic. For example, the two-dimensional Fourier transform
of a rotationally symmetric filter or signal is rotationally
symmetric. Therefore, in the Fourier domain, one can easily derive
a rotationally symmetric counterpart of a one-dimensional filter
which approximates the filtering characteristics introduced by the
reconstruction algorithm. Because a sequence of linear filtering
steps is equivalent to multiplying their respective Fourier
transforms, the resulting two-dimensional Fourier transform can be
used to derive a two-dimensional filter or combination of
one-dimensional filters. This derived filter or filters constitute
the complementary filter which, in combination with the perceived
reconstruction filtering effect, produce filtered slices 80 that
appear to have been filtered by a rotationally symmetric filter.
The rotationally symmetric filtering effect has the particular
benefit of equally enhancing the contrast of structures at any
orientation.
[0048] In addition to the linear filters discussed above,
non-linear and adaptive filters may also be employed as the
complementary filter. For example, median filters may be employed
as complementary filters and may have the advantage of removing
image noise while retaining edge information. Similarly,
complementary filters may be designed based upon order statistics
or robust statistics which have a high-pass character and which can
enhance structures of a given size, such as small structures, or of
a given shape. Likewise, other types of non-linear filters, such as
polynomial filters, can be used as complementary filters.
[0049] Similarly, the complementary filtering steps in the process
may consist of a multiscale filtering method which acts as a
high-pass filter. The multiscale filtering method may be used to
selectively enhance image structures of a given size. Multiscale
filtering methods may include wavelet transforms, wavelet packet
transforms, Laplacian pyramid representations, as well as other
methods that can be developed using appropriate smoothing filters
at different scales, and linear combinations of such filters. The
multiscale filter methods in general can be used to decompose an
image into a sequence of images of different scale, and the
original image can be obtained by recombining the images at
different scales in a suitable way. By introducing different
weighting factors in the recombination process, one obtains a
similar image as the original image, where structures at different
scales appear to be enhanced or suppressed.
[0050] In selecting an appropriate complementary filter, factors
such as the specific characteristics of the reconstruction
filtering effect, such as impulse response, may be considered.
These characteristics may vary with the specific system geometry
used during image acquisition, particularly the angular range over
which projections were acquired. In addition, filter selection may
consider such factors as computational speed, implementation in the
spatial or frequency domain, denoising characteristics of filters,
and the desirability of special high-pass filters such as unsharp
masking filters. In addition to the linear, non-linear, and
multi-scale filtering methodologies discussed above, other types of
filters which complement the perceived filter effect introduced
during reconstruction 72 to improve the contrast symmetry in
filtered slices 80 are suitable complementary filters.
[0051] While application of the complementary filter 78 to the
reconstructed slices 74 is one method of improving contrast
symmetry in the reconstructed slices, an alternative method is
depicted in FIG. 7. In the method of FIG. 7, the complementary
filter 78 is applied to the projection data 70 as opposed to the
reconstructed slices 74. A filtered projection data set 82 results
from the application of the complementary filter to the projection
data 70, and it is this filtered projection data 82 to which the
reconstruction algorithm is applied 72. Reconstructed slices 74 are
thereby formed which constitute the reconstructed volume 76. To the
extent that the reconstruction process and the complementary
filtering process are both linear in nature, it is arbitrary
whether the step of applying the complementary filer 78 occurs
prior or subsequent to the application of the reconstruction
algorithm 72. For computational reasons, however, it may be
advantageous to apply the complementary filter 78 to the projection
data 70 as opposed to the reconstructed slices 74. In particular,
because the number of projections is typically less than the number
of reconstructed slices, it is generally more computationally
efficient to filter the projections. Furthermore, since some
reconstruction algorithms consist of a filtering step followed by a
backprojection step, it may be computationally advantageous to
integrate the filtering that is part of the reconstruction and the
complementary filter into a single filtering step.
[0052] If, however, the reconstruction process or the complementary
filtering process are non-linear or non-additive, the order in
which the complementary filter is applied 78 and the reconstruction
algorithm is applied 72 is selected accordingly. In addition, in
such non-linear reconstructions, there may be advantages to
applying a complementary filter both to the projection data 70 and
the reconstructed slices 74, as depicted in FIG. 8. In such
instances the application of the complementary filter 78 may be
split between the two applications, or separate complementary
filters may be applied in each instance to achieve the desired
contrast symmetry in the filtered slices 80.
[0053] The application of a complementary filter, whether by the
techniques discussed in regard to FIGS. 6, 7, or 8, provides a
mechanism for correcting the orientation specific contrast
variations attributable to the reconstruction algorithm or other
factors. These techniques may be used in conjunction with linear or
elongated scan paths 56 to introduce contrast symmetry into the
final image, thereby improving the diagnostic value of the image.
In addition, it is not necessary that the source 12 move to produce
the contrast asymmetries addressed. Indeed, the patient 18 or the
detector 22 may instead be moved relative to the source 12. The
discussed techniques are also applicable in situations in which the
patient 18 or detector 22 are moved to produce the noted
asymmetries. While the techniques have been discussed in the
context of medical imaging, other fields such as non-destructive
evaluation and testing or other non-invasive imaging situations may
also utilize these techniques. Indeed, the described techniques may
be applicable in any situation where the goal is to reconstruct
three-dimensional information about an imaged object from
projection radiographs.
[0054] The invention may be susceptible to various modifications
and alternative forms, and specific embodiments have been shown by
way of example in the drawings and have been described in detail
herein. However, it should be understood Rather, the invention is
to cover all modifications, equivalents, and alternatives falling
within the spirit and scope of the invention as defined by the
following appended claims.
* * * * *