U.S. patent application number 11/511656 was filed with the patent office on 2006-12-28 for imaging chain for digital tomosynthesis on a flat panel detector.
This patent application is currently assigned to GE Medical Systems Global Technology Company, Inc.. Invention is credited to Gopal B. Avinash, Bernhard E. H. Claus, Jeffrey W. Eberhard, Kadri N. Jabri, John P. Kaufhold, Stephen W. Metz, John M. Sabol.
Application Number | 20060291711 11/511656 |
Document ID | / |
Family ID | 33552725 |
Filed Date | 2006-12-28 |
United States Patent
Application |
20060291711 |
Kind Code |
A1 |
Jabri; Kadri N. ; et
al. |
December 28, 2006 |
Imaging chain for digital tomosynthesis on a flat panel
detector
Abstract
A method of creating and displaying images resulting from
digital tomosynthesis performed on a subject using a flat panel
detector is disclosed. The method includes the step of acquiring a
series of x-ray images of the subject, where each x-ray image is
acquired at different angles relative to the subject. The method
also includes the steps of applying a first set of corrective
measures to the series of images, reconstructing the series of
images into a series of slices through the subject, and applying a
second set of corrective measures to the slices. The method further
includes the step of displaying the images or slices according to
at least one of a plurality of display options.
Inventors: |
Jabri; Kadri N.; (Waukesha,
WI) ; Avinash; Gopal B.; (New Berlin, WI) ;
Metz; Stephen W.; (Greenfield, WI) ; Sabol; John
M.; (Sussex, WI) ; Eberhard; Jeffrey W.;
(Albany, NY) ; Claus; Bernhard E. H.; (Niskayuna,
NY) ; Kaufhold; John P.; (Schenectady, NY) |
Correspondence
Address: |
Patrick S. Yoder;FLETCHER YODER
P. O. Box 692289
Houston
TX
77269-2289
US
|
Assignee: |
GE Medical Systems Global
Technology Company, Inc.
|
Family ID: |
33552725 |
Appl. No.: |
11/511656 |
Filed: |
August 29, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10613591 |
Jul 3, 2003 |
|
|
|
11511656 |
Aug 29, 2006 |
|
|
|
Current U.S.
Class: |
382/132 |
Current CPC
Class: |
G06T 11/005 20130101;
G06T 5/00 20130101; G06T 2207/10112 20130101; G06T 11/008 20130101;
G06T 2207/30004 20130101; G06T 2211/436 20130101 |
Class at
Publication: |
382/132 |
International
Class: |
G06K 9/00 20060101
G06K009/00 |
Claims
1.-54. (canceled)
55. A method of creating and displaying images resulting from
digital tomosynthesis performed on a subject using a flat panel
detector comprising: positioning a subject between an x-ray source
and a flat panel detector; acquiring image data from the detector
at a plurality of different angular positions of the source
relative to the subject; processing the image data to form a
two-dimensional image and a tomosynthesis image; selectively using
an anti-scatter grid during acquisition of at least a portion of
the image data; and selectively displaying the two-dimensional and
tomosynthesis images.
56. The method of claim 55, wherein during acquisition of the image
data the x-ray source is moved with respect to the subject.
57. The method of claim 55, wherein during acquisition of the image
data the detector is moved with respect to the subject.
58. The method of claim 55, further comprising identifying a
feature of interest in at least one of the two-dimensional image
and the tomosynthesis image.
59. An imaging system comprising: an x-ray source, and a flat panel
x-ray detector, the source and/or the detector being selectively
movable with respect to a subject disposed therebetween to
different angular positions with respect to the subject; a computer
coupled to the source and to the detector for acquiring image data
from the detector during emission of x-rays from the source at a
plurality of angular positions of the source and/or the detector
with respect to the subject; and an anti-scatter grid selectively
positionable with respect to the source and detector during
acquisition of the image data; the computer being configured to
form at least one two-dimensional projection image and
tomosynthesis images based upon the image data.
60. The imaging system of claim 59, wherein during acquisition of
the image data the x-ray source is moved with respect to the
subject.
61. The imaging system of claim 59, wherein during acquisition of
the image data the detector is moved with respect to the
subject.
62. The imaging system of claim 59, wherein the computer is further
configured to identify a feature of interest in at least one of the
two-dimensional image and the tomosynthesis image.
63. An imaging system comprising: an x-ray source, and a flat panel
x-ray detector, the source and/or the detector being selectively
movable with respect to a subject disposed therebetween to
different angular positions with respect to the subject; a computer
coupled to the source and to the detector for acquiring image data
from the detector during emission of x-rays from the source at a
plurality of angular positions of the source and/or the detector
with respect to the subject; the computer being configured to form
at least one two-dimensional projection image and tomosynthesis
images of selected slices through the subject based upon the image
data; and a display for displaying two dimensional projection image
and the tomosynthesis images.
64. The imaging system of claim 63, wherein during acquisition of
the image data the x-ray source is moved with respect to the
subject.
65. The imaging system of claim 63, wherein during acquisition of
the image data the detector is moved with respect to the
subject.
66. The imaging system of claim 63, wherein the computer is further
configured to identify a feature of interest in at least one of the
two-dimensional image and the tomosynthesis image.
67. An imaging system comprising: an x-ray source, and a flat panel
x-ray detector, the source and/or the detector being selectively
movable with respect to a subject disposed therebetween to
different angular positions with respect to the subject; a computer
coupled to the source and to the detector for acquiring image data
from the detector during emission of x-rays from the source at a
plurality of angular positions of the source and/or the detector
with respect to the subject; the computer being configured to form
at least one two-dimensional projection image for analysis and a
plurality of tomosynthesis images of selected slices through the
subject based upon the image data; and a display for selectively
displaying two dimensional projection image for analysis and the
plurality of tomosynthesis images.
68. The imaging system of claim 67, wherein during acquisition of
the image data the x-ray source is moved with respect to the
subject.
69. The imaging system of claim 67, wherein during acquisition of
the image data the detector is moved with respect to the
subject.
70. The imaging system of claim 67, wherein the computer is further
configured to identify a feature of interest in at least one of the
two-dimensional image and the tomosynthesis image.
71. An imaging system comprising: an x-ray source, and a flat panel
x-ray detector, the source and/or the detector being selectively
movable with respect to a subject disposed therebetween to
different angular positions with respect to the subject; a computer
coupled to the source and to the detector for acquiring image data
from the detector suitable for reconstruction in a two-dimensional
image or in a three-dimensional tomosynthesis image during emission
of x-rays from the source at a plurality of angular positions of
the source and/or the detector with respect to the subject; the
computer being configured to selectively form at least one or a
two-dimensional projection image and a plurality of tomosynthesis
images based upon the image data; and a display for selectively
displaying two dimensional projection image or the plurality of
tomosynthesis images.
72. The imaging system of claim 71, wherein during acquisition of
the image data the x-ray source is moved with respect to the
subject.
73. The imaging system of claim 71, wherein during acquisition of
the image data the detector is moved with respect to the
subject.
74. The imaging system of claim 71, wherein the computer is further
configured to identify a feature of interest in at least one of the
two-dimensional image and the tomosynthesis image.
75. A method of creating and displaying images using a flat panel
detector comprising: positioning a subject between an x-ray source
and a flat panel detector; acquiring image data from the detector
suitable for reconstruction in a two-dimensional image or in a
three-dimensional tomosynthesis image during emission of x-rays
from the source; selectively processing the image data to form a
two-dimensional image or a plurality of tomosynthesis images; and
selectively displaying the two-dimensional image or the
tomosynthesis images.
Description
BACKGROUND OF THE INVENTION
[0001] The present invention relates generally to the field of
digital tomosynthesis. More specifically, the present invention
relates to an imaging chain for digital tomosynthesis on a flat
panel detector.
[0002] Digital tomosynthesis imaging is a technique that requires
the acquisition of multiple x-ray images at different angles
relative to the patient within a short time interval. Once these
images have been acquired, a reconstruction algorithm is applied to
the data represented by the images to reconstruct "slices" through
the patient. These slices, which are essentially re-constructed
x-ray images of selected planes within an object or patient, may
eliminate any structures underlying or overlying a particular area
or region of interest and thereby allow for improved diagnosis and
treatment.
[0003] The standard digital radiography (DR) image acquisition,
processing, and display chain was not designed with tomosynthesis
in mind. As a result, the use of standard digital radiography
processes and procedures presents a number of potential problems
when used in the performance of digital tomosynthesis. For example,
patient motion in-between the successive acquisitions may result in
images that include motion artifacts. Similarly, physiologic motion
(e.g., motion of the heart, lungs, etc.) in-between the successive
acquisitions may also result in images that include motion
artifacts. Other potential problems, such as intensity and
resolution non-uniformities, may arise as a result of the
angulation of the source of the x-rays relative to the detector.
Yet another potential problem is that the use of large angulation
ranges may result in increased scatter when no grid is used.
Furthermore, errors and uncertainty in the positioning of the
source and the detector may result in image reconstruction
artifacts. Still another potential problem is that the reduced
exposure used in tomosynthesis (relative to the standard single
acquisition) may result in increased noise being present in the
resulting images.
[0004] While various efforts have been made to address some of
these potential problems, these efforts have generally been
narrowly focused in one particular problem area. Moreover, these
efforts have generally failed to address several potential
opportunities that may be possible due to the additional
information and data provided by digital tomosynthesis. One such
opportunity involves the non-disruptive incorporation of
three-dimensional imaging techniques into a traditional
two-dimensional imaging system and workflow. Another such
opportunity relates to the application of computer aided detection
(CAD) algorithms to the additional image information that is
generated by tomosynthesis. Still another opportunity is presented
to create and utilize new visualization techniques that will
enhance the diagnostic value of the additional information
generated by tomosynthesis.
[0005] It would be advantageous to provide a system or method of
addressing, overcoming, or reducing the impact of more than a
narrow subset of the problems that may arise as a result of using
the standard digital radiography image acquisition, processing, and
display chain for tomosynthesis. It would also be advantageous to
provide a system or method that capitalizes on any one or more of
the potential opportunities presented by digital tomosynthesis.
Accordingly, it would be advantageous to provide a system or method
that has any one or more of these or other advantageous
features.
SUMMARY OF THE INVENTION
[0006] The present invention relates to a method of creating and
displaying images resulting from digital tomosynthesis performed on
a subject using a flat panel detector. The method comprises the
step of acquiring a series of x-ray images of the subject, where
each x-ray image is acquired at different angles relative to the
subject. The method also comprises the steps of applying a first
set of corrective measures to the series of images, reconstructing
the series of images into a series of slices through the subject,
and applying a second set of corrective measures to the slices. The
method further comprises the step of displaying the images or
slices according to at least one of a plurality of display
options.
[0007] The present invention also relates to a system for creating
and displaying images of the internal structures of a subject
resulting from digital tomosynthesis performed with a flat panel
digital detector. The system comprises a means for acquiring a
series of x-ray images of the subject, where each x-ray image is
acquired at different angles relative to the subject. The system
also comprises a means for applying a first set of corrective
measures to the series of images, a means for reconstructing the
series of images into a series of slices through the subject, and a
means for applying a second set of corrective measures to the
slices. The system further comprises a means for displaying the
images or slices according to at least one of a plurality of
display options.
[0008] The present invention further relates to a method of
creating and displaying images of the anatomy of a patient using
digital tomosynthesis performed with a flat panel detector and
other equipment. The method includes the step of receiving inputs
relating to options for acquiring x-ray images of the patient,
where the options allow for the selection of at least one of a
field of view, a method of controlling the dose of the x-rays, the
energy level or levels at which the images will be acquired, how a
source and a detector will move while the images are acquired,
whether a large field of view is desired, the acquisition paths of
the source and the detector, and the characteristics of slices to
be constructed from the x-ray images. The method also comprises the
steps of acquiring a single x-ray image of the patient and
adjusting parameters relating to the acquisition of x-ray images.
The parameters relating to the acquisition of the x-ray images
include at least one of x-ray technique parameters, filtration
techniques, position of acquisition, and angle of the acquisition.
The method further includes continuing to acquire a single x-ray
image and to then adjust the acquisition parameters until a
sufficient number of images have been acquired. The method also
includes applying detector corrections, intensity corrections, and
geometric corrections to one or more of the images and performing
at least one of frequency filtering for structure enhancement,
tissue equalization, spatial filtering, and image resizing on one
or more of the acquired images. The method further includes the
steps of constructing at least one slice through the patient by
applying a 3D reconstruction algorithm to the data represented by
the acquired images, removing artifacts from one or more of the
slices, and enhancing information provided in the one or more
slices. Additionally, the method includes the step of optimizing
the display of one or more of the slices by performing at least one
of edge enhancement, tissue equalization, display window level
adjustment, and display window width adjustment. The method also
includes the step of displaying one or more of the slices as one of
a two-dimensional or three-dimensional image or set of images.
[0009] The present invention still further relates to a method of
adjusting the acquisition parameters for the acquisition of images
during tomosynthesis performed on a subject. The method comprises
the step of acquiring a first image of the subject, where the first
image provides information relating to the subject. The method also
comprises the steps of selecting the acquisition parameters for the
acquisition of a second image based on the information provided by
the first image and acquiring the second image according to the
selected acquisition parameters.
[0010] Other features and advantages of the invention will become
apparent to those skilled in the art upon review of the following
detailed description, claims, and drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1 is a schematic diagram of a tomosynthesis system
according to one embodiment of the invention.
[0012] FIG. 2 is a flowchart showing the steps in a tomosynthesis
imaging chain according to a preferred embodiment.
[0013] FIG. 3 is a flowchart showing the first step illustrated in
the tomosynthesis imaging chain of FIG. 2.
[0014] FIG. 4 is a flowchart showing the second step illustrated in
the tomosynthesis imaging chain imaging chain of FIG. 2.
[0015] FIG. 5 is a flowchart showing the third step illustrated in
the tomosynthesis imaging chain imaging chain of FIG. 2.
[0016] FIG. 6 is a flowchart showing the fourth step illustrated in
the tomosynthesis imaging chain imaging chain of FIG. 2.
[0017] FIG. 7 is a flowchart showing the fifth step illustrated in
the tomosynthesis imaging chain imaging chain of FIG. 2.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0018] Referring to FIG. 1, a tomosynthesis system 10 is shown
schematically according to a preferred embodiment. Tomosynthesis
system 10 includes an x-ray source 20, a detector 30, a computer
40, and supporting structure 50.
[0019] X-ray source 20 is directed toward a subject 21 (e.g.,
object, patient, etc.) and is configured to emit a beam of x-rays
22 at desired times. Once x-rays 22 are emitted, they pass through
subject 21 and are picked up by, or hit, detector 30.
[0020] Detector 30 (e.g., x-ray detector, digital radiography
detector, flat panel detector, flat detector, etc.) may be any one
of a variety of different detectors conventionally known within the
art or that will become available in the future (e.g., energy
discriminating detectors that are theoretically capable of
acquiring high and low energy images simultaneously). However,
according to a preferred embodiment, detector 30 is a flat panel
digital detector. When x-rays 22 are picked up by detector 30, they
are converted into electrical signals that are sent to computer 40.
The electrical signals will vary depending on a number of factors,
including the angle at which x-rays 22 hit detector 30, the
intensity of the different x-rays that hit detector 30, and a
number of other factors. Based on these electrical signals,
computer 40 is then capable of creating an image of the internal
structures of subject 21.
[0021] Computer 40 (e.g., processor, controller, etc.) includes
processing circuitry that executes stored program logic and may be
any one of a variety of different computers, processors, or
controllers (or combination thereof) that are available for and
compatible with the various types of equipment and devices used in
tomosynthesis system 10. Through its various processors and
controllers computer 40 controls the operation and function of
source 20 and detector 30. For example, computer 40 may control,
among other functions and operations, when source 20 emits x-rays,
how detector 30 reads and conveys information or signals after the
x-rays hit detector 30, and how source 20 and detector 30 move
relative to one another and relative to subject 21. Computer 40
also controls how information (e.g. images or data acquired during
the tomosynthesis operation) is processed and displayed. The
different processing steps performed by computer 40 are dictated
and controlled by software designed to allow computer 40 to perform
the various operations underlying tomosynthesis. Information may
also be stored in computer 40 for later retrieval and use.
[0022] During the tomosynthesis operation or process, multiple
images of subject 21 are acquired from different perspectives or
angles. In order to acquire the images from different perspectives,
any one or more of source 20, detector 30, and subject 21 may move
relative to one or more of the others while the images are being
acquired. This motion may take place at the same time the images
are being acquired or in-between the different image acquisitions.
The movement of source 20, detector 30, and/or subject 21 (which
may be accomplished through a movable table or support structure,
which is not shown) are generally controlled by computer 40 based
on information entered into computer 40 by someone operating the
tomosynthesis equipment, based on pre-defined acquisition
protocols, or based on information that has already been acquired
by computer 40.
[0023] Referring to FIG. 2, an imaging chain 100 that is utilized
in the tomosynthesis process is shown according to a preferred
embodiment. Imaging chain 100 includes steps 200, 300, 400, 500,
600, 700, and 800.
[0024] During step 200 (e.g., the acquisition step), the patient is
prepared for the tomosynthesis process and the x-ray images are
acquired. Referring now to FIG. 3, step 200 can be broken down into
sub-steps 210, 220, 230, 240, 250, and 260. At sub-step 210, the
patient and x-ray equipment are prepared for the acquisition of
x-ray images. This preparation includes generally determining where
the x-rays will be focused, placing the patient in the appropriate
location, and preparing the x-ray equipment to take images or
acquisitions in the desired region of the patient. This may be done
through the use of external markers, by placing fiducial markers on
the patient, by placing calibration objects in the field of view,
through the use of light-field or laser positioning aids (e.g. a
cross-hair projected onto the patient that corresponds to the
target area), and/or by placing or selecting automatic exposure
control sensor(s), etc. According to alternative embodiments, other
methods may be used to prepare the patient and x-ray equipment.
Moreover, any of the methods may be used individually or in
combination with other methods.
[0025] At sub-step 220, a "pre-tomosynthesis" image or acquisition
(e.g. a localizer acquisition) is acquired in order to get
information relating to patient positioning, patient
characteristics, and/or acquisition characteristics or parameters.
With respect to patient characteristics, the image may provide
information such as body thickness and general anatomy and may
additionally help with the location and identification of any
implantable devices or non-standard structures (e.g. a missing
lung, an enlarged heart, etc.). This information, along with other
information the pre-tomosynthesis image may provide, may be used by
computer 40, or by the operator in a semi-automatic mode, as a
basis for optimizing the parameters or characteristics of
subsequent acquisitions. For example, the information provided by
the pre-tomosynthesis image may be used as a basis for optimizing
the energy level of the x-rays used in the subsequent acquisitions,
the pulse duration, the tube current, the tube current duration,
etc. The pre-tomosynthesis image may be acquired using an equal or
lower dose of x-rays than is used to acquire images in later steps
in imaging chain 100 (described below). According to one
embodiment, the image generated during sub-step 220 may not be
utilized during the reconstruction process (described below).
According to an alternative embodiment, the pre-tomosynthesis image
may be "re-used" in later steps of imaging chain 100 (e.g. one less
image may be needed during subsequent steps that would otherwise be
required in the absence of the localizer acquisition). According to
another alternative embodiment, sub-step 220 may not be part of
step 200 and may not be included in imaging chain 100. According to
another alternative embodiment, the information provided by any
image acquired during the tomosynthesis process may be used as a
basis for optimizing the acquisition parameters or characteristics
of subsequent acquisitions.
[0026] At sub-step 230 (e.g., the graphical prescription step), a
variety of parameters relating to the images that will be taken
during later steps in imaging chain 100 are selected and set. These
parameters relate to the field of view, the method used to control
the dose of the x-rays, the energy level of the x-rays, how the
x-ray source will be moved during the acquisitions, whether the
acquisition will require a field of view larger than the detector
area, the acquisition paths of the source and detector, the slice
characteristics, and the presence or absence of an anti-scatter
grid.
[0027] The field of view may be selected by specifying a region of
interest within the localizer acquisition taken in sub-step 220.
This may be done interactively on computer 40, which displays the
localizer acquisition, by indicating where in the resulting image
the x-rays should be targeted or focused. The field of view may
also be selected by specifying and entering coordinates consisting
of reference points defined relative to the patient. In addition, a
volume of interest within the patient may be defined by specifying
a region of interest in the image as well as a start and end height
above the detector (e.g., a thickness of the volume of interest).
The number of slices to be reconstructed and the slice separation
may also be defined. Slice separation may be predetermined,
determined by the acquisition configuration (e.g., the maximum
angle), or selected by the operator. In combination with the volume
of interest, a variable-opening collimator may be controlled so as
to optimally cover the volume of interest, while minimizing the
dose of x-rays received by the patient.
[0028] There are at least three methods that may be used to control
the dose of x-rays received by the patient. The first method is to
use ion chambers to automatically control the exposure of the
patient to x-rays. The second method is to fix the time/pulse-width
of the x-rays used to generate the acquisitions. The third method
is to automatically calculate an optimal dose using information
from the localizer image or from a previous acquisition in the
series of acquisitions acquired during the tomosynthesis process.
According to alternative embodiments, other conventional methods of
controlling the dose of x-rays received by a patient or subject may
be used.
[0029] The energy level of the x-rays can be set to a single energy
level or to multiple energy levels. For example, at each position
and angle of acquisition, a single image can be acquired at a
specific energy level, or several images can be acquired at
different energy levels. Moreover, the energy level may also vary
as a function of the projection angle.
[0030] With respect to selecting how the x-ray source will move
during acquisitions, the acquisition can be made while the x-ray
source is moving (e.g; a continuous acquisition) or after the
source settles into each position (e.g. a step & shoot
acquisition). The detector may also move, either during the
exposure, which may tend to minimize blurring in the continuous
scan, and/or between exposures, which tends to optimize the covered
volume. To achieve better image quality, or minimize required
corrections, the detector may be tilted towards the incident x-ray
beam.
[0031] If a large field of view (e.g. a field of view that is
larger than the detector area) is required, tomosynthesis system 10
may be configured to take multiple tomosynthesis sweeps in
succession, and then "paste" or "stitch" the corresponding acquired
images together before passing them to the next step in the imaging
chain. In an alternative embodiment, the acquired images are used
to reconstruct mote than one volume of interest, and the pasting or
stitching is performed after the 3D reconstruction.
[0032] The paths along which the source and detector travel during
the process of acquiring images may also be defined. These paths
(which may be one-dimensional, two-dimensional, three-dimensional,
etc.) are defined by the position or angular orientation (e.g.,
tilt) of the source and the detector as well as the angle of the
source relative to the detector. Moreover, factors such as the type
of clinical application, the portion of the anatomy that is of
interest, the volume that is of interest, and the size of the
patient may be taken into account in selecting the desired
path.
[0033] Other parameters that may be adjusted relate to the slice
characteristics. These parameters relate to the number of slices,
the slice thickness (which may be variable or fixed), the slice
orientation (e.g. the angle of the slices with respect to detector
plane), the start depth, and the end depth. It is also possible to
reconstruct on non-planar slices, where the shape or curvature of
the slices may be adapted to the anatomy to be imaged.
[0034] Still another parameter or option that may be selected is
whether to include the anti-scatter grid. If an anti-scatter grid
is selected, one of a plurality of available grids may then be
selected.
[0035] At sub-step 240, the time at which images are acquired is
linked to certain physiological signals or events, which is
referred to as physiologic gating. Physiologic gating helps to
maintain uniformity between the different acquisitions and to
increase the quality of the results of the tomosynthesis process.
According to one embodiment, a physiological signal such as a
patient's heart rhythm (EKG) or breathing cycle is detected and is
used as a basis for triggering the acquisition of images. According
to this embodiment (referred to as prospective physiologic gating),
the timing of the acquisition of images is linked to the
physiological events such that acquisitions are taken at certain
points or at certain intervals in the physiologic cycle. According
to another embodiment, the physiological events are recorded at the
time acquisitions are taken. According to this embodiment (referred
to as retrospective physiologic gating), the point or interval of
the physiological cycle at which the patient is in when the images
are acquired is taken into account in the reconstruction and
processing of the images in the later steps of imaging chain 100.
According to another alternative embodiment, sub-step 240 is not
included within imaging chain 100.
[0036] At sub-step 250 an acquisition is taken according to the
settings selected and applied in the previous steps. Then at
sub-step 260, the acquisition parameters are adjusted and another
acquisition is taken. Such adjustments to the acquisition
parameters may include, but are not limited to, adjustments to the
x-ray technique parameters (e.g. the energy level of the x-rays,
the pulse duration, the tube current, the tube current duration,
etc.), the filtration, the position of acquisition, the angle of
acquisition, etc. These adjustments are made to provide the variety
of different images (e.g., datasets) that will later be
reconstructed into the desired view. In making these adjustments,
the x-ray technique parameters may be the same (e.g. constant,
fixed, pre-determined) for all images or the parameters may be
varied between images. If adjustments are made to the x-ray
technique parameters between any of the images, the adjustments may
be based on information acquired from the pre-tomosynthesis image,
on information acquired from any previous image or images, or on
other relevant information. Sub-step 260 is repeated until a
sufficient number of images (e.g., datasets) have been obtained to
allow computer 40 to reconstruct the desired volume of interest.
The acquisition of a sufficient number of images from different
perspectives or acquisition angles allows computer 40 to construct
a three-dimensional dataset by suitably combining the individual
datasets that are represented by particular images. According to
alternative embodiments, the acquisitions may be taken while the
detector is moving or while the detector is stationary.
[0037] As illustrated in FIG. 2, the acquisitions taken in step 200
are processed at step 300. Step 300 (e.g. "pre-processing")
involves the processing of the images or acquisitions taken in step
200 to correct or modify various attributes or characteristics of
the images. Like step 200, step 300 can be broken down into a
number of different sub-steps, which are illustrated in FIG. 4.
[0038] At sub-step 310, various corrections (e.g., detector
corrections, etc.) are made to correct properties of the images
that arise as a result of the use of a detector, and in particular,
a flat panel digital detector. These corrections include bad
pixel/line correction, gain map correction, corrections specific to
dual energy acquisitions (if used) such as laggy pixel corrections,
etc.
[0039] At sub-step 320, intensity corrections are made. Intensity
corrections include corrections of variations due to the imaging
geometry, such as (1/r.sup.2) attenuation, heel effect, and tube
angulation. Intensity corrections may also include corrections to
the sensitivity map of the detector, corrections to offset the
effects of Modulation Transfer Function (MTF) variations, etc.
Finally, intensity corrections may also include corrections of
intensity variation due to use of different x-ray energies at
different positions/angles.
[0040] At sub-step 330, scatter corrections, which are particularly
important when no scatter grid is used during the acquisition of
the images, are made to reduce the effects of scatter. Scatter
corrections can be made using scatter reduction algorithms that use
information from multiple energy images to perform the correction.
Scatter correction can depend on the angle of acquisition or be
angle-independent.
[0041] Sub-steps 310, 320, and 330 can be used in combination to
achieve quantitative images in situations such as where the values
at each pixel correspond to the line integral of the attenuation
coefficient along the corresponding ray. Reference calibration
measurements also may be used to achieve quantitative images.
[0042] At sub-step 340, geometric corrections are made to reduce
the effects of any non-uniformities in the equipment setup or
operation. Such non-uniformities may include, for example,
deflection or sag in supporting structure 50 of tomosynthesis
system 10, which may cause the source and the detector to be
slightly out of position with respect to one another.
Non-uniformities may also include, among other things, jitter in
the track (e.g, railing, channel, guide, etc.--not shown) along
which source 20 and detector 30 move. Geometric corrections may be
based on calibration events or runs that are performed once after
the installation of tomosynthesis system 10 or they may be based on
calibration events or runs that are repeated on a periodic basis.
According to alternative embodiments, the geometric corrections may
be based on image information using, for example, fiducial or
anatomical markers. According to other alternative embodiments,
geometric corrections may be relative (i.e., the geometry used may
not be the "true" geometry) or absolute without compromising the
reconstructed image quality.
[0043] At sub-step 350, motion corrections are made to account for
any motion of the patient that may have occurred between
acquisitions (e.g., contractions of the heart, expansion or
contraction of the lungs, external movement, etc.). Motion
corrections are made by aligning (e.g., registering) the images
based on the anticipated position of external fiducial markers or
anatomic landmarks, including aligning or registering the multiple
energy images acquired at a single position and angle.
[0044] At sub-step 360, material decomposition is performed.
Material decomposition is applicable when multiple energy
acquisitions are used and serves to create separate images of
different tissue types (e.g., creates a separate image of bone and
a separate image of soft-tissue). Techniques such as
log-subtraction or basis material decomposition may be used to
perform the material decomposition.
[0045] At sub-step 370, any "noise" present in the images is
removed or reduced. The noise reduction process is based on noise
reduction algorithms. These algorithms may be applied to images
independently, or the algorithms may share information across
images.
[0046] At sub-step 380, various filtering techniques (e.g.
pre-reconstruction filtering) may be applied to the acquisitions.
Such filtering techniques may include frequency filtering for
specific tissue and/or structure enhancement, tissue equalization,
spatial filtering, image resizing/shrinking, etc. These operations
or techniques can be tailored to the specific reconstruction
technique or techniques used, to the acquisition parameters, and to
various attributes of the patient.
[0047] According to alternative embodiments, each of sub-steps 330,
350, 360, and 370 are optional steps that may not be included
within imaging chain 100.
[0048] As illustrated in FIG. 2, the acquisitions processed at a
step 300 are reconstructed at step 400. Step 400 (e.g.
"reconstruction") involves using the data and information from the
acquired images to construct an image (e.g. a "slice,"
reconstructed image, etc.) of the patient. Referring now to FIG. 5,
step 400 may include sub-steps 410, 420, and 430.
[0049] At sub-step 410, the acquisitions obtained in the previous
steps are reconstructed (e.g. constructed, transformed, rendered,
etc.) into one or more slices through the patient (or other object)
using a 3D reconstruction algorithm. The reconstruction algorithm
may employ a cone-beam geometry (which may allow for precise
measurements of the size of objects, but which may be
computationally somewhat slow), or a parallel beam geometry (which
is computationally fast, but which may result in variances of the
physical distances between pixels as a function of the height of
the reconstructed slice or image). The techniques through which the
acquisitions are reconstructed include shift and add, filtered back
projection (FBP), generalized filtered back projection (GFBP),
Fourier reconstruction, objective function-based reconstruction,
variations of the algebraic reconstruction technique (ART), matrix
inversion tomo-synthesis (MITS), order statistics-based
backprojection (OSBP), or any combination or these or other
reconstruction techniques. The reconstruction also may make use of
prior information, which may include, but is not limited to, a
geometric model of the relevant anatomy or physical constraints of
a patient. Such prior information may also include information
pertaining to the point in the physiological cycle at which the
patient or subject was in when a particular image was acquired.
Such prior information may further include information pertaining
to the chemical composition and associated attenuation spectra of
tissues in the body. Such prior information may additionally
include previously acquired medical scans of the patient, such as
x-ray tomosynthesis, CT, MR, and/or ultrasound imagery. Moreover,
the reconstruction may involve using additional images (e.g., using
an additional lateral (LAT) view in addition to the
posterio-anterior (PA) tomosynthesis sequence, or using a PA and a
LAT tomosynthesis sequence).
[0050] At sub-step 420, a deconvolution algorithm is used to help
remove any blur that may arise from sub-step 410. According to an
alternative embodiment, sub-step 420 may be excluded from step 400
and from imaging chain 100.
[0051] At sub-step 430, patient information is input into the 3D
reconstruction process (sub-step 410) and/or the deconvolution
process (sub-step 420) to improve or optimize the overall
reconstruction process. The patient information may include
information relating to current or historical physical and
pathological conditions (e.g., size, composition, abnormal anatomy,
etc.) and/or to the acquisition parameters of previous acquisitions
(e.g., the energy level of the x-rays, the pulse duration, the tube
current, the tube current duration, the filtration, the position of
acquisition, the angle of acquisition, etc.). To use such
information in the reconstruction process, patient qualitative
and/or quantitative model(s) are formed. The use of the patient
information in this manner may help to optimize reconstruction with
respect to the parameters of acquisition and/or the patient or
imaged anatomy. According to an alternative embodiment, sub-step
430 is an optional step that may be excluded from step 400 and from
imaging chain 100.
[0052] Referring now to FIGS. 2 and 6, the images (e.g. slices)
reconstructed in step 400 are then subjected to further processing
at step 500. Step 500 (e.g. "post-processing") involves the
processing of the images reconstructed in step 400 to correct,
remove, adjust, enhance, etc. various attributes or characteristics
of the images. The post-processing of step 500 can be broken down
into sub-steps, which are illustrated in FIG. 6.
[0053] At sub-step 510, the reconstructed images are filtered to
remove potential artifacts or attributes, such as streaking, that
may arise as a result of the reconstruction step 400. The
information conveyed by the images may also be enhanced. According
to alternative embodiments, this enhancement of the image
information may include, among other things, the removal of ribs
and direction filtering.
[0054] At sub-step 520, any residual motion artifacts contained
within the images created during step 400 are removed or reduced.
Such artifacts may include soft-tissue detail blurring, bone edge
blurring, heart contour shadowing, overshoot/undershoot at the
edges of an organ, etc.
[0055] At sub-step 530, noise reduction algorithms similar to those
utilized in sub-step 370 are applied to the reconstructed images to
reduce or eliminate the effects of "noise" within the images. The
noise reduction algorithms may be applied to the images
independently, or the algorithms may share information across
images.
[0056] At sub-step 540, various attributes of the reconstructed
images that relate to the presentation of those images are
processed. This processing may include edge enhancement, tissue
equalization, and the adjustment of the display window level and
window width for optimal display. Look-up tables for clinical
displays that are specific to certain applications also may be
applied. Appropriate dynamic range management (DRM) algorithms also
may be applied.
[0057] At sub-step 550, material decomposition techniques,
including log-subtraction and basis material decomposition, are
applied to the images when three-dimensional data sets from
multiple energy acquisitions have been reconstructed
separately.
[0058] According to alternative embodiments, each of sub-steps 510,
520, 530, 540 and 550 are optional steps that may not be included
within imaging chain 100.
[0059] Referring now to FIGS. 2 and 7, step 600 consists of
presenting and/or analyzing the information processed in step 500.
Like the previous steps, sub-step 600 can be broken down into
several sub-steps.
[0060] At sub-step 610, the region-of-interest (e.g., the
particular part of the subject or patient one wishes to examine) is
selected for display and visualization. According to alternative
embodiments, the selection of the region-of-interest may be
interactive (e.g., selected manually) or the selection may be
automatic or semi-automatic. An automatic or semi-automatic
selection may be based on the automatic localization of anotomical
features or other distinct features of the subject or patient.
[0061] At sub-step 620, a specific structure or tissue is segmented
for display and visualization. Such segmentation may be done for
each slice individually or it may be based on information acquired
across the different slices.
[0062] At sub-step 630, the image (or the data on which the image
is based) is reformatted and/or re-mapped. The reformatting and/or
re-mapping of sub-step 630 may include Multi-Planar Reformatting
(MPR) for "slicing" data sets at different angles, Maximum
Intensity Projection (MIP), or various other reformatting or
re-mapping techniques.
[0063] At sub-step 640, the image may be rendered. The rendering
may be surface or volume rendering and may include the adjustment
of transparency levels.
[0064] At sub-step 650, the rendered data set is displayed.
According to alternative embodiments, the viewing perspective and
other parameters may be controlled interactively or they may run in
a loop using predetermined settings. In one embodiment, some of the
display parameters may depend on specific parameters of the
acquisition. For example, the maximum viewing angle may be limited
as a function of the tube angles utilized during the
acquisition.
[0065] As an alternative to sub-steps 640 and 650, the data may be
viewed as a two-dimensional set of images in sub-step 660. This may
be done by looking at the images side-by-side, by looking at the
images in a cine loop according to a temporal display, by
interactively toggling between the different slices, or by using
any one of a plurality of other different two-dimensional viewing
techniques. In one embodiment, the two-dimensional images can be
generated from the reconstructed slices (e.g., by taking the
average of appropriate subsets of slices).
[0066] According to alternative embodiments, any one or more of
sub-steps 610, 620, 630, 640, and 650 is optional and may be
excluded from imaging chain 100.
[0067] Referring now to FIG. 2, imaging chain 100 also includes
step 700. Step 700 includes using computer 40 to aid or assist in
the processing and/or diagnosis of various attributes or
characteristics embodied within the acquisitions and corresponding
data, which is known as computer-aided detection (CAD). Step 700 is
performed using processing and diagnosis algorithms, which can be
general radiography algorithms or which can be tailored to
tomosynthesis slices and/or three-dimensional datasets. The CAD
algorithm may act on the projection images (e.g. the images upon
which the reconstructed slices are based), the reconstructed
slices, the full three-dimensional dataset, or any combination of
these. CAD may include a consistency check, for example in the case
of CAD acting on the projection images, where the suspicious
regions that are detected are linked via the reconstructed
three-dimensional geometry. Moreover, CAD results may be used as a
basis for the automatic choice of the region-of-interest for
display. Furthermore, using the results of segmentation and/or
quantitative images, CAD can provide quantitative results, such as
the size and/or thickness of lesions. CAD can also include temporal
analysis of datasets, for example temporal subtraction images,
combined with registration techniques. According to an alternative
embodiment, step 700 is an optional step that may be excluded from
imaging chain 100.
[0068] Referring still to FIG. 2, step 800 includes storing the
images and data for future retrieval, analysis, comparison, etc.
Once the images and data have been stored, they can be used as
input into a CAD system or can be directly viewed at a later
time.
[0069] According to a preferred embodiment, the various steps and
sub-steps described above are performed in the order in which they
are described and depicted in FIGS. 1-7. However, according to
alternative embodiments, the steps and sub-steps may be performed
in any order or sequence.
[0070] While the embodiments and application of the invention
illustrated in the figures and described above are preferred, it
should be understood that these embodiments are offered by way of
example only. Accordingly, the present invention is not limited to
a particular embodiment, but extends to various modifications that
nevertheless fall within the scope of the appended claims.
* * * * *