U.S. patent application number 10/861880 was filed with the patent office on 2005-02-10 for region of interest methods and systems for ultrasound imaging.
This patent application is currently assigned to Siemens Medical Solutions USA, Inc.. Invention is credited to Gardner, Edward A., Kane, Richard M., Main, Joan C..
Application Number | 20050033123 10/861880 |
Document ID | / |
Family ID | 34118825 |
Filed Date | 2005-02-10 |
United States Patent
Application |
20050033123 |
Kind Code |
A1 |
Gardner, Edward A. ; et
al. |
February 10, 2005 |
Region of interest methods and systems for ultrasound imaging
Abstract
In one embodiment, at least one visual characteristic from a
selected image is used to automatically select an ultrasound image
from a set of images from a plurality of heart cycles. In another
embodiment, motion correction is performed on ultrasound images
that are automatically selected from a plurality of ultrasound
images associated with the same phase of the heart cycle. In yet
another embodiment, ultrasound images are automatically selected
from a set of images based on a time interval that is within a
tolerance range from a reference phase of the heart cycle. In
another embodiment, a stored user-preference of a phase of the
heart cycle is used to automatically select an ultrasound image
from a set of images. In yet another embodiment, a user-defined
region of interest is placed on an ultrasound image associated with
a selected phase of a heart cycle.
Inventors: |
Gardner, Edward A.; (San
Jose, CA) ; Kane, Richard M.; (Los Altos, CA)
; Main, Joan C.; (Mountain View, CA) |
Correspondence
Address: |
Siemens Corporation
Attn: Elsa Keller, Legal Administrator
Intellectual Property Department
170 Wood Avenue South
Iselin
NJ
08830
US
|
Assignee: |
Siemens Medical Solutions USA,
Inc.
|
Family ID: |
34118825 |
Appl. No.: |
10/861880 |
Filed: |
June 3, 2004 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60490324 |
Jul 25, 2003 |
|
|
|
Current U.S.
Class: |
600/300 ;
600/407 |
Current CPC
Class: |
G01S 7/52088 20130101;
A61B 8/14 20130101; A61B 8/483 20130101; A61B 8/5284 20130101; G06T
7/20 20130101; G06T 2207/30048 20130101; A61B 8/0891 20130101; G01S
7/52066 20130101; A61B 8/463 20130101; G06T 2207/10132 20130101;
A61B 8/481 20130101; A61B 5/7207 20130101; G01S 7/52036 20130101;
A61B 8/5276 20130101; A61B 8/543 20130101; G06T 2207/20104
20130101; A61B 5/352 20210101; G01S 15/8981 20130101; A61B 8/0883
20130101; G06T 7/0016 20130101 |
Class at
Publication: |
600/300 ;
600/407 |
International
Class: |
G06K 009/00; A61B
005/00 |
Claims
What is claimed is:
1. A method for automatically selecting an ultrasound image from a
set of ultrasound images, the method comprising: (a) providing a
plurality of sets of ultrasound images associated with a respective
plurality of heart cycles; (b) selecting a first ultrasound image
from a first set of ultrasound images associated with a first heart
cycle; and (c) automatically selecting an ultrasound image from
each of the other sets of ultrasound images by comparing at least
one visual characteristic of the first ultrasound image with at
least one visual characteristic of the ultrasound images in each of
the other sets.
2. The invention of claim 1, wherein (c) comprises determining
which ultrasound image in each of the other sets comprises at least
one visual characteristic that most-closely matches the at least
one visual characteristic of the first ultrasound image.
3. The invention of claim 1, wherein the ultrasound images are of a
heart, and wherein the at least one visual characteristic comprises
a shape of the heart.
4. The invention of claim 1, wherein the ultrasound images are of a
heart comprising a valve, and wherein the at least one visual
characteristic comprises a state of the valve.
5. The invention of claim 1, wherein the ultrasound images comprise
real-time images.
6. The invention of claim 1, wherein the ultrasound images comprise
contrast-enhanced ultrasound images.
7. The invention of claim 1 further comprising: (d) displaying the
ultrasound images automatically selected in (c).
8. The invention of claim 7, wherein (d) further comprises
displaying the ultrasound image selected in (b).
9. The invention of claim 1 further comprising: (d) performing a
quantification analysis on the ultrasound images automatically
selected in (c).
10. The invention of claim 9, wherein (d) comprises performing a
quantification analysis on the ultrasound images automatically
selected in (c) and the ultrasound image selected in (b).
11. The invention of claim 1 further comprising: (d) generating a
set of time intensity curves using the ultrasound images
automatically selected in (c).
12. The invention of claim 1 further comprising: (d) performing
motion correction on the ultrasound images automatically selected
in (c).
13. The invention of claim 12, wherein (d) comprises performing
motion correction on the ultrasound images automatically selected
in (c) and the ultrasound image selected in (b).
14. The invention of claim 1, wherein the first ultrasound image is
selected by a user from a plurality of displayed ultrasound
images.
15. The invention of claim 1, wherein the first ultrasound image is
automatically selected by identifying which image from the first
set is associated with a phase of the heart cycle indicated by a
stored user-preference.
16. The invention of claim 1 further comprising placing at least
one region of interest on at least one of the first image and an
automatically-selected image.
17. A method for motion correcting a set of ultrasound images
associated with a phase of a heart cycle, the method comprising:
(a) providing a plurality of sets of ultrasound images associated
with a respective plurality of heart cycles; (b) selecting a phase
of a heart cycle; (c) for each of the plurality of sets,
automatically selecting an ultrasound image that is associated with
the selected phase of the heart cycle; and (d) performing motion
correction on the automatically-selected ultrasound images.
18. The invention of claim 17, wherein the motion correction
comprises a sum of absolute differences (SAD) technique.
19. The invention of claim 17 further comprising placing a region
of interest on an image associated with the selected phase before
(d), and wherein (d) comprises changing the location of the region
of interest.
20. The invention of claim 17, wherein the motion correction
registers ultrasound images together.
21. The invention of claim 17, wherein the phase of the heart cycle
is selected in (b) by a stored user-preference.
22. The invention of claim 17, wherein the phase of the heart cycle
is selected in (b) by user selection of one of a plurality of
displayed ultrasound images, each of the displayed ultrasound
images being associated with a respective phase of the heart
cycle.
23. The invention of claim 22, wherein the user-selected image
comprises at least one visual characteristic, and wherein the
ultrasound images are automatically selected in (c) by comparing
the at least one visual characteristic of the user-selected
ultrasound image with at least one visual characteristic of the
ultrasound images in each of the plurality of sets.
24. The invention of claim 17, wherein the ultrasound images are
automatically selected in (c) by comparing (1) a time interval
between a reference phase in the heart cycle and a phase associated
with an ultrasound images in a set with (2) a time interval between
the reference phase and the selected phase.
25. The invention of claim 24, wherein the reference phase
comprises a preceding R-wave.
26. The invention of claim 24, wherein the reference phase
comprises a succeeding R-wave.
27. The invention of claim 24, wherein the reference phase
comprises a T-wave.
28. The invention of claim 17, wherein the ultrasound images
comprise real-time images.
29. The invention of claim 17, wherein the ultrasound images
comprise contrast-enhanced ultrasound images.
30. The invention of claim 17 further comprising displaying the
ultrasound images associated with the selected phase.
31. The invention of claim 30, wherein the displaying is performed
before the motion correction in (c).
32. The invention of claim 30, wherein the displaying is performed
after the motion correction in (c).
33. The invention of claim 17 further comprising: (e) performing a
quantification analysis on the motion-corrected ultrasound
images.
34. The invention of claim 17 further comprising: (e) generating a
set of time intensity curves using the motion-corrected ultrasound
images.
35. The invention of claim 17 further comprising placing at least
one region of interest on an automatically-selected ultrasound
image.
36. A method for motion correcting a set of ultrasound images
associated with a phase of a heart cycle, the method comprising:
(a) providing a plurality of ultrasound images associated with a
single phase of a heart cycle; and (b) performing motion correction
on the plurality of ultrasound images.
37. The invention of claim 36, wherein the plurality of ultrasound
images comprise gated images.
38. The invention of claim 36, wherein the plurality of ultrasound
images comprise real-time images.
39. The invention of claim 36, wherein the plurality of ultrasound
images comprise contrast-enhanced ultrasound images.
40. The invention of claim 36, wherein the plurality of ultrasound
images are manually selected by a user from a real-time image
clip.
41. The invention of claim 36, wherein the motion correction
comprises a sum of absolute differences (SAD) technique.
42. A method for automatically selecting ultrasound images
associated with a phase of a heart cycle, the method comprising:
(a) providing a plurality of sets of ultrasound images associated
with a respective plurality of heart cycles, wherein each
ultrasound image is characterized by a time interval between an
acquisition time of the ultrasound image and a reference phase in
the heart cycle; (b) selecting a phase of the heart cycle, wherein
the selected phase is characterized by a time interval between the
selected phase and the reference phase; and (c) automatically
selecting a plurality of ultrasound images from each heart cycle
that are characterized by a time interval that is within a
tolerance range of the time interval between the selected phase and
the reference phase.
43. The invention of claim 42, wherein the automatically-selected
plurality of ultrasound images are from the diastole portion of the
heart cycle.
44. The invention of claim 42, wherein the tolerance range varies
with the heart cycle.
45. The invention of claim 44, wherein the tolerance range is
greater for diastole than it is for systole.
46. The invention of claim 42, wherein the reference phase
comprises a preceding R-wave.
47. The invention of claim 42, wherein the reference phase
comprises a succeeding R-wave.
48. The invention of claim 42, wherein the reference phase
comprises a T-wave.
49. The invention of claim 42 further comprising; (d) motion
correcting the automatically-selected plurality of ultrasound
images.
50. The invention of claim 42 further comprising: (d) displaying
the automatically-selected plurality of ultrasound images.
51. The invention of claim 42 further comprising: (d) performing a
quantification analysis on the automatically-selected plurality of
ultrasound images.
52. The invention of claim 42 further comprising: (d) generating a
set of time intensity curves using the automatically-selected
plurality of ultrasound images.
53. The invention of claim 42, wherein the phase of the heart cycle
is selected in (b) by a stored user-preference.
54. The invention of claim 42, wherein the phase of the heart cycle
is selected in (b) by user selection of one of a plurality of
displayed ultrasound images, each of the displayed ultrasound
images being associated with a respective phase of the heart
cycle.
55. The invention of claim 42, wherein the ultrasound images
comprise real-time images.
56. The invention of claim 42, wherein the ultrasound images
comprise contrast-enhanced ultrasound images.
57. The invention of claim 42 further comprising placing at least
one region of interest on an automatically-selected ultrasound
image.
58. A method for automatically selecting ultrasound images
associated with a phase of a heart cycle, the method comprising:
(a) providing a plurality of sets of ultrasound images associated
with a respective plurality of heart cycles; (b) retrieving a
stored user-preference of a phase of the heart cycle; and (c) for
each of the plurality of sets of ultrasound images, automatically
selecting an ultrasound image that is associated with the phase
specified in the stored user-preference.
59. The invention of claim 58 further comprising; (d) performing
motion correction on the automatically-selected ultrasound
images.
60. The invention of claim 58 further comprising; (d) displaying
the automatically-selected ultrasound images.
61. The invention of claim 58 further comprising; (d) performing a
quantification analysis on the automatically-selected ultrasound
images.
62. The invention of claim 58 further comprising; (d) generating a
set of time intensity curves using the automatically-selected
ultrasound images.
63. The invention of claim 58, wherein the ultrasound images are
automatically selected in (c) by comparing at least one visual
characteristic of an ultrasound image associated with the phase
specified in the stored user preference with at least one visual
characteristic of the ultrasound images in each of the plurality of
sets.
64. The invention of claim 58, wherein the ultrasound images are
automatically selected in (c) by comparing (1) a time interval
between a reference phase in the heart cycle and a phase associated
with an ultrasound images in a set with (2) a time interval between
the reference phase and the phase specified in the stored user
preference.
65. The invention of claim 58, wherein the ultrasound images
comprise real-time images.
66. The invention of claim 58, wherein the ultrasound images
comprise contrast-enhanced ultrasound images.
67. The invention of claim 58 further comprising placing at least
one region of interest on an automatically-selected image.
68. A method for automatically selecting ultrasound images
associated with a phase of a heart cycle, the method comprising:
(a) providing a plurality of sets of ultrasound images associated
with a respective plurality of heart cycles; (b) selecting a phase
of a heart cycle; (c) for each of the plurality of sets,
automatically selecting an ultrasound image that is associated with
the selected phase of the heart cycle; and (d) placing at least one
user-defined region of interest on an ultrasound image associated
with the selected phase of the heart cycle.
69. The invention of claim 68, wherein the at least one
user-defined region of interest comprises more than nine
pixels.
70. The invention of claim 68, wherein the at least one
user-defined region of interest comprises less than an entire
imaged tissue.
71. The invention of claim 68, wherein the at least one
user-defined region of interest comprises less than an entire
myocardium.
72. The invention of claim 68 further comprising performing motion
correction on the automatically-selected ultrasound images.
73. The invention of claim 72, wherein the motion correction is
performed before (d).
74. The invention of claim 72, wherein the motion correction is
performed after (d).
75. The invention of claim 68, wherein the phase of the heart cycle
is selected in (b) by a stored user-preference.
76. The invention of claim 68, wherein the phase of the heart cycle
is selected in (b) by user selection of one of a plurality of
displayed ultrasound images, each of the displayed ultrasound
images being associated with a respective phase of the heart
cycle.
77. The invention of claim 76, wherein the user-selected image
comprises at least one visual characteristic, and wherein the
ultrasound images are automatically selected in (c) by comparing
the at least one visual characteristic of the user-selected image
with at least one visual characteristic of the ultrasound images in
each of the plurality of sets.
78. The invention of claim 68, wherein the ultrasound images are
automatically selected in (c) by comparing (1) a time interval
between a reference phase in the heart cycle and a phase associated
with an ultrasound images in a set with (2) a time interval between
the reference phase and the selected phase.
79. The invention of claim 78, wherein the reference phase
comprises a preceding R-wave.
80. The invention of claim 78, wherein the reference phase
comprises a succeeding R-wave.
81. The invention of claim 78, wherein the reference phase
comprises a T-wave.
82. The invention of claim 68, wherein the ultrasound images
comprise real-time images.
83. The invention of claim 68, wherein the ultrasound images
comprise contrast-enhanced ultrasound images.
84. The invention of claim 68 further comprising displaying the
automatically-selected ultrasound images.
85. The invention of claim 68 further comprising: (e) performing a
quantification analysis on the automatically-selected ultrasound
images.
86. The invention of claim 68 further comprising: (e) generating a
time-intensity curve for the at lease one region of interest.
87. A medical diagnostic ultrasound imaging system comprising: a
transducer; a beamformer in communication with the transducer; a
processor in communication with the beamformer; and a display
device in communication with the processor; wherein the processor
is operative to perform one or more of the methods in claims 1, 17,
36, 42, 58, or 68.
88. An ultrasound image review station comprising: a display
device; a user interface device; and a processor in communication
with the display device and the user interface device, wherein the
processor is operative to perform one or more of the methods in
claims 1, 17, 36, 42, 58, or 68.
Description
RELATED APPLICATION
[0001] This application claims the benefit of U.S. Provisional
Application No. 60/490,324, filed Jul. 25, 2003, which is hereby
incorporated by reference herein.
BACKGROUND
[0002] Contrast agents can be used in a medical ultrasound
examination to enhance diagnosis. The rate of contrast agent
enhancement of tissue relates to the rate of blood flow to the
tissue and can be used to diagnose a variety of disease states.
While contrast agent quantification can be performed over an entire
heart cycle, techniques have been discussed for performing
quantification on specific parts of the heart cycle. For example,
U.S. patent application Publication No. US2003/0114759A1 to Skyba
et al. describes an ultrasonic imaging system and method for
displaying tissue perfusion and other parameters varying with time
in which gated or ungated images are used to enable a parametric
display to be keyed to specific phases of the heart cycle.
Additionally, QLAB Software by Philips Medical Systems has been
described as containing custom tools to "auto trim" relative to an
ECG trigger for quantification of specific parts of a cardiac
cycle.
SUMMARY
[0003] The present invention is defined by the following claims,
and nothing in this section should be taken as a limitation on
those claims.
[0004] By way of introduction, the embodiments described below
relate to phase selection for cardiac contrast assessment. In one
embodiment, at least one visual characteristic from a selected
image is used to automatically select an ultrasound image from a
set of images from a plurality of heart cycles. In another
embodiment, motion correction is performed on ultrasound images
that are automatically selected from a plurality of ultrasound
images associated with the same phase of the heart cycle. In yet
another embodiment, ultrasound images are automatically selected
from a set of images based on a time interval that is within a
tolerance range from a reference phase of the heart cycle. In
another embodiment, a stored user-preference of a phase of the
heart cycle is used to automatically select an ultrasound image
from a set of images. In yet another embodiment, a user-defined
region of interest is placed on an ultrasound image associated with
a selected phase of a heart cycle. Other embodiments are provided,
and each of the embodiments described herein can be used alone or
in combination with one another.
[0005] The embodiments will now be described with reference to the
attached drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] FIG. 1 is a block diagram of a medical diagnostic ultrasound
imaging system of an embodiment.
[0007] FIG. 2 is a schematic of ultrasound images showing the time
course of contrast agent into a heart.
[0008] FIG. 3 shows time intensity curves of an embodiment.
[0009] FIG. 4 is a flow chart of a method of an embodiment.
[0010] FIG. 5 illustrates a plurality of sets of ultrasound images
of an embodiment.
[0011] FIG. 6 is a timeline showing an ECG trace and times when
image frames of an embodiment were acquired.
BRIEF DESCRIPTION OF THE PRESENTLY PREFERRED EMBODIMENTS
[0012] Introduction
[0013] Turning to the drawings, FIG. 1 is a block diagram of a
medical diagnostic ultrasound imaging system 100 that can be used
with the embodiments described herein, which generally relate to
diagnostic ultrasonic imaging with contrast agents. As shown in
FIG. 1, the ultrasound system 5 comprises a transducer probe 10, a
beamformer 15, a processor 20, a display device 25, an ECG device
30, and a user interface 35, each of which is in communication with
the other through one or more named or unnamed components. Two
components can be in communication with each other through a wired
or wireless connection. The term "processor" broadly refers to the
appropriate hardware and/or software components of the ultrasound
system 5 that can be used to implement the functionality described
herein. The ultrasound system 5 can comprise additional components,
which are not shown in FIG. 1 for simplicity.
[0014] During an ultrasound examination, a sonographer contacts the
transducer probe 10 with a patient, and the ultrasound system's
processor 20 causes the beamformer 15 to apply a voltage to the
transducer 10 to cause it to vibrate and emit an ultrasonic beam
into the portion of the patient's body in contact with the
transducer 10. Ultrasonic energy reflected from the patient's body
impinges on the transducer 10, and the resulting voltages created
by the transducer 10 are received by the beamformer 15. The
processor 20 processes the sensed voltages to create an ultrasound
image and displays the image on the display device 25. The ECG
device 30 captures ECG information (e.g., a heart cycle waveform)
from the patient. The ECG information can be displayed with an
ultrasound image during image acquisition. Additionally, the ECG
information can be stored with an ultrasound image on a storage
device internal or external to the ultrasound system 5 for later
review.
[0015] The ultrasound system 5 can be used for contrast agent
imaging to enhance diagnosis. Venous injection of contrast agent
causes an increase in the ultrasound signal when the contrast agent
washes into the tissue being imaged. In operation, a high-intensity
ultrasound pulse is transmitted to destroy contrast agent in a
tissue, and then the high-intensity ultrasound pulse is turned off
and the tissue is imaged as new contrast-agent-filled blood
perfuses into the tissue. The rate of the contrast agent
enhancement (or "wash-in") relates to the rate of blood flow to the
tissue and ultimately to tissue perfusion. Because of this, the
time course of contrast enhancement can be assessed to diagnose a
variety of disease states from coronary artery disease to tumor
neovasculature to liver metastases. As used herein, the phrase
"assessment of the time course of contrast enhancement" refers to
either qualitative or quantitative assessment of contrast
ultrasound images. All of these techniques rely on comparisons
between images acquired at different times in order to see the
variation caused by contrast agent enhancement. These techniques
will now be discussed.
[0016] Qualitative assessment is the simplest method of determining
the time course of contrast enhancement and relies on observing, by
eye, the visible changes in an ultrasound image caused by contrast
agent. Qualitative assessment had been done successfully for
coronary perfusion assessment (Cwajg et al., "Detection of
Angiographically Significant Coronary Artery Disease with
Accelerated Intermittent Imaging and Intravenous Administration of
Ultrasound Contrast Materials," American Heart J. 139: 675-683
(2000)) and characterization of liver diseases. Diagnoses can be
made by simply observing the contrast enhancement in a "live"
ultrasound image or by reviewing a stored image clip.
[0017] Quantitative assessment (or analysis) refers to time
intensity curve analysis, parametric imaging, displaying results of
calculations, or any other contrast enhancement assessment
technique (now existing or later developed) other than pure
qualitative analysis. Time intensity curve (TIC) analysis allows
more precise assessment of the time course of contrast enhancement
than can be made by eye alone. TIC's present the variation of the
average signal level or number of color pixels in a number of
regions of interest (ROI's) in graphical form to identify contrast
enhancement differences between the regions. The data is also fit
to functional curves, the parameters of which are used to make
absolute determinations of flow characteristics (Wei et al., "Basis
for Detection of Stenosis Using Venous Administration of
Microbubbles During Myocardial Contrast Echocardiography: Bolus or
Continuous Infusion," JACC 32: 252-60 (July 1998)). These
parameters can be related to coronary flow reserve (Wei et al.,
"Noninvasive Quantification of Coronary Blood Flow Reserve in
Humans Using Myocardial Contrast Echocardiography," Circulation
103: 2560-2565 (2001)) and can be used to make diagnoses.
[0018] FIGS. 2 and 3 illustrate the use of TICs. FIG. 2 is a
schematic of ultrasound images showing the time course of contrast
agent into a heart. The image at time 0 seconds was taken before
the start of contrast agent wash-in, and the images at times 1
second, 2 seconds, and 3 seconds were taken during contrast
wash-in. As shown in these images, the signal from tissue grows
over time, but enhancement is slower in the top part of the image,
indicating a problem in blood flow. Quantification is accomplished
by placing two regions of interest 40, 50 at different regions of
the heart image. Changes in the pixel intensity in the two regions
of interest 40, 50 is indicative of contrast wash-in in those
regions 40, 50, and the average intensity from within those regions
40, 50 is plotted in time-intensity curves (see FIG. 3). The
time-intensity curves show that the contrast enhancement in region
of interest 50 is slower than the contrast enhancement in region of
interest 40 by virtue of the reduced signal at times 1 second and 2
seconds. The data points in the time intensity curves are fitted
with smooth functions 60, 70, and parameters of these functions are
related to the blood flow in the tissue.
[0019] TIC analysis has had limited penetration into the market due
to the relatively time-consuming analysis required and the lack of
spatial information derived from a small number of ROI's. Removing
this obstacle while maintaining much of its sensitivity is
parametric imaging. Parametric imaging combines the spatial
resolution and ease of use of qualitative assessment with the
sensitivity of quantitative curve fitting. In parametric imaging,
the variation over time of signal level from each pixel is fitted
by some function that relates to a physiological phenomenon.
Selected parameters from the function for each pixel can then be
combined into an image to show variation in the entire series in a
static image. For contrast agent imaging, some parameter or
combination of parameters related to contrast wash-in is displayed
in this parametric image. In this way, spatial variation of the
wash-in time course can be easily assessed.
[0020] Motion Effects on the Assessment of the Time Course of
Contrast Enhancement
[0021] During a myocardial contrast imaging exam, cardiac
ultrasound images are acquired at a high frame rate (typically,
more than 15 frames per second) in order to provide wall motion
information for qualitative wall motion assessment as well as to
facilitate maintaining the scan plane. However, this high frame
rate provides little additional information for quantification
assessment, and cardiac motion shown by this high frame rate can
make quantification difficult. For example, cardiac motion can
cause substantial changes in intensity from different regions.
These cyclic changes are apparent in time intensity curves
generated from each frame in a real-time clip. It is also difficult
or impossible to maintain the same tissue in a ROI through the
cardiac cycle. Thickening causes the tissue in a fixed area to
change, and any rotation of the heart will also cause different
tissue to be in the image plane at different phases of the heart
cycle.
[0022] One solution to this problem is to select one image from
each heart cycle for analysis. (As described below, in some
situations, more than one image from each heart cycle can be
selected.) The selection of images from the same phase of the heart
cycle prevents contractile cardiac motion from interfering with
quantification of myocardial perfusion because subsequent analysis
(preferably including ROI placement) is based on the set of images
acquired at the selected phase instead of all of the acquired
images. (As used herein, "set" refers to a group of one or more
than one member.)
[0023] FIG. 4 is a flow chart of a method for phase selection for
cardiac contrast assessment of an embodiment. As shown in FIG. 4, a
plurality of sets of ultrasound images associated with a respective
plurality of heart cycles is provided (act 100). Next, a phase of
the heart cycle is selected (act 110), and ultrasound images
associated with the selected phase are automatically selected (act
120). Next, qualitative and/or quantitative assessment is
performed. With qualitative assessment, the automatically-selected
ultrasound images are displayed for user assessment (act 130). With
quantification assessment involving time intensity curves, one or
more regions of interest are placed on an ultrasound image
associated with the selected phase (act 150), and quantification
analysis is performed (act 160). As shown in FIG. 4, with
parametric imaging, no regions of interest are necessary (since
quantification is done on each pixel).
[0024] Motion correction (act 140) can be used with either
qualitative or quantitative assessment. It is important to note
that while FIG. 4 shows a method with multiple acts, each of these
acts can be used alone or in combination with one another.
Additionally, these acts can be performed on an ultrasound system
or an image review station (e.g., a dedicated workstation or a
personal computer with a processor, display device, and user
interface). These acts will now be discussed.
[0025] Providing a Plurality of Sets of Ultrasound Images
[0026] The first act of the method is to provide a plurality of
sets of ultrasound images associated with a respective plurality of
heart cycles (act 100). These images can be stored, for example, in
a storage device in an ultrasound system or image review station,
on removable media (e.g., a magneto-optical disc), or in a network
location accessible by the ultrasound system or image review
station.
[0027] With some of the embodiments below, it is preferred that the
stored images be acquired using "real-time imaging." As used
herein, the term "real-time imaging" refers to acquiring ultrasound
images with a sufficiently high frame rate such that enough
ultrasound images are acquired to allow the selection of an image
associated with a desired phase of the heart cycle. "Real-time
images" refers to images acquired during "real-time imaging."
"Real-time imaging" is different from gated imaging, in which a
portion of the heart cycle, such as an R-wave, triggers the
acquisition of a single ultrasound image per heart cycle. "Gated
images" refer to images acquired using gated imaging. Without
intending to set a lower limit to the definition of "real-time
imaging," an example of real-time imaging is acquiring ultrasound
images at a frame rate of more than 15 frames per second. It should
be noted that, as used herein, the term "real time" does not
indicate when images are displayed with respect to a time of
acquisition. Accordingly, images played from a stored clip can be
"real time" images even though they are being displayed at a time
after acquisition.
[0028] Selecting a Phase of the Heart Cycle
[0029] Next, a phase of the heart cycle is selected (act 110). This
act can be performed manually or automatically. FIG. 5 will be used
to illustrate the manual operation. FIG. 5 shows a display of four
sets 200, 210, 220, 230 of ultrasound images of an apical
four-chamber view of the heart over four heart cycles. Each of the
displayed images is associated with a particular phase of the heart
cycle, and selection one of the displayed images selects a phase in
the heart cycle. In operation, a user selects a displayed image to
represent the cardiac phase that he is interested in based on the
shape of the heart in the image or the state of the ECG signal
associated with that image. For example, selection of image A
selects a portion of the heart cycle that is X ms after the
preceding R-wave.
[0030] In the automatic operation, a user can set a phase
preference (e.g., X ms after the R-wave) that is stored in a
storage device of the ultrasound system. In future examinations,
the ultrasound system retrieves the stored user-preference to
automatically select the phase rather than requiring the user to
select a displayed image or take some other action to manually
select the desired phase. A combination of the automatic and manual
operations can be used. For example, the automatic operation can be
used to initially select the phase, and, if the user is
dissatisfied with this selection, the manual operation can be
used.
[0031] Automatically Selecting Ultrasound Images Associated with
the Selected Phase
[0032] For each of the plurality of sets of ultrasound images, an
image associated with the selected phase of the heart cycle is
automatically selected (act 120). This selection can be performed,
for example, based on a time interval from a part of the heart
cycle, such as the R-wave, or from image information. Each of these
techniques will now be described.
[0033] Cardiac ultrasound machines provide a subsystem that
measures physiological signals (such as the ECG signal) and present
the information along with the ultrasound images as a reference, as
shown in FIG. 5. The ultrasound system typically detects when the
R-wave occurs in each heartbeat. The times for each R-wave can be
stored along with the times of acquisition of each of the contrast
images. The difference between the timestamp for each image and the
time of the previous R-wave produces an interval that is analogous
to the cardiac phase. Cardiac phase can be defined in a number of
ways. The simplest method is to define only two phases of the
heart--systole and diastole. This is similar to dividing a sinusoid
into a negative phase and a positive phase. Sinusoidal phase can be
more specifically defined using a phase angle, and, in a similar
fashion, the phase of the cardiac cycle can be divided more finely.
Although the phase angle does not fit for the cardiac cycle as it
does for sinusoids, the phase can be defined as changing
continuously during a cycle where the heart is in the same state
during the same phase in all cycles. Irregularities can be
considered as variations in the rate of phase change during a
heartbeat or between heartbeats.
[0034] In operation, the time interval between a reference phase
(e.g., a preceding R-wave) and the selected phase is determined.
That time interval is then compared with the time interval between
the reference phase and the phase associated with each image in a
set of images from a heart cycle. An image in each cardiac cycle
can be selected that has an interval from the preceding R-wave that
most closely matches the interval for the image associated with the
selected phased. With reference to FIG. 6, a timeline 300 shows an
ECG trace and the times when image frames were acquired. Each
rectangle indicates the time for one ultrasound frame. When a user
selects image A, the ultrasound system automatically identifies
frames B, C, D, E, F, and G as having a similar phase as A based on
the ECG signal characteristics, such as the time interval from the
preceding R-wave or some other characteristic of the ECG signal. In
this way, the ECG information is used to sample an image clip at a
rate of one frame per heartbeat at the selected phase to retain
only images A-G for subsequent analysis.
[0035] If the imaging frame rate is irregular (possibly due to
contrast destruction), there may not be an image that was acquired
with a similar interval. In this case, the system preferably
identifies that the match is poor and does not provide any images
for some of the cardiac cycles. For example, a poor match can be
defined as when the best match differs by more than 100 ms.
Further, the above method relies on each cardiac cycle being
identical. When there are irregularities, the interval between an
image and the preceding R-wave may not give a good indication of
the phase of the heart in that image. The interval to the
succeeding R-wave can also be used to identify the cardiac phase in
an image in question, although it should be acknowledged that the
length of diastole varies to a much greater extent than does the
length of systole. Other characteristics of the ECG signal can also
be used to get a better estimate of the phase of a particular
image. For example, the T-wave can be identified in order to better
identify end-systole. This time can be used as another reference
when determining the phase of a particular frame.
[0036] Instead of using ECG information, an image can be
automatically selected based on one or more visual characteristics
of an image, such as the shape of the heart and/or the state of the
valves. In this way, visual characteristics are used to identify a
phase in the heart cycle. In operation, an image from a set of
images associated with a heart cycle is selected. This image can be
manually selected by a user or automatically selected by
identifying which image from the set is associated with a phase in
a stored user-preference. Then, at least one visual characteristic
from the selected image is compared to at least one visual
characteristic of the ultrasound images from the other sets. The
images that most-closely match (or match using some other criteria)
would be automatically selected. With reference again to the
illustration in FIG. 5, if the user selects image A, the ultrasound
system would compare image characteristics such as position of the
valve leaflets and/or the shape of the myocardium in image A with
corresponding visual characteristics in other images in the clip to
select the images that match most closely. In this example, A
closely matches images B, C, and D, so images A, B, C, and D are
used for subsequent analysis. (Alternatively, only the
automatically-selected images (images B, C, and D) can be used for
subsequent analysis and/or display.)
[0037] In the above discussion, only one frame per cardiac cycle
was selected for analysis. In some cases, however, the heart does
not change significantly over portions of the heart cycle. For
example, in diastole, there are periods where the heart is
comparatively still. Several neighboring images from these quiet
periods can be used in quantification (to provide more data points)
without the complication of heart motion. To do this based on the
time interval from the previous R-wave, the system would identify
all the images that had intervals within some tolerance range of
the interval for the user-selected image. Because the speed of the
heart changes dramatically during the heart cycle, this tolerance
range preferably varies based on the interval (e.g., longer for
intervals within diastole and shorter for intervals within
systole).
[0038] Qualitative Assessment
[0039] During qualitative assessment, the ultrasound images
associated with the desired phase are displayed (act 130). By
visually observing the images, the user can detect visible changes
caused by contrast agent to make a diagnosis.
[0040] Quantitative Assessment
[0041] As described above, quantitative assessment (or analysis)
refers to time intensity curve analysis, parametric imaging,
displaying results of calculations, or any other contrast
enhancement assessment technique (now existing or later developed)
other than pure qualitative analysis. Quantitative assessment often
relies on the placement of at least one region of interest.
Although the region(s) of interest can be placed on an image before
or after the automatic selection of images described above, placing
the ROI(s) on an automatically-selected image allows a user to
rapidly scroll through an image clip to check ROI placement in all
the relevant images because the data set has been reduced to images
at a particular cardiac phase. After the ROI(s) have been placed,
quantification can be performed. For example, time intensity curves
can be drawn for each phase of interest in the cardiac cycle. In
one presently preferred embodiment, at least one region of interest
used in quantification is user-defined and comprises more than nine
pixels but covers less than the entire imaged tissue (e.g., the
entire myocardium). As used herein, the term "pixel" refers to any
one of the small discrete basic elements from which an image is
composed.
[0042] Motion Correction
[0043] By selecting images from the same phase of the cardiac
cycle, contractile cardiac motion is removed before further
analysis. Contractile motion makes up a large part of the motion of
the heart, but patient breathing and inadvertent probe movement can
also cause motion of the heart in an ultrasound image. Preferably,
this motion is also tracked in order to effectively quantify
changes in the ultrasound signal received from the tissue. Many
techniques have been developed to automatically determine the
motion between pairs of images. While these techniques have been
used, motion correction in a real-time image clip can be very
difficult because of contractile cardiac motion. The phase
selection techniques described above simplify motion correction
because they allow motion tracking to be performed on a reduced set
of images. By first identifying the phase to align, motion
correction processing time can be reduced by a factor of 10-30 due
to the reduction in the number of frames processed.
[0044] The motion between pairs of images can be used to change the
location of the ROI's so that each ROI always encompasses the same
tissue. Alternatively, after the motion has been determined, the
images can be registered together so that the motion between the
images is removed. Simple registration can allow the entire image
to move as a unit to remove bulk motion. More complicated
techniques register different parts of the image independently and
stretch or shrink the intervening image in order to accommodate
relative movement in an image.
[0045] The most successful motion detection techniques use
correlation-based algorithms such as minimizing the sum of absolute
differences (SAD). The SAD technique is a good choice because it is
well-suited for rapid calculation. Using this technique, the
absolute difference is calculated between each point in the
original image and points in the subsequent image corresponding to
each of many potential motions. The best estimate of the actual
motion is chosen as the movement of the second image that produces
the minimum of the sum of the absolute differences. When ROI's are
tracked to eliminate motion, the SAD is preferably calculated
between each ROI in the original image and possible motions of that
ROI in the subsequent image. The minimum of the SAD for each ROI
will provide an estimate of the position of each ROI in the
subsequent image. When simple, bulk-movement registration is used,
the SAD is preferably calculated for a large region in the center
of the image. This will estimate the motion of the most important
part of the image in order to eliminate that motion through
registration. When more complicated registration techniques are
used that accommodate relative movement in the image, the SAD is
preferably calculated for many regions throughout the image. The
movement of each of these regions can then be estimated based on
the minimum of these SAD calculations and error checking. In this
technique, more weight is preferably placed on the displacement
estimates from some regions when calculating the displacement field
for the image.
[0046] There are several alternatives that can be used with these
embodiments. For example, while the embodiments described above
have been illustrated using contrast-enhanced ultrasound images,
these techniques can also be used with ultrasound images that are
free of contrast agent. Further, as noted above, these embodiments
can be used with real-time or gated images. For example, motion
correction can be performed on gated images or images manually
selected by a user from a real-time image clip instead of
automatically-selected images. Further, motion correction can take
place before or after region of interest placement and can be used
in conjunction with either qualitative or quantitative assessment.
Finally, as noted above, each of the embodiments described herein
can be used alone or in combination with one another.
[0047] It is intended that the foregoing detailed description be
understood as an illustration of selected forms that the invention
can take and not as a definition of the invention. It is only the
following claims, including all equivalents, that are intended to
define the scope of this invention.
* * * * *