U.S. patent application number 14/460043 was filed with the patent office on 2014-12-04 for system and method of ultrasound image processing.
The applicant listed for this patent is Echometrix, LLC. Invention is credited to Hirohito Kobayashi, Nathan D. Miller.
Application Number | 20140357996 14/460043 |
Document ID | / |
Family ID | 46020292 |
Filed Date | 2014-12-04 |
United States Patent
Application |
20140357996 |
Kind Code |
A1 |
Miller; Nathan D. ; et
al. |
December 4, 2014 |
SYSTEM AND METHOD OF ULTRASOUND IMAGE PROCESSING
Abstract
An ultrasound system includes an ultrasound transducer adapted
to obtain a dynamic series of echo signals of a subject tissue at
different deformation states, and an image processor for generating
and displaying ultrasound images of the tissue. The processor is
configured to generate dynamic images that correspond to the
dynamic series of echo signals, identify a plurality of pixels
within a region of interest (ROI) of a first of the generated
images, evaluate local tissue mechanical behavior by tracking the
displacement, deformation, and echo intensity of the identified
plurality of pixels from the first image to subsequent images based
on groups of pixels that correspond to each of the identified
plurality of pixels, determine tissue functionality in the subject
at the tracked pixel locations, and display the tissue
functionality in dynamic images that corresponds to the tracked
pixel locations.
Inventors: |
Miller; Nathan D.;
(Middleton, WI) ; Kobayashi; Hirohito; (Madison,
WI) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Echometrix, LLC |
Madison |
WI |
US |
|
|
Family ID: |
46020292 |
Appl. No.: |
14/460043 |
Filed: |
August 14, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13293499 |
Nov 10, 2011 |
8840555 |
|
|
14460043 |
|
|
|
|
61412071 |
Nov 10, 2010 |
|
|
|
Current U.S.
Class: |
600/438 |
Current CPC
Class: |
G06T 2207/10132
20130101; G06T 7/248 20170101; G06T 2207/30004 20130101; A61B
8/5223 20130101; G06T 7/0016 20130101; A61B 8/485 20130101; A61B
8/4254 20130101; G01S 7/52042 20130101; A61B 8/469 20130101 |
Class at
Publication: |
600/438 |
International
Class: |
A61B 8/08 20060101
A61B008/08; A61B 8/00 20060101 A61B008/00 |
Claims
1. A non-transitory computer readable storage medium having stored
thereon a computer program comprising instructions which when
executed by one or more computers cause the one or more computers
to: access a first image and a second image generated from a
dynamic series of echo signals of a tissue acquired using an
ultrasound transducer, wherein the second image represents the
tissue at a different state of deformation than the first image;
identify a location of a first plurality of pixels representing the
tissue in the first image; assume a location of a second plurality
of pixels representing the tissue in the second image; compare echo
signals of the first plurality of pixels to echo signals of the
second plurality of pixels; derive actual locations of the second
plurality of pixels based on the comparison; calculate a
deformation of the tissue between the first image and the second
image from the derived actual locations of the second plurality of
pixels; and output functionality information of the tissue based on
the deformation.
2. The non-transitory computer readable storage medium of claim 1
wherein the instructions cause the computer to compute the
functionality information of the tissue based on a comparison of
echo intensities between the first image and the second image with
respect to the deformation.
3. The non-transitory computer readable storage medium of claim 1
wherein the instructions cause the computer to overlay the
functionality information on at least one of the first image and
the second image.
4. The non-transitory computer readable storage medium of claim 1
wherein the first plurality of pixels comprises a first target
pixel and a plurality of pixels located in a neighbor-region of the
first target pixel; wherein the second plurality of pixels
comprises a second target pixel and a plurality of pixels located
in a neighbor-region of the second target pixel; and wherein the
instructions further cause the computer to: identify coordinates of
the first target pixel; assume coordinates of the second target
pixel; compare an echo intensity of the first target pixel to an
echo intensity of the second target pixel; and determine whether
the second target pixel represents the same tissue as the first
target pixel based on the comparison.
5. The non-transitory computer readable storage medium of claim 1
wherein the first image comprises a representation of the tissue in
a non-deformed state.
6. The non-transitory computer readable storage medium of claim 5
wherein the instructions further cause the computer to iteratively
adjust an assumed location of the second plurality of pixels if a
difference between the echo signals of the second plurality of
pixels and the echo signals of the first plurality of pixels
exceeds a threshold.
7. A method comprising: acquiring a dynamic series of echo signals
of a tissue using an ultrasound transducer; generating a first
image and a second image from the dynamic series of echo signals,
wherein the second image represents the tissue at a different state
of deformation than the first image; identifying a location of a
first plurality of pixels representing the tissue in the first
image; assuming a location of a second plurality of pixels
representing the tissue in the second image; comparing echo signals
of the first plurality of pixels to echo signals of the second
plurality of pixels; deriving actual locations of the second
plurality of pixels based on the comparison; calculating a
deformation of the tissue between the first image and the second
image from the derived actual locations of the second plurality of
pixels; and outputting functionality information of the tissue
based on the deformation.
8. The method of claim 7 further comprising computing the
functionality information of the tissue based on a comparison of
echo intensities between the first image and the second image with
respect to the deformation.
9. The method of claim 7 further comprising deriving the actual
locations of the second plurality of pixels by iteratively
adjusting an assumed location of the second plurality of
pixels.
10. The method of claim 7 further comprising: manually selecting a
first pixel of the first plurality of pixels in the first image;
and automatically selecting other pixels of the first plurality of
pixels, the other pixels comprising pixels located in a
neighbor-region of the first pixel.
11. The method of claim 7 further comprising simultaneously
monitoring a displacement of the tissue, a deformation of the
tissue, and an echo intensity of the first plurality of pixels.
12. The method of claim 7 wherein outputting functionality
information of the tissue comprises at least one of: generating a
composite image having the functionality information overlaid on an
image of the tissue; generating a histogram; and generating a
probability density function plot.
13. An ultrasound system comprising: an ultrasound transducer
adapted to obtain a dynamic series of echo signals of a tissue at
different deformation states; and one or more processors configured
to: access a series of images that correspond to the dynamic series
of echo signals; identify a target pixel within a first image of
the series of images, the target pixel representing tissue of a
patient; define a neighbor-region of pixels surrounding the target
pixel; simultaneously monitor a displacement of the tissue
represented within the target pixel, a deformation of tissue
represented within the neighbor-region of pixels, and an echo
intensity of pixels representing the tissue from the first image to
at least one subsequent image of the series of images; and generate
a functionality indicator for the tissue based on the monitored
displacement, echo intensity, and deformation.
14. The ultrasound system of claim 13 further comprising: a first
processor configured to generate the series of images that
correspond to the dynamic series of echo signals; and a second
processor configured to: access the series of images; identify the
target pixel; define the neighbor-region of pixels; monitor the
displacement, tissue deformation, and echo intensity; and generate
the functionality indicator.
15. The ultrasound system of claim 13 wherein the functionality
indicator comprises one of a histogram and a probability density
function plot.
16. The ultrasound system of claim 13 wherein the image processor
is further configured to: select a plurality of target pixels with
the first image of the series of images; define a plurality of
neighbor-regions of pixels, each of the plurality of
neighbor-regions of pixels surrounding a respective target pixel;
and monitor a displacement of tissue represented within the
plurality of target pixels, an echo intensity of the plurality of
target pixels, and a deformation of the plurality of
neighbor-regions of pixels from the first image to subsequent
images of the series of images.
17. The ultrasound system of claim 13 wherein the first image
comprises an image of the tissue in an undeformed state.
18. The ultrasound system of claim 13 wherein the image processor
is further configured to automatically define the neighbor-region
of pixels following selection of the target pixel.
19. The ultrasound system of claim 13 wherein the image processor
is further configured to generate a composite image comprising an
image of the tissue having the displayed functionality overlaid
thereon.
20. The ultrasound system of claim 19 wherein the displayed
functionality is displayed according to a tissue functionality
color map.
21. The ultrasound system of claim 13 wherein the image processor
is further configured to: identify coordinates of the target pixel
in the first image; estimate a displacement of the tissue
represented by the target pixel from the first image to a
subsequent image; identify coordinates of a displaced target pixel
in the subsequent image based on the estimated displacement;
compare an echo intensity of the target pixel in the first image to
an echo intensity of the displaced target pixel; and determine
whether the displaced target pixel represents the same tissue as
the target pixel based on the comparison.
22. The ultrasound system of claim 21 wherein the image processor
is configured to: use a mathematical norm to compare the echo
intensities; and if the mathematical norm exceeds a threshold,
estimate new coordinates of the target pixel and the
neighbor-region; and if the mathematical norm is less than the
threshold, use the estimated coordinates as the coordinates of the
target pixel and the neighbor-region in the subsequent image.
23. The ultrasound system of claim 21 wherein the processor is
further configured to iteratively estimate new coordinates of the
target pixel and the neighbor region.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] The present application is a continuation of and claims
priority to U.S. patent application Ser. No. 13/293,499 filed Nov.
10, 2011, which claims priority to U.S. Provisional Application
61/412,071 filed Nov. 10, 2010, the disclosures of which are
incorporated herein in their entirety.
BACKGROUND OF THE INVENTION
[0002] Embodiments of the invention relate generally to ultrasound
systems and methods for using ultrasound systems. More
particularly, embodiments of the invention relate to a system and
method for ultrasound image processing.
[0003] Conventional ultrasonic imaging provides a mapping of
ultrasonic echo signals onto an image plane where the intensity of
the echo, caused principally by relatively small differences in
material properties between adjacent material types, is mapped to
brightness of pixels on the image plane. While such images serve to
distinguish rough structure within the body, they provide limited
insight into the physical properties of the imaged materials.
Ultrasound-based diagnostic medical imaging techniques are used to
visualize muscles, tendons, and many internal organs, to capture
their size, structure and any pathological lesions with real time
tomographic images. Ultrasound has been used by healthcare
providers to image the human body for at least 50 years and has
become one of the most widely used diagnostic tools in modern
medicine. The technology is relatively inexpensive and portable,
especially when compared with other medical imaging modalities.
[0004] Ultrasound technologies can be used to visualize and discern
various medical information from soft tissues. However, the
mechanical material properties (i.e. load versus deformation) of
soft tissues such as tendon and ligament are nonlinear,
deformation-dependent, and can reflect the pathological state of
the tissue. Finding a non-invasive way of assessing these
properties is a difficult task, but ultrasound techniques and
systems can be utilized.
[0005] Today, most musculoskeletal pathologies are diagnosed by
observing images captured through modalities such as MRI or
Ultrasound. Often, key image texture changes affecting tissue
pathology are observed. Yet, this observation-based diagnosis is
highly subjective and observer-dependent. Hence an economical, yet
objective ultrasound assessment method or imaging technology has
been sought. To satisfy this clinical demand, additional ultrasound
imaging technologies have been developed.
[0006] It is well known that tissue mechanical functionality
(stiffness-strain relation and all other properties that can be
deduced from this relation) is a function of mechanical behavior
(deformation and displacement and all other properties that can be
deduced from deformation and displacement). The tissue mechanical
functionality is specific to each tissue type and tissue health
status. Hence, a properly and reliably evaluated tissue mechanical
functionality through a wide range of mechanical behavior can be a
reliable metric for diagnosis or monitor tissue health.
[0007] With the advancement of ultrasound technology, ultrasound
technology allows a fast, low cost, non-invasive and reliable
measurement of both tissue mechanical functionality and tissue
mechanical behavior.
[0008] However, known methods of tissue assessment with ultrasound
do not provide an objective measure of the status of a pathology.
Instead, typically an operator (such as a medical practitioner)
observes behavior of a tissue in dynamic ultrasound images and
makes a subjective determination as to the status.
[0009] It would therefore be desirable to have a system and method
capable of objectively determining a status of the pathology with
the tissue mechanical functionalities deduced via ultrasound
dynamic image (CINE image) analysis.
BRIEF DESCRIPTION OF THE INVENTION
[0010] Embodiments of the invention provide a system and method of
measuring behavior of a subject and providing an objective
assessment which can be used in a diagnosis.
[0011] In accordance with one aspect of the invention, an
ultrasound system includes an ultrasound transducer adapted to
obtain a dynamic series of echo signals of a subject tissue at
different deformation states, and an image processor for generating
and displaying ultrasound images of the tissue. The processor is
configured to generate dynamic images that correspond to the
dynamic series of echo signals, identify a plurality of pixels
within a region of interest (ROI) of a first of the generated
images, evaluate local tissue mechanical behavior by tracking the
displacement, deformation, and echo intensity of the identified
plurality of pixels from the first image to subsequent images based
on groups of pixels that correspond to each of the identified
plurality of pixels, determine tissue functionality in the subject
at the tracked pixel locations, and display the tissue
functionality in dynamic images that corresponds to the tracked
pixel locations.
[0012] In accordance with another aspect of the invention, a method
of determining a deformed state of a tissue in ultrasound images,
the method includes selecting pixels that are within a region of
interest (ROI) of a first ultrasound image of a tissue, wherein the
tissue is at a first state of deformation, identifying pixels that
surround the selected pixels in the first ultrasound image,
evaluating a local tissue mechanical behavior by tracking the
selected pixels from the first ultrasound image to subsequent
locations in subsequent ultrasound images using the identified
pixels that surround the selected pixels, wherein the subsequent
ultrasound images correspond to different states of tissue
deformation, determining functionality of the tissue at the
subsequent locations of the identified pixels, and displaying the
functionality at their original or subsequent locations in an image
of the tissue.
[0013] In accordance with yet another aspect of the invention, a
non-transitory computer readable storage medium having stored
thereon a computer program comprising instructions which when
executed by a computer cause the computer to obtain a dynamic
series of echo signals of a subject taken using an ultrasound
transducer with a tissue of the subject at different states of
deformation, generate first and second images using the obtained
series of echo signals, identify target pixels and their
neighboring pixels in the first image, assume locations of the
target pixels and their neighboring pixels in the second image,
calculate values of a mathematical norm based on the target pixels
and their neighboring pixels in the first image and based on their
assumed location in the second image, derive actual locations of at
least the target pixels in the second image based on the values of
the mathematical norm, and overlay deduced functionality or
mechanical behavior of the tissue that correspond to their actual
locations in the original or second image.
[0014] In accordance with still another aspect of the invention, a
method for monitoring tissue pathology includes generating a first
tissue functionality histogram/probability density function plot
based at least in part upon motion or deformation of a first pixel
between a first ultrasound image at a first deformed state and a
second ultrasound image at a second deformed state, calculating a
first ordinal scale aspect ratio based upon the first tissue
functionality histogram, generating a second tissue functionality
histogram/probability density function plot based at least in part
upon motion of a second pixel between a third ultrasound image at a
third deformed state and a fourth ultrasound image at a fourth
deformed state, calculating a second ordinal scale aspect ratio
based upon the second tissue functionality histogram/probability
density function plot, and generating a time series plot based on
the first ordinal scale aspect ratio and the second ordinal scale
aspect ratio as an indicator of a tissue pathology.
[0015] Various other features and advantages will be made apparent
from the following detailed description and the drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings illustrate embodiments presently contemplated
for carrying out the invention.
[0017] In the drawings:
[0018] FIG. 1 is a force-deformation plot illustrating behavior of
intact versus damaged tissue.
[0019] FIG. 2 is a stiffness plot as a function of material
deformation.
[0020] FIG. 3 is a schematic block diagram of an ultrasound scanner
suitable for use with the present invention.
[0021] FIG. 4 is a block diagram illustrating implementation of
region-of-interest (ROI) and fiber-of-interest (FOI) tracking,
according to embodiments of the invention.
[0022] FIG. 5 illustrates an ROI of a tissue showing boundary
pixels and their surrounding pixels used for motion and deformation
determination, according to an embodiment of the invention.
[0023] FIG. 6 is an algorithm for pixel tracking according to an
embodiment of the invention.
[0024] FIG. 7 is a graphical illustration of an original pixel and
its surrounding pixels, and the same pixel in a different location
and in a distorted state.
[0025] FIG. 8 is an illustration of a strain field color map ROI of
a tissue obtained according to an embodiment of the invention.
[0026] FIG. 9 is an illustration of stiffness-gradient fields
(color map) overlaid onto the ROI according to an embodiment of the
invention.
[0027] FIG. 10 is a histogram illustrating a height and width of
the histogram for evaluation tissue condition according to an
embodiment of the invention.
DETAILED DESCRIPTION
[0028] Terms used to describe aspects of the present invention are
given their ordinary and common meaning unless specifically defined
herein. For the purposes of the present application, stress is the
force distributed on a unit area of material, i.e., normalized
force. The stress causing stretching deformation is expressed as
positive stress. The stress causing contraction/compression
deformation is expressed as negative stress. Strain is normalized
deformation of material. Stretching deformation is expressed with
positive strain. Contraction/compression deformation is expressed
with negative strain. Stress-strain relation is a stress versus
strain plot.
[0029] Stiffness of a material is the slope of a stress versus
strain relationship. Because each different type of material has a
different stiffness, the slope of the stress-strain relationship
can be an important index for material identification. This slope
can be a constant for the simplest materials, but when the stress
versus strain relationship is linear these are called linear
materials. However, some materials such as soft biological tissues
are typically not necessarily linear, and as will be further
discussed, their linearity may vary as a function of whether the
tissue healthy or not. Such materials have a nonlinear
stress-strain relationship with a stiffness (slope of relationship)
that changes with increasing strain. Soft biological tissue (and
rubber-like materials as well) are usually less stiff (a lower
slope) at low levels of strain and stiffer (a higher slope) at
higher levels of strain. Generally, the nonlinear stress-strain
relation of a material is very specific to each material and
considered as a signature of each material.
[0030] A "stiffness gradient" is due to the nonlinear nature of the
stress-strain relation and a typical stiffness-strain relation
deduced from a stress-strain relation is also a relationship
(nonlinear) that changes with varying strain. The slope of this
stiffness-strain relation is thus named the stiffness gradient. A
set of parameters describing this curve can be considered as
indices of material stiffness caused by applied strain. Generally,
the stiffness gradient is also considered as a parameter that is
specific to each material.
[0031] Derivation of exemplary stiffness gradients for biological
tissues is illustrated in FIGS. 1 and 2. Referring to FIG. 1,
generic displacement versus load is shown for exemplary biological
tissues, such as tendons, that are intact (i.e., healthy) and
damaged (i.e., injured and not healed). In general a healthy
biological tissue will exhibit a different stiffness than that of
an injured tissue. Thus, load-displacement curve 10 includes
exemplary measured data for two measurements of intact or healthy
tissue 12 and for two measurements of damaged or injured tissue 14.
As known in the art, stiffness of a material is generally known as
the ratio of the force and deformation. As can be seen, at small
deformation 16, medium deformation 18, and large deformation 20,
damaged tissue 14 has a lower slope than that of injured tissue 14.
In this example, the injured tissue has a lower corresponding load
for a comparable displacement than does a healthy tissue. In other
words, for a given load, the healthy tissue will displace less than
will the damaged tissue. The general slope of measurements 12 and
14 is shown in FIG. 2 (thus, the healthy 12 and damaged 14 curves
are only illustrated as single respective curves). Taking the
derivative of the load curves thereby yields stiffness plot 22.
Stiffness plot 22 includes healthy tissue curve 24 and injured
tissue curve 26 which, as can be seen, are not linear. In this
example, the healthy tissue curve 24 has a larger slope 28 for
relatively smaller deformation than that 30 of injured tissue curve
26, and their slope difference decreases with increasing
deformation. It is the slope of these curves that can be taken
advantage of, according to the invention. Thus, as known in the
art, different magnitude of stiffness is discernible in the echoes
of ultrasound measurements.
[0032] Acoustic impedance is defined as the mathematical product of
density and acoustic velocity, or square root of the product
between density and stiffness. Ultrasound echo reflection is caused
at the interface between two materials with different acoustic
impedance. In the case of a biological tissue, the density of most
tissues is a similar value (close to the density of water), the
impedance difference of tissues are directly linked to stiffness
differences. However, even the stiffness differences of different
tissue types at a not-deformed state (not-loaded state) are minute,
hence echo reflection is small that results in an unclear
ultrasound image.
[0033] The acoustoelastic (AE) effect in elastic media occurs when
acoustic waves propagate through deformed elastic media, acoustic
characteristics (acoustic impedance, wave velocity, reflected echo
magnitude) depend on material properties and magnitude of applied
deformation. This phenomenon is called acoustoelasticity and the
theory of acoustoelasticity provides a set of equations to analyze.
Thus, when a biological tissue such as a tendon is flexed, the
flexing causes a change in the amount of tension, which can be
observed using ultrasonic images, according to the invention. Thus,
in principle and as known in the art, deforming a tendon while
receiving echoes from an ultrasound imaging apparatus can yield
ultrasound images that highlight changes in the stiffness of the
material that can be indicative of a level of injury in the tendon
by processing acquired ultrasound dynamic image (also widely known
as CINE image). Thus, such measurements can be employed to monitor
a tissue as it heals, and ultrasound images may be taken over a
period of weeks or months, as examples, to provide a means of
observing the status of the tissue.
[0034] However, the act of flexing the tendon causes motion of the
tendon to occur. Thus, in order to view an injured tendon the
tendon is, paradoxically, caused to move (by the patient flexing a
toe, for instance), which prevents a reference frame from being
maintained from one image to the next during a dynamic (cine)
ultrasound imaging session. As such, according to the invention, a
region-of-interest (ROI) is established which can be maintained
from image to image in a dynamic session, enabling both tissue
functionality such as stiffness and tissue mechanical behavior such
as tissue strain in an injured tendon, for instance, to be
monitored despite it being moved in order to cause deformation and
displacement of the tendon.
[0035] According to the invention, the ROI for analysis can be
selected by a user on an image frame of dynamic ultrasound images.
In one embodiment, a fiber-of-interest (FOI) for analysis can be
selected by a user on an image frame of dynamic ultrasound images.
According to one embodiment, the ROI and FOI can be selected on a
first image (image of tissue resting state without deformation) of
the dynamic ultrasound images and in another embodiment the ROI or
FOI are selected on other than the first image frame of ultrasound
dynamic images.
[0036] As implied, deformation-dependent stiffness is directly
related to the echo intensity. Hence the deformation-dependent
stiffness is calculated by assuming the echo intensity at any
region to be "stiffness" that is modeled as the linear combination
of finite numbers of constants and the same strain evaluated at the
pixel. Deformation-dependent stress is a deformation-dependent
function that is evaluated by integrating "deformation-dependent
stiffness."
[0037] Pattern recognition aims to classify data based either on a
priori knowledge or on statistical information extracted from
patterns. The patterns to be classified are usually groups of
measurements or observations, defining points in an appropriate
multidimensional space. The method and apparatus described herein
can be directly implemented onto ultrasound hardware system for on
site processing or delivered to any computer system as image
post-processing software.
[0038] Referring to FIG. 3, an acoustoelastographic imaging system
50 suitable for use with the present invention may employ an
ultrasonic imaging machine 52 alone or in combination with an
external computer 54. According to one embodiment, computer 54
includes more than one core or CPU 31 for performing calculations,
enabling quicker analysis of a region-of-interest, as will be
further described. Generally, the ultrasonic imaging machine 52
provides the necessary hardware and/or software to collect and
process ultrasonic echo signals by a processor 56 held within the
ultrasonic imaging machine 52 or in the external computer 54.
[0039] An ultrasonic transducer 58 associated with the ultrasonic
imaging machine 52 may transmit an ultrasonic beam 60 toward a
region of interest 62 within the patient 64 on a table 65 to
produce echo signals 66 that may be received by the ultrasonic
transducer 58 and converted to an electrical echo signals 66. For
the construction of an image, multiple rays within ultrasonic beam
60 and echo signals 66 will be acquired through different voxels 68
in the patient so as to obtain an "echo set" of echo signals 66
from a plurality of voxels 68 within the region of interest 62.
[0040] The electrical echo signals 66 may be received by interface
circuitry 70 of the ultrasonic imaging machine 52, the interface
circuitry 70 providing amplification, digitization, and other
signal processing as is understood in the art. The digitized echo
signals may then be transmitted to a memory 72 for storage and
subsequent processing by the processor 56 as will be described. The
processed echo signals 66 may be used to construct an image
displayed on graphical display 74. Input commands from an operator
may be received via a keyboard 76 or cursor control device 78, such
as a mouse, attached to the processor 56 via an interface 80 as is
well understood in the art. A position sensor 82 may be attached to
the ultrasonic transducer 58 to indicate orientation of the
ultrasonic transducer 58 through an electrical position signal 84
also provided to the interface circuitry 70.
[0041] At least one embodiment of the method includes an algorithm
that evaluates material properties of user selected ROI and/or FOI
of deforming medium by analyzing captured ultrasound dynamic
images. Additionally, as will be described, data input/output is a
feature that manages image upload and recording the deduced
composite data image. The ROI analysis includes ROI selection,
tracking, deformation evaluation, ROI echo intensity monitor and
material property evaluation will be achieved in this feature. The
deduced ROI data can then be visualized. In another embodiment, FOI
tracking is provided. FOI selection, tracking and deformation
evaluation will be achieved in a similar methodology as ROI
tracking, with the exception that FOI tracking is linear. The
deduced FOI data will be also display in this feature.
[0042] Referring now to FIG. 4, the illustrated method 100 consists
of three key parts: data I/O 102, ROI analysis 104 and FOI analysis
106. The data I/O elements 108 of method 100 include ultrasound
dynamic images 110 widely known as CINE images of the deforming
target material has to be properly captured by a user with an
ultrasound device such as that described with respect to FIG. 3.
This method can analyze CINE images formatted with most image
format including DICOM format that is widely accepted as a standard
medical image saving format. Properly captured images can be
uploaded 112 into a data key pad for single image analysis or
multiple images analysis for data compare. Both uploaded original
images and analyzed images can be transferred to movie re-player
and recorder 114 for viewing and recording moving dynamic pictures.
When the analyzed the images are viewed at 114, multiple deduced
data can be overlaid onto an original image and presented as a
composite image using either ROI tracking 104 or FOI tracking 106.
The newly created composite data image can be save as any image
format to be viewed by other software. Saved composite data images
can be uploaded into data I/O block 108 for re-analysis on a
different ROI/FOI. Uploaded images can be transferred to ROI
analysis block for analysis. To achieve swift and proper AE
analysis, a proper ROI is selected by a user with ROI selection
function 116.
[0043] ROI selection can be achieved by a mouse click action on
multiple pixels (including more than three pixels to properly
define a region in an image). Once the ROI is selected and defined
by the user 116, ROI inner pixels to be analyzed will be
automatically defined by software. The number and location of ROI
inner pixels will be automatically determined based on a default or
user selected pixel spacing (density) of inner pixels. The user can
select multiple ROIs for simultaneous analysis and comparison.
According to the invention and as will be described, a small
neighbor pixels surrounding the defined pixel will be automatically
defined by an analysis algorithm. The shape of and size of neighbor
region can be set default or selected by user.
[0044] Once the ROIs are selected by a user, the proposed method
and system can automatically initiate ROI tracking feature 118. ROI
tracking will be described further with respect to FIGS. 5 and 6.
In the disclosed process, three key parameters, pixel displacement,
pixel echo intensity and deformation of neighbor region surrounding
each defined pixels (ROI border/inner pixels) are simultaneously
evaluated and monitored. The pixel displacement monitor and
evaluation of neighbor region deformation are simultaneously
achieved with the tracking algorithm described below.
[0045] Referring now to FIG. 5, ROI 200 is defined by, in this
illustration, four corner pixels 202. ROI 200 in this embodiment is
of a portion of a tendon 204. It is contemplated that, although ROI
200 is illustrated having four pixels 202 that form boundary 206,
ROI 200 may be formed having three or more pixels 202 which form an
area or ROI 200 therebetween.
[0046] According to the invention, ROI tracking 118 of method 100
includes tracking motion of the ROI by using pixels from the
dynamic ultrasound images that surround 208 pixels 202. Thus,
referring to FIG. 6, ROI tracking 300 starts 302 and pixels are
identified 304 to track motion of the ROI. Such pixels may include
but are not limited to pixels 202 of portion of a tendon 204 in
FIG. 5. Once pixels for determining displacement of the ROI are
determined at step 304, pixels surrounding the identified pixels
are identified at step 306.
[0047] Pixel displacement and pixel neighbor-region deformation may
be calculated by optimizing properly defined mathematical norm,
according to the invention. Two example mathematical norms are
presented herein. In each of these two mathematical norms,
displacement of a pixel from one dynamic image to the next is
calculated by assuming a displacement and deformation of an element
308, calculating the value of the norm 310, and comparing it to a
threshold 312. If the value of the norm is not within a
predetermined threshold 314, then the assumed displacement and
deformation is perturbed or otherwise altered 316, and the norm is
recalculated at step 310. The process iteratively repeats until,
when the value of the norm is within the given threshold 318, then
the process ends 320. The process of pixel displacement and
neighbor-region deformation calculation just described can be done
on a number of pixels within the ROI. Thus, dynamic images obtained
by, for instance, flexing a tendon, can be evaluated in order to
track the movement or displacement of pixels that define the ROI
and pixels within the ROI from image to image.
[0048] As stated, pixel displacement may be calculated with any one
of many mathematical norms. Two example norms are described here
after. Referring to FIG. 7, an X-Y plot 400 includes an original
element F 402 having pixel "22" 404 at its center. Pixel "22" 404
corresponds to a selected pixel to determine its motion from image
to image, such as one of pixels 202 of FIG. 5. Further, as stated
with respect to FIG. 5, surrounding pixels 208 may include a radius
of pixels that may include 10, 15, or more pixels in radius from
pixel "22" 404. For simplicity of illustration purposes, however,
only 1 pixel surrounding pixel "22" 404 is illustrated. Such are
labeled in FIG. 7 as surrounding pixels 406 (elements 11, 12, 13,
21, 23, 31, 32, and 33). X-Y plot 400 also illustrates deformed and
displaced element G 408 having center pixel "22*" 410 with
surrounding pixels 412 (elements 11*, 12*, 13*, 21*, 23*, 31*, 32*,
and 33*).
[0049] By example, assume a original pixel (pixel 22 404) is
displaced u and v to new location 22* 410. Along with pixel
movement, the original square shaped surround neighbor of element F
402 has deformed and translated to be G(x*,y*) 408. It is also
assumed that the magnitude of both displacement and deformation
take place are relatively small and can be assumed to be bounded by
a small search boundary. In this circumstance, the coordinate
relation between each pixel in original and deformed neighbor
region for general three dimensional case be related by the
following relation:
{right arrow over (x)}*={right arrow over (x)}+{right arrow over
(u)}+(.gradient..sub.x{right arrow over (u)}).DELTA.{right arrow
over (x)}; Eqn.0.
[0050] Here {right arrow over (x)}, and .DELTA.{right arrow over
(x)}, represent known coordinates vector of center pixel and known
relative distances vector of other pixels from center pixel in the
original neighbor region. The unknown displacements vectors of
center pixel are represented by {right arrow over (u)}. However, to
make the discussion simple, the two dimensional case will be used
here after. In the two dimensional case the coordinate relation
between each pixel in original and deformed neighbor region can be
related by
x * = x + u + .differential. u .differential. x .DELTA. x +
.differential. u .differential. y .DELTA. y ; and Eqn . 1 ; y * = y
+ v + .differential. v .differential. x .DELTA. x + .differential.
v .differential. y .DELTA. y ; Eqn . 2. ##EQU00001##
[0051] Here x, y, .DELTA.x, and .DELTA.y represent known x, y
coordinates of center pixel 404 and known relative distances of
other pixels from center pixel 404 in the original neighbor region
402. Unknown displacements of center pixel 404 are represented by u
and v. Similarly, the unknown deformation and rotation of
neighbor-region (element) are represented by the differentials
.differential. u .differential. x , .differential. u .differential.
y , .differential. v .differential. x , and .differential. v
.differential. y . ##EQU00002##
[0052] Mathematically defined norms for assessing the deformation
of center pixel 404 include evaluation of the cross-correlation
coefficient r.sub.ij:
r ij ( u , v , .differential. u .differential. x , .differential. u
.differential. y , .differential. v .differential. x ,
.differential. v .differential. y ) = 1 - ( i j [ F ( x i , y j ) -
F _ ] [ G ( x i * , y j * ) - G _ ] ) i j [ F ( x i , y j ) - F _ ]
2 i j [ G ( x i * , y j * ) - G _ ] 2 ; Eqn . 3. ##EQU00003##
[0053] or a norm of pixel intensity difference:
[0053]
s(i,j)=.SIGMA..sub.i.SIGMA..sub.j[F(x.sub.i,y.sub.j)-G(x*.sub.i,y-
*.sub.j)].sup.2; Eqn. 4.
[0054] Here F(x.sub.i,y.sub.j) is the gray-scale value at a point
(x.sub.i,y.sub.j) in the original image and G(x.sub.i*,y.sub.j*) is
the gray-scale value at a point (x.sub.i*,y.sub.j*) in the deformed
image. The mean values of the sub-image F and G are denoted by F
and G.
[0055] With these mathematical relations, the step by step process
for tracking center pixel by finding best match of neighbor-region
can be described as follows, as described also with respect to FIG.
6:
[0056] Step 1: Assume initial displacement u, v and deformation
du/dx, du/dy, dv/dx and, dv/dy and estimate new pixel coordinate
x*, y* and neighbor region G(x* and y*) by equations 1 and 2.
[0057] Step 2: Evaluate mathematically defined norms (either Eqn. 3
or 4).
[0058] Step 3: Check the magnitude of the evaluated norm. If the
value is below a defined threshold and/or change in u, v and
deformation du/dx, du/dy, dv/dx and, dv/dy falls below a defined
threshold, the new location of the neighbor region is considered
found and terminate tracking.
[0059] Step 4: If the evaluated norm is relatively large, then
perturb displacement u, v and deformation du/dx, du/dy, dv/dx and,
dv/dy slightly and repeat process step 1 through 3 until properly
small value of norm is achieved.
[0060] The disclosed process can be executed with any existing
nonlinear optimization mathematical theory.
[0061] Upon the completion of the evaluation of local tissue
mechanical behavior that includes pixel displacement and tracking,
neighbor-region deformation evaluation and echo intensity change
monitoring, this algorithm will evaluate the tissue functionality
with deduced tissue mechanical behavior information. Here, the
normalized material stiffness and deformation dependent normalized
stiffness (stiffness gradient) are deduced from recoded pixel echo
intensity and material deformation.
[0062] Since the deduced pixel displacement and pixel
neighbor-region deformation are evaluated numerical value of the
displacement and deformations of all pixels within ROI can be
transferred onto subsequent images. Thus, returning to FIG. 4,
after the above description related to ROI tracking (that includes
mechanical behavior such as tissue movement and deformation), at
step 120 tissue functionality or property evaluation is performed,
and ROI property presentation occurs at step 122 where ROI
locations and deformations are overlaid with tissue functionality
such as a stiffness-strain relationship.
[0063] Referring still to FIG. 4, in a FOI assessment a
practitioner selects typically two points in one of a series of
dynamic images at step 124. The movement of two points are tracked
at step 126 in much the same fashion as described above with
respect to FIGS. 5-7. At step 128 a property of the tissue is
evaluated. However, because only two points were selected, the
property that is being assessed is limited to relative motion of
the two points. In such fashion, a practitioner can determine,
based on the evaluation, whether the tissue is undergoing
deformation in the dynamic images and at the selected location. The
property is presented at step 130 to the practitioner. Thus, this
method also evaluates the regional material stiffness-strain for
any pixel neighbor region defined at each pixel in ROI as part of
ROI material property evaluation feature. Echo intensity is
directly related to the stiffness. Hence the concept of the
deformation-dependent stiffness is developed by assuming the echo
intensity at any region to be indicative of stiffness-strain that
is modeled as the linear combination of finite numbers of constants
and same numbers of strain evaluated at the pixel.
[0064] This algorithm models the deformation-dependent stiffness in
any sub-region in an ROI at arbitral image frame number, say N
frame, as a linear combination of m numbers (picked by user) of
unknown constant a.sub.i and m numbers of strains that are
evaluated at the same point/region at N numbers of image frames
that included current image frame (m, m-1, m-2, N-m+1).
C.sub.N(.epsilon..sub.N)=a.sub.1.epsilon..sub.N+a.sub.2.epsilon..sub.N-1-
+ . . . +a.sub.m.epsilon..sub.N+m-1
[0065] Unknown constants a.sub.i can be deduced by solving m
numbers of linear equations with measured N sets of stiffness
C.sub.N (.epsilon..sub.N) and strains evaluated at the same
neighbor sub-region. Deduce unknown constant a.sub.i can be
feedback into above relation to describe deformation dependent
material stiffness.
[0066] Finally, this algorithm can evaluate deformation-dependent
normalized stress for any sub-regions by integrating
deformation-dependent stiffness deduced in previous step.
[0067] Once the tissue functionality and mechanical behavior are
evaluated in ROI property evaluation feature, all data are passed
to an ROI presentation feature for qualitative and quantitative
data presentation.
[0068] First, the deduced numerical value of the tissue
functionality and mechanical behavior of all pixels within ROI can
be translated into a color scheme and a color image and be
produced. In addition, the same deduced numerical values of all
pixels within ROI can be used to evaluate the probability density
function and/or histogram and presented as a plot.
[0069] The algorithm described in this document can be directly
implemented onto ultrasound hardware system for on site processing
or delivered to any computer system as image post-processing
software. The overview architecture of this algorithm is presented
in FIG. 1.
[0070] As Numerical Data Output, all of the deduced the data can be
saved with a simple viewable format such as ASCII format. ASCII
format data can be easily accessed with any existing software that
is available on market.
[0071] According to at least one embodiment, a fully automated
tracking and ultrasound analysis is a solution for practical daily
swift clinical use. To achieve that goal, currently used manual ROI
selection may be replaced by an automated or systematic ROI
selection. Thus, a more automated ROI selection may be achieved by
one of following methods.
[0072] Method 1: Implementation of automated tendon/ligament
identifier: The image texture of tendon/ligament is different from
other type of tissues, such as fat, muscle and skin, hence
automated tissue "differentiation", also known as "segmentation",
is possible by implementing an existing digital image
processing/machine vision algorithm.
[0073] Method 2: Tracking of all pixels in view space: Treating
whole image as ROI and track all key pixels. Tracking all pixels
can be time consuming. However, by splitting the tracking procedure
into multiple cores, such as cores 31 of computer 30 of FIG. 3.
Hence tracking all pixels can be accomplished.
[0074] Alternatively, the ROI tracking can be performed by
utilizing an optical flow methodology. The direct output of the
optical flow method is the velocity (speed in x and y direction) of
a target pixel. There are two additional steps that are used to
evaluate the deformation (stretch=Post motion length between two
pixels-pre-motion length of two pixels) necessary for evaluation.
First, the "time" factor has to be multiplied to the deduced
velocity to find out the displacement of the pixel. Second, the
deformation can be evaluated from pixel displacement.
[0075] In yet another alternative example, the ROI tracking can be
performed utilizing a region-matching, or box matching,
methodology. This method contains some similarities with the first
ROI tracking methodology described above. First, a small sub-region
is created around a target pixel to be tracked. The best matching
sub-region through matching texture with optimization is then
found.
[0076] In another embodiment, a method for monitoring tissue
pathology is provided. In the deduced tissue functionality color
map, each pixel contains tissue functionality data. A tissue
functionality histogram/probability density function plot generated
from the deduced tissue functionality. The plot can be plotted with
the Frequency (frequency of pixels) on the y-axis and the deduced
tissue functionality on the x-axis. The plot is then normalized to
reduce the possible bias caused by differences in the region of
interest. One method of normalization takes all frequency
(frequency of pixels) normalized by the maximum frequency value of
the deduced from other ultrasound data obtained for a particular
subject from a particular data set. The aspect ratio for a
particular normalized histogram is calculated. The height of the
histogram (normalized frequency) is divided by the width of the
histogram (range of tissue functionality and mechanical behavior
distribution) to calculate the aspect ratio. The aspect ratio is
converted and utilized as an ordinal scale to differentiate tissue
pathology differences over time. To assist with this
differentiation, a time series plot is generated for the normalized
aspect ratios obtained for a particular subject at different times.
The time series plot can graphically provide data for monitoring
tissue pathology changes. Based upon the time series plot certain
information about the monitored tissue can be deduced.
[0077] By example, a patient presented with a soft tissue injury
can be monitored over time. At selected points, the injured tissue
can be visualized through ultrasound and the tissue functionality
and mechanical behavior histograms generated. The normalized aspect
ratio can be plotted and provide an ongoing ability to monitor the
injured tissue. Furthermore, a single patient can be compared to
existing patients data for similar injuries and determine if the
injury is healing at an expected rate, slow rate, or at an
accelerated rate. Referring to FIGS. 8-10, subsequent steps of at
least one embodiment of the aspect ratio analysis process are
illustrated. Referring to FIG. 8, a mechanical behavior field is
illustrated (typically done in color for shown here in black/white)
which is a result of the ROI analysis in one of the images that is
subsequent to the image of FIG. 5. Referring to FIG. 8, ROI 200 of
FIG. 5 has moved from its location in the image (from where it was
in FIG. 5) and has clearly distorted in shape as well. Referring to
FIG. 9, the tissue functionality fields are obtained using methods
known in the art. That is, relationships between elements can be
established using known techniques and using measured echoes at the
subsequent images.
NC.sub.N(.epsilon..sub.N)=A.sub.1.epsilon..sub.1+A.sub.2.epsilon..sub.2+
. . . +A.sub.N.epsilon..sub.N
[0078] In the above equation, tissue functionality NC(.epsilon.),
measured strains .epsilon., and relations are established from
frame to frame via the variables A. Tissue functionality
NC(.epsilon.) and strain .epsilon. are measured as a function of
time. With the above relation, the time factor can be struck and
simple tissue functionality relations may be evaluated. Once the
tissue functionality is established, a tissue functionality
gradient can be evaluated using the relationship as follows:
SG 2 - 1 = NC ( 2 ) - NC ( 1 ) 2 - 1 ##EQU00004##
[0079] As such, the tissue functionality gradient can be calculated
and is specific to tissue type or tissue health (and therefore
varies over time as a tissue heals, for instance). Thus, tissue
functionality fields are calculated and illustrated in FIG. 9.
Referring now to FIG. 10, the calculated tissue functionality
fields from FIG. 9 can be presented in histogram form 400 having
tissue functionality level along the x-axis and number of
occurrences as counts in the y-axis. The illustrated curves of FIG.
10 include an injured tissue 402 corresponding to pixel stresses
obtained from FIG. 9. However, as the tissue heals over time, the
tissue will be illustrated having a more uniform color and lower
stress levels. Thus, histogram 404 illustrates a tissue that is
significantly healthier than that illustrated in histogram 402.
That is, histogram 402 is shallower and wider (having pixels with
higher stress) when compared to healthy tissue of histogram 404.
Thus, the illustrated histograms provide an objective measure of
health that can be simply evaluated by obtaining a ratio of the
height 406 of the histogram 404 versus its width 408. Further, the
cutoff for determining width 408 can be obtained based on a number
of counts (roughly 4 in FIG. 8), but can be any number based on
imaging circumstances, pixel size, etc. . . . . And, it is to be
recognized that the ratio of height 406/width 408 will increase
with time as the tissue heals. That is, less pixels will tend to
have a high stress level and more will lean toward an increased
mean value. The ratio can be ascertained over time, according to
the invention, as a means to objectively determine the health of a
tissue from dynamic images obtained while flexing the tissue, and
over a period of weeks or months.
[0080] In another embodiment, there is implementation of automated
tendon/ligament identifier: The image texture of tendon/ligament is
different from other type of tissues, such as fat, muscle and skin,
hence automated tissue "differentiation", also known as
"segmentation", may be possible by implementing existing digital
image processing/machine vision algorithm.
[0081] In another embodiment, tracking of all pixels in view space:
Treating whole image as ROI and track all key pixels. Tracking all
pixels can be time consuming. However, by splitting the tracking
procedure into multiple cores, the calculation speed can be
drastically reduced.
[0082] Furthermore, the ROI data points, including the neighboring
points included within the algorithm, the distance between these
points (.DELTA.x, .DELTA.y) be modified. in current code, default
pixel distance is set at 10 pixels. However, a user can key in any
pixel spacing. Generally, smaller pixel spacing results in a higher
analysis resolution. On the other hand, larger pixel spacing
results on a lower analysis resolution.
[0083] The following scenario provides an example: A patient is
properly scanned with any ultrasound system. The dynamic (video)
ultrasound images (CINE mode) of gradually deforming tissue are
captured. Captured dynamic (video) ultrasound images are sent to
PAC (Picture Archive Center) as part of DICOM (Digital Imaging and
Communications in Medicine) format patient's information. In image
diagnosis, radiologist retrieves patient's DICOM data, selects the
ROI (or perhaps first performs one or more FOI determinations) and
applies an embodiment of the present invention. Then, a Radiologist
more accurately diagnoses pathology with deduced data from
supplemental information.
[0084] Another example includes: A is patient properly scanned with
any ultrasound system. The dynamic (video) ultrasound images (CINE
mode) of gradually deforming tissue are captured. B1) Captured
dynamic (video) ultrasound images sent to computer connected to
ultrasound system. B2) ROI selected and an embodiment of the
present invention for ROI tracking is applied on-site in the
computer connected to ultrasound system. The deduced results are
put into DICOM format patient file and sent to PAC. D)
Diagnosis.
[0085] In yet another example, the ROI tracking method is employed:
A) The patient properly scanned with any ultrasound system. The
dynamic (video) ultrasound images (CINE mode) of gradually
deforming tissue are captured. In the captured dynamic (video)
ultrasound images, the ROIs are selected and analyzed with the ROI
tracking methods programmed into ultrasound system. The deduced
results are put into DICOM format patient file and sent to PAC. D)
Diagnosis.
[0086] Currently ROI is selected by applying clicking action on the
pixels that defines ROI Alternatively, a user can select ROI by
encircling ROI with mouse movement. For example, the pencil feature
in Microsoft paint program. Similarly, defining ROI with encircling
ROI with tablet pen on touch screen can be an option.
[0087] The pixel tracking can be achieved by actually tracking the
texture of sub-regions (currently circles) surrounding each target
pixel. The size of disk-shaped sub-region can influence the
tracking result. Larger sub-regions track better. Currently, the
radius of sub-region disk is set at 20 pixels, but this can range
significantly depending upon system user's selection. The size of
the sub-region can be optimized. A differently shaped sub-region
(square or rectangular) may be chosen alternatively.
[0088] Some of the major concept elements behind the ROI tracking
methods are the evaluation of deformation-dependent tissue
functionality named "stiffness gradient" by measuring the echo
intensity change within the ROI. Indeed the echo intensity within
ROI and sub-region (around target pixel)) change from
frame-to-frame. However the magnitude change of echo intensity or
texture (intensity of group of pixels) is ASSUMED to be small if
TWO consecutive images are compared. Hence, tracking a single pixel
can be properly achieved by tracking the sub-region surrounding the
target pixel. The process of tracking texture is executed by
slightly moving sub-region disk around target pixel and comparing
texture in each step. Currently, a mathematically defined norm
(summation of echo intensity differences of all pixels within
sub-region) is used as the parameter to check how well newly found
sub-region matches with original sub-region. If the both
sub-regions are a good match, the echo intensity difference between
each pixel is zero, hence the norm is zero. A newly found
sub-region that output the minimum norm is ASSUMED as the perfect
match and the pixel in its center is considered as the new location
of target pixel.
[0089] If two sub-regions match well, the norm evaluated will be
zero. In reality a zero norm can typically not be reached.
Therefore a newly found sub-region that output the minimum norm is
ASSUMED as a good match and the pixel in its center is considered
as the new location of target pixel. Hence setting up a rigor yet
achievable the criteria for norm will be important. On the other
hand, if a large norm is chosen as the criterion for matching, the
matching texture can be found fast yet not rigorous enough. Hence,
setting proper norm criterion for terminating sub-region reach will
be practical.
[0090] Another embodiment includes a feature to present Stiffness
Gradient Histogram of fixed size disk-shaped sub-ROI is added. ROI
analysis can be completed and Stiffness Gradient color map is
presented, user can move cursor to any color coated pixel within
the ROI color map to select the target pixel by clicking action.
Once the target pixel is selected, a fixed size (this size can be
selected by user) disk-shaped sub-ROI will be created around the
target pixel for sub-ROI Stiffness Gradient assessment. Multiple
ROIs and sub-ROIs can be used to generate a Stiffness Gradient
Histogram which can be selected from same analyzed image (case) or
different analyzed image (case) for more standardized compare.
[0091] The drawback of the ROI Stiffness Gradient Histogram
evaluated from user-selected ROI (flexible size and shape) is the
size dependency. If ROIs with significantly different size are
compared, the deduced Stiffness Gradient Histograms will not be
comparable. To solve this drawback, a feature for selecting a fixed
size and disk-shaped sub-ROIs and evaluation of sub-ROI histograms
are included as an embodiment of the invention.
[0092] In addition, following statistical indices are also
evaluated from a Stiffness Gradient Histogram and presented on the
view window Mean: Mean value of the histogram Variance: Second
order moment around mean value that represents the "wideness" of
the histogram
[0093] Skewness: Third order moment around mean value that
represents the "measure of the asymmetry" of the histogram
Kurtosis: Fourth order moment about mean value that represents the
"peakedness" of the histogram Aspect Ratio: Also a parameter shows
the "peakedness" of the histogram.
[0094] It is to be understood that the invention is not limited in
its application to the details of construction and the arrangement
of components set forth in the following description or illustrated
in the following drawings. The invention is capable of other
embodiments and of being practiced or of being carried out in
various ways.
[0095] The disclosed method utilizes unique tissue (material)
properties in a selected ROI and a tracking algorithm and
ultrasound echo single/image analysis algorithm to achieve
evaluation of tissue type/status specific mechanical
functionality.
[0096] A technical contribution for the disclosed method and
apparatus is that it provides for a computer implemented for
ultrasound image processing.
[0097] In accordance with still another aspect of the invention, a
method for deducing simple image output or numerical scale for
monitoring tissue pathology includes feature that evaluates and
presents of a local or total tissue stiffness-strain relation with
selected and processed ROI (region of interest). Then the stiffness
gradient is further evaluated as over-all slope of the deduced
tissue stiffness-strain relation. The deduced tissue functionality
(stiffness, stiffness-gradient) or tissue mechanical behavior
(tissue deformation, displacement, rate of deformation and rate of
displacement) at each image pixel can be presented as easy
understand color map. The same information can be also presented in
the form of histogram plot or probability distribution function
plot. Finally, the shape aspect ratio of these plots can be
calculated as a numerical single number ordinal scale that
represents the tissue health. All these information can be
evaluated from the ultrasound images captured from same
pathological tissue location of the same patient at different
visits. Finally deduced information can be compared and used to
monitor heal or progress of tissue pathology.
[0098] One skilled in the art will appreciate that embodiments of
the invention may be interfaced to and controlled by a computer
readable storage medium having stored thereon a computer program.
The computer readable storage medium includes a plurality of
components such as one or more of electronic components, hardware
components, and/or computer software components. These components
may include one or more computer readable storage media that
generally stores instructions such as software, firmware and/or
assembly language for performing one or more portions of one or
more implementations or embodiments of a sequence. These computer
readable storage media are generally non-transitory and/or
tangible. Examples of such a computer readable storage medium
include a recordable data storage medium of a computer and/or
storage device. The computer readable storage media may employ, for
example, one or more of a magnetic, electrical, optical,
biological, and/or atomic data storage medium. Further, such media
may take the form of, for example, floppy disks, magnetic tapes,
CD-ROMs, DVD-ROMs, hard disk drives, and/or electronic memory.
Other forms of non-transitory and/or tangible computer readable
storage media not list may be employed with embodiments of the
invention.
[0099] A number of such components can be combined or divided in an
implementation of a system. Further, such components may include a
set and/or series of computer instructions written in or
implemented with any of a number of programming languages, as will
be appreciated by those skilled in the art. In addition, other
forms of computer readable media such as a carrier wave may be
employed to embody a computer data signal representing a sequence
of instructions that when executed by one or more computers causes
the one or more computers to perform one or more portions of one or
more implementations or embodiments of a sequence.
[0100] In accordance with one embodiment of the invention, an
ultrasound system includes an ultrasound transducer adapted to
obtain a dynamic series of echo signals of a subject tissue at
different deformation states, and an image processor for generating
and displaying ultrasound images of the tissue. The processor is
configured to generate dynamic images that correspond to the
dynamic series of echo signals, identify a plurality of pixels
within a region of interest (ROI) of a first of the generated
images, evaluate local tissue mechanical behavior by tracking the
displacement, deformation, and echo intensity of the identified
plurality of pixels from the first image to subsequent images based
on groups of pixels that correspond to each of the identified
plurality of pixels, determine tissue functionality in the subject
at the tracked pixel locations, and display the tissue
functionality in dynamic images that corresponds to the tracked
pixel locations.
[0101] In accordance with another embodiment of the invention, a
method of determining a deformed state of a tissue in ultrasound
images, the method includes selecting pixels that are within a
region of interest (ROI) of a first ultrasound image of a tissue,
wherein the tissue is at a first state of deformation, identifying
pixels that surround the selected pixels in the first ultrasound
image, evaluating a local tissue mechanical behavior by tracking
the selected pixels from the first ultrasound image to subsequent
locations in subsequent ultrasound images using the identified
pixels that surround the selected pixels, wherein the subsequent
ultrasound images correspond to different states of tissue
deformation, determining functionality of the tissue at the
subsequent locations of the identified pixels, and displaying the
functionality at their original or subsequent locations in an image
of the tissue.
[0102] In accordance with yet another embodiment of the invention,
a non-transitory computer readable storage medium having stored
thereon a computer program comprising instructions which when
executed by a computer cause the computer to obtain a dynamic
series of echo signals of a subject taken using an ultrasound
transducer with a tissue of the subject at different states of
deformation, generate first and second images using the obtained
series of echo signals, identify target pixels and their
neighboring pixels in the first image, assume locations of the
target pixels and their neighboring pixels in the second image,
calculate values of a mathematical norm based on the target pixels
and their neighboring pixels in the first image and based on their
assumed location in the second image, derive actual locations of at
least the target pixels in the second image based on the values of
the mathematical norm, and overlay deduced functionality or
mechanical behavior of the tissue that correspond to their actual
locations in the original or second image.
[0103] In accordance with still another embodiment of the
invention, a method for monitoring tissue pathology includes
generating a first tissue functionality histogram/probability
density function plot based at least in part upon motion or
deformation of a first pixel between a first ultrasound image at a
first deformed state and a second ultrasound image at a second
deformed state, calculating a first ordinal scale aspect ratio
based upon the first tissue functionality histogram, generating a
second tissue functionality histogram/probability density function
plot based at least in part upon motion of a second pixel between a
third ultrasound image at a third deformed state and a fourth
ultrasound image at a fourth deformed state, calculating a second
ordinal scale aspect ratio based upon the second tissue
functionality histogram/probability density function plot, and
generating a time series plot based on the first ordinal scale
aspect ratio and the second ordinal scale aspect ratio as an
indicator of a tissue pathology.
[0104] This written description uses examples to disclose the
invention, including the best mode, and also to enable any person
skilled in the art to practice the invention, including making and
using any devices or systems and performing any incorporated
methods. The patentable scope of the invention is defined by the
claims, and may include other examples that occur to those skilled
in the art. Such other examples are intended to be within the scope
of the claims if they have structural elements that do not differ
from the literal language of the claims, or if they include
equivalent structural elements with insubstantial differences from
the literal languages of the claims.
* * * * *