U.S. patent application number 15/265463 was filed with the patent office on 2017-01-05 for systems and methods for image processing for modeling changes in patient-specific blood vessel geometry and boundary conditions.
The applicant listed for this patent is HeartFlow, Inc.. Invention is credited to Gilwoo CHOI, David EBERLE, Leo GRADY, Hyun Jin KIM, Sethuraman SANKARAN, Michiel SCHAAP, Charles A. TAYLOR.
Application Number | 20170004280 15/265463 |
Document ID | / |
Family ID | 54142391 |
Filed Date | 2017-01-05 |
United States Patent
Application |
20170004280 |
Kind Code |
A1 |
TAYLOR; Charles A. ; et
al. |
January 5, 2017 |
SYSTEMS AND METHODS FOR IMAGE PROCESSING FOR MODELING CHANGES IN
PATIENT-SPECIFIC BLOOD VESSEL GEOMETRY AND BOUNDARY CONDITIONS
Abstract
Systems and methods are disclosed for modeling changes in
patient-specific blood vessel geometry and boundary conditions
resulting from changes in blood flow or pressure. One method
includes determining, using a processor, a first anatomic model of
one or more blood vessels of a patient; determining a biomechanical
model of the one or more blood vessels based on at least the first
anatomic model; determining one or more parameters associated with
a physiological state of the patient; and creating a second
anatomic model based on the biomechanical model and the one or more
parameters associated with the physiological state.
Inventors: |
TAYLOR; Charles A.; (San
Mateo, CA) ; KIM; Hyun Jin; (San Mateo, CA) ;
SANKARAN; Sethuraman; (Palo Alto, CA) ; SCHAAP;
Michiel; (Mountain View, CA) ; EBERLE; David;
(San Francisco, CA) ; CHOI; Gilwoo; (Mountain
View, CA) ; GRADY; Leo; (Millbrae, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
HeartFlow, Inc. |
Redwood City |
CA |
US |
|
|
Family ID: |
54142391 |
Appl. No.: |
15/265463 |
Filed: |
September 14, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14323128 |
Jul 3, 2014 |
|
|
|
15265463 |
|
|
|
|
14317726 |
Jun 27, 2014 |
9390232 |
|
|
14323128 |
|
|
|
|
61969573 |
Mar 24, 2014 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06F 30/20 20200101;
G16H 50/50 20180101 |
International
Class: |
G06F 19/00 20060101
G06F019/00; G06F 17/50 20060101 G06F017/50 |
Claims
1-20. (canceled)
21. A method for determining a quantity of interest of a patient,
comprising: receiving patient data of the patient at a first
physiological state; determining a value of a quantity of interest
of the patient at the first physiological state based on the
patient data, the quantity of interest representing a medical
characteristic of the patient; extracting features from the patient
data, wherein the extracted features are used to determine the
quantity of interest to be determined for the patient at a second
physiological state; and mapping the value of the quantity of
interest of the patient at the first physiological state to a value
of the quantity of interest of the patient at the second
physiological state using the extracted features.
22. The method as recited in claim 21, wherein mapping the value of
the quantity of interest of the patient at the first physiological
state to the value of the quantity of interest of the patient at
the second physiological state further comprises: mapping the value
of the quantity of interest of the patient at the first
physiological state to the value of the quantity of interest of the
patient at the second physiological state without using data of the
patient at the second physiological state.
23. The method as recited in claim 21, wherein the quantity of
interest of the patient at the first physiological state is a same
quantity of interest as the quantity of interest of the patient at
the second physiological state.
24. The method as recited in claim 21, wherein the quantity of
interest of the patient at the first physiological state is
different from the quantity of interest of the patient at the
second physiological state.
25. The method as recited in claim 21, wherein mapping the value of
the quantity of interest of the patient at the first physiological
state to the value of the quantity of interest of the patient at
the second physiological state further comprises: applying a
trained machine learning function to the value of the quantity of
interest of the patient at the first physiological state, the
machine learning function representing a relationship between the
quantity of interest of a set of patients at the first
physiological state and the quantity of interest of the set of
patients at the second physiological state.
26. The method as recited in claim 25, wherein the trained machine
learning function is based on training data comprising quantities
of interest of the set of patients at the first physiological state
and corresponding quantities of interest of the set of patients at
the second physiological state.
27. The method as recited in claim 21, wherein the patient data
comprises medical image data of the patient, and determining the
value of the quantity of interest of the patient at the first
physiological state comprises: determining the value of the
quantity of interest of the patient at the first physiological
state based on a patient-specific simulation of blood flow
performed using boundary conditions corresponding to the first
physiological state determined based on the medical image data of
the patient.
28. The method as recited in claim 21, the patient data comprises
medical image data of the patient, and extracting features from the
patient data comprises: processing the medical image data of the
patient to determine measurements of the patient.
29. An apparatus for determining a quantity of interest of a
patient, the apparatus executing a method comprising: receiving
patient data of the patient at a first physiological state;
determining a value of a quantity of interest of the patient at the
first physiological state based on the patient data, the quantity
of interest representing a medical characteristic of the patient;
extracting features from the patient data, wherein the extracted
features are based on the quantity of interest to be determined for
the patient at a second physiological state; and mapping the value
of the quantity of interest of the patient at the first
physiological state to a value of the quantity of interest of the
patient at the second physiological state using the extracted
features.
30. The apparatus as recited in claim 29, wherein mapping the value
of the quantity of interest of the patient at the first
physiological state to the value of the quantity of interest of the
patient at the second physiological state further comprises:
mapping the value of the quantity of interest of the patient at the
first physiological state to the value of the quantity of interest
of the patient at the second physiological state without using data
of the patient at the second physiological state.
31. The apparatus as recited in claim 29, wherein the quantity of
interest of the patient at the first physiological state is a same
quantity of interest as the quantity of interest of the patient at
the second physiological state.
32. The apparatus as recited in claim 29, wherein the quantity of
interest of the patient at the first physiological state is
different from the quantity of interest of the patient at the
second physiological state.
33. The apparatus as recited in claim 29, wherein mapping the value
of the quantity of interest of the patient at the first
physiological state to the value of the quantity of interest of the
patient at the second physiological state further comprises:
applying a trained machine learning function to the value of the
quantity of interest of the patient at the first physiological
state, the machine learning function representing a relationship
between the quantity of interest of a set of patients at the first
physiological state and the quantity of interest of the set of
patients at the second physiological state.
34. The apparatus as recited in claim 33, wherein the trained
machine learning function is based on training data comprising
quantities of interest of the set of patients at the first
physiological state and corresponding quantities of interest of the
set of patients at the second physiological state.
35. A non-transitory computer readable medium storing computer
program instructions for determining a quantity of interest of a
patient, the computer program instructions when executed by a
processor cause the processor to perform operations comprising:
receiving patient data of the patient at a first physiological
state; determining a value of a quantity of interest of the patient
at the first physiological state based on the patient data, the
quantity of interest representing a medical characteristic of the
patient; extracting features from the patient data, wherein the
extracted features are used to determine the quantity of interest
to be determined for the patient at a second physiological state;
and mapping the value of the quantity of interest of the patient at
the first physiological state to a value of the quantity of
interest of the patient at the second physiological state using the
extracted features.
36. The non-transitory computer readable medium as recited in claim
35, wherein mapping the value of the quantity of interest of the
patient at the first physiological state to the value of the
quantity of interest of the patient at the second physiological
state further comprises: mapping the value of the quantity of
interest of the patient at the first physiological state to the
value of the quantity of interest of the patient at the second
physiological state without using data of the patient at the second
physiological state.
37. The non-transitory computer readable medium as recited in claim
35, wherein the patient data comprises medical image data of the
patient, and determining the value of the quantity of interest of
the patient at the first physiological state comprises: determining
the value of the quantity of interest of the patient at the first
physiological state based on a patient-specific computational fluid
dynamics simulation of blood flow performed using boundary
conditions corresponding to the first physiological state
determined based on the medical image data of the patient.
38. The non-transitory computer readable medium as recited in claim
35, the patient data comprises medical image data of the patient,
and extracting features from the patient data comprises: processing
the medical image data of the patient to determine measurements of
the patient.
39. A method for determining fractional flow reserve (FFR) for a
coronary stenosis of a patient at a hyperemia state, comprising:
receiving patient data of the patient at a rest state; calculating
a value of a pressure over the coronary stenosis of the patient at
the rest state based on the patient data; extracting features from
the patient data; mapping the value of the pressure over the
coronary stenosis of the patient at the rest state to a value of
the pressure over the coronary stenosis of the patient at the
hyperemia state using the extracted features; and outputting the
FFR for the coronary stenosis of the patient based on the pressure
over the coronary stenosis of the patient at the hyperemia
state.
40. The method as recited in claim 39, wherein the value of the
pressure over the coronary stenosis of the patient at the rest
state comprises a value of a pressure distal to the coronary
stenosis for the patient at the rest state and a value of a
pressure proximal to the coronary stenosis for the patient at the
hyperemia state, and wherein the mapping comprising: mapping the
value of the pressure distal to the coronary stenosis for the
patient at the rest state to a value of the pressure distal to the
coronary stenosis for the patient at the hyperemia state; and
mapping the value of the pressure proximal to the coronary stenosis
for the patient at the rest state to a value of the pressure
proximal to the coronary stenosis for the patient at the hyperemia
state.
41. The method as recited in claim 39, wherein outputting the FFR
comprises: calculating the FFR based on the value of the pressure
distal to the coronary stenosis for the patient at the hyperemia
state and the value of the pressure proximal to the coronary
stenosis for the patient at the hyperemia state.
42. The method as recited in claim 39, wherein the value of the
pressure over the coronary stenosis of the patient at the rest
state comprises a ratio of a value of a pressure distal to the
coronary stenosis for the patient at the rest state and a value of
a pressure proximal to the coronary stenosis for the patient at the
hyperemia state, and wherein the mapping comprises: mapping the
ratio of the pressure over the coronary stenosis of the patient at
the rest state to a ratio of the pressure over the coronary
stenosis of the patient at the hyperemia state.
Description
RELATED APPLICATION(S)
[0001] This application claims priority to U.S. Provisional
Application No. 61/969,573 filed Mar. 24, 2014, the entire
disclosure of which is hereby incorporated herein by reference in
its entirety.
FIELD OF THE INVENTION
[0002] Various embodiments of the present disclosure relate
generally to medical modeling and related methods. More
specifically, particular embodiments of the present disclosure
relate to systems and methods for modeling changes in
patient-specific blood vessel geometry and boundary conditions
resulting from changes in blood flow or pressure.
BACKGROUND
[0003] Coronary artery disease may cause the blood vessels
providing blood to the heart to develop lesions, such as a stenosis
(abnormal narrowing of a blood vessel). As a result, blood flow to
the heart may be restricted. A patient suffering from coronary
artery disease may experience chest pain, referred to as chronic
stable angina during physical exertion or unstable angina when the
patient is at rest. A more severe manifestation of disease may lead
to myocardial infarction, or heart attack.
[0004] A need exists to provide more accurate data relating to
coronary lesions, e.g., size, shape, location, functional
significance (e.g., whether the lesion impacts blood flow), etc.
Patients suffering from chest pain and/or exhibiting symptoms of
coronary artery disease may be subjected to one or more tests that
may provide some indirect evidence relating to coronary lesions.
For example, noninvasive tests may include electrocardiograms,
biomarker evaluation from blood tests, treadmill tests,
echocardiography, single positron emission computed tomography
(SPECT), and positron emission tomography (PET). These noninvasive
tests, however, typically do not provide a direct assessment of
coronary lesions or assess blood flow rates. The noninvasive tests
may provide indirect evidence of coronary lesions by looking for
changes in electrical activity of the heart (e.g., using
electrocardiography (ECG)), motion of the myocardium (e.g., using
stress echocardiography), perfusion of the myocardium (e.g., using
PET or SPECT), or metabolic changes (e.g., using biomarkers).
[0005] For example, anatomic data may be obtained noninvasively
using coronary computed tomographic angiography (CCTA). CCTA may be
used for imaging of patients with chest pain and involves using
computed tomography (CT) technology to image the heart and the
coronary arteries following an intravenous infusion of a contrast
agent. However, obtaining anatomic data using CCTA often means that
models based on the anatomic data reflect a patient's state as
he/she is undergoing imaging (e.g., CCTA imaging). Therefore,
anatomic models for assessing blood flow rates are based on patient
conditions during an imaging procedure. For example,
patient-specific anatomic models for simulating arterial blood flow
are often obtained while a patient is in a baseline condition
during imaging and prior to treatment. However, various forms of
treatment may affect anatomy and consequently, blood flow.
[0006] In other words, a patent's state may change due to any array
of medical procedures and/or health conditions. Meanwhile, models
for assessing blood flow may fail to reflect the change in state.
As a result, there is a need for methods and systems accounting for
changes in a patient's physiological state in indirect assessments
of blood flow rates. In particular, there is a need for methods and
systems for creating an anatomical model based on a patient's
change in state in order to improve the accuracy of a simulation
performed using the model. More specifically, creating an
anatomical model may entail modeling changes in patient-specific
blood vessel geometry and boundary conditions.
[0007] The foregoing general description and the following detailed
description are exemplary and explanatory only and are not
restrictive of the disclosure.
SUMMARY
[0008] According to certain aspects of the present disclosure,
systems and methods are disclosed for anatomical modeling. One
method includes: determining, using a processor, a first anatomic
model of one or more blood vessels of a patient; determining a
biomechanical model of the one or more blood vessels based on at
least the first anatomic model; determining one or more parameters
associated with a physiological state of the patient; and creating
a second anatomic model based on the biomechanical model and the
one or more parameters associated with the physiological state.
[0009] In accordance with another embodiment, a system for
anatomical modeling comprises: a data storage device storing
instructions for anatomical modeling; and a processor configured
for: determining, using a processor, a first anatomic model of one
or more blood vessels of a patient; determining a biomechanical
model of the one or more blood vessels based on at least the first
anatomic model; determining one or more parameters associated with
a physiological state of the patient; and creating a second
anatomic model based on the biomechanical model and the one or more
parameters associated with the physiological state.
[0010] In accordance with yet another embodiment, a non-transitory
computer readable medium for use on a computer system containing
computer-executable programming instructions for anatomical
modeling is provided. The method includes: determining, using a
processor, a first anatomic model of one or more blood vessels of a
patient; determining a biomechanical model of the one or more blood
vessels based on at least the first anatomic model; determining one
or more parameters associated with a physiological state of the
patient; and creating a second anatomic model based on the
biomechanical model and the one or more parameters associated with
the physiological state.
[0011] Additional objects and advantages of the disclosed
embodiments will be set forth in part in the description that
follows, and in part will be apparent from the description, or may
be learned by practice of the disclosed embodiments. The objects
and advantages of the disclosed embodiments will be realized and
attained by means of the elements and combinations particularly
pointed out in the appended claims.
[0012] It is to be understood that both the foregoing general
description and the following detailed description are exemplary
and explanatory only and are not restrictive of the disclosed
embodiments, as claimed.
BRIEF DESCRIPTION OF THE DRAWINGS
[0013] The accompanying drawings, which are incorporated in and
constitute a part of this specification, illustrate various
exemplary embodiments and together with the description, serve to
explain the principles of the disclosed embodiments.
[0014] FIG. 1 is a block diagram of an exemplary system and network
for modeling changes in patient-specific blood vessel geometry and
boundary conditions, according to an exemplary embodiment of the
present disclosure.
[0015] FIG. 2 is a block diagram of an exemplary method of changing
geometry and boundary conditions in a blood flow simulation arising
from different states of a patient, according to an exemplary
embodiment of the present disclosure.
[0016] FIG. 3 is a block diagram of an exemplary method of
determining a second state model of conditions, according to an
exemplary embodiment of the present disclosure.
[0017] FIG. 4 is a block diagram of an exemplary method of
determining an updated geometric model based on the second state
conditions, according to an exemplary embodiment of the present
disclosure.
[0018] FIG. 5 is a block diagram of an exemplary method of
determining geometry responses to different physiologic conditions,
according to an exemplary embodiment of the present disclosure.
DESCRIPTION OF THE EMBODIMENTS
[0019] Reference will now be made in detail to the exemplary
embodiments of the invention, examples of which are illustrated in
the accompanying drawings. Wherever possible, the same reference
numbers will be used throughout the drawings to refer to the same
or like parts.
[0020] Often, patient-specific anatomic models for simulating
arterial blood flow are based on image data associated with one
state. In one example, coronary artery anatomic data may be
obtained under baseline or resting conditions. In another example,
coronary artery anatomic data may be obtained based on an anatomic
state achieved during imaging, including states that increase blood
vessel size and blood flow to improve image quality. Geometric
models may be created and boundary conditions assigned based on the
image data from a baseline condition or imaging conditions.
Simulations modeling reversible, physiological states (e.g., blood
flow simulations associated with drugs, exercise, and/or treatment)
are often performed based on the anatomic and geometric model
associated with the first state. However, drugs, exercise, and/or
treatment may all cause changes in blood vessel geometry and
boundary conditions from the first state. For example, a geometry
of a patient's anatomy may change due to various conditions or
treatments, including administration of drugs (e.g., adenosine or
other drugs to increase blood flow), simulations of medical
conditions (e.g., simulated hyperemia), simulations of physical
activities or conditions (e.g., exercise), angioplasty, surgery
(e.g., stenting or bypass grafting), etc. Therefore, a desire
exists for patient-specific models for simulating arterial blood
flow that may account for a representation of a patient's state,
where the patient's state may differ from a state from which the
anatomic model was built. Simulating arterial blood flow using a
patient-specific model reflecting a second state may improve
accuracy of simulation results. Particularly, the present
disclosure is directed to second state(s) that may include
reversible, physiological states. Furthermore, simulations and
models based on the second state may further be applied to model
possible treatments that may affect geometry (e.g., angioplasty,
stenting, and/or bypass surgery). For example, a geometric change
to a model may be made (e.g., to model stenting), based on
patient-specific models that reflect a second state. The following
discussion outlines various scenarios where an anatomic and
biomechanical model under which simulations are performed, may not
accurately represent a patient's state.
[0021] In one embodiment, simulations may be performed using
patient-specific anatomic models based on image data obtained under
resting conditions. Geometric models and boundary condition models
based on these baseline conditions may then be used as input to
computer models in order to predict flow and pressure under a
physiologic state, including during the administration of adenosine
or other drugs to increase blood flow and simulate exercise, or
after angioplasty and stenting or bypass grafting. The patient's
anatomy may be at a state distinct from the first, resting state,
in light of one or more treatments or conditions. Therefore, a
desire exists for patient-specific models for simulating arterial
blood flow to take into account a second state reflecting patient
anatomy at a non-resting state.
[0022] In another embodiment, a patient-specific model extracted
from image data may be based on a state distinct from a baseline
state. For instance, in the case of coronary artery anatomic data,
beta blockers may be used to reduce heart rate, while nitrates may
be administered to dilate large coronary arteries. Both drugs may
be administered to improve image quality. For example, beta
blockers used to slow the heart may affect blood pressure and
hence, the size of a vessel; and nitrates used during coronary
computed tomography (CT) angiography may increase flow by relaxing
smooth muscle cells in blood vessels, decreasing their tension (or
tone), and increasing the size of the blood vessels. The increased
size and flow through the vessels improves image quality. The
administration of beta blockers and/or nitrates may cause geometry
and physiologic conditions to change to a state that may be
different from a baseline state. However, the new state of the
arteries from the administration of beta blockers and/or nitrates
changes geometry and physiologic conditions to a state that may be
different from a baseline. In other words, modeled changes in blood
flow and pressure may cause changes in patient-specific geometric
models and boundary conditions, since local vessel size may be
affected by local pressure and smooth muscle tone of the vessels
(which can be affected by administration of nitrates, adenosine,
papaverine, adenosine triphosphate (ATP), etc.). However, an image
created from baseline conditions may not account for the affect
that drugs may have on anatomy geometry and boundary conditions. A
blood flow simulation performed under the state may be expected to
yield diagnostic data different from that attained prior to the
administration of the drugs. Thus, the present disclosure is
directed to a new approach including changing geometry and boundary
conditions in a blood flow simulation to model an original baseline
or resting state of arteries prior to administration of drugs.
[0023] A specific example of the above embodiment may include
modeling of increased flow as occurs during simulated hyperemia.
Such modeling may be performed to calculated fractional flow
reserve or coronary flow reserve. The simulations of increased flow
may be typically performed based on coronary anatomic data obtained
under baseline or resting conditions. In reality, the data is often
obtained subsequent to administration of beta blockers and/or
nitrates. More specifically, simulation of increased blood flow
through coronary arteries may result in pressure changes along the
coronary arteries, especially downstream of a coronary artery
stenosis. The metric of FFR may be calculated from the ratio of
downstream pressure to aortic pressure. As a result during the
simulation of hyperemia (performed using anatomic data obtained at
baseline), blood pressure may be significantly lower at points
downstream of the vessel rather than at points upstream of the
vessel. Also, blood pressure may be significantly lower during the
hyperemic state than during the resting state. The blood vessels
may diminish in size (i.e., "deflate") due to the reduced pressure.
Such changes in vessel size may affect tightness of a coronary
artery stenosis or the caliber of vessels downstream from the
stenosis. This in turn may affect the accuracy of the hyperemic
simulation and accuracy of the predicted FFR, as compared to
measured data (obtained during actual administration of
vasodilators causing increased flow and pressure reduction along
the length of the vessel). Thus, the present disclosure is directed
to a new approach for changing geometry and boundary conditions in
a blood flow simulation model of the hyperemic state of arteries
using image data obtained without administration of drugs to
increase blood flow.
[0024] Furthermore, treatment recommendations may be improved with
modeling taking into account changes in patient-specific blood
vessel geometry and boundary conditions. For example, percutaneous
coronary intervention (PCI) or coronary artery bypass grafting
(CABG) is often used to treat patients with coronary artery
disease. Computer models are often used to predict changes in blood
flow or pressure resulting from the treatments to aid the physician
in deciding how best to treat a given patient. Patient-specific
models for simulating PCI or CABG may be created from pre-treatment
image data, then modified to incorporate a treatment plan. The
modifications are generally restricted to geometric changes in
diseased segments to account for dilation of stenosis with PCI or
creation of an alternate conduit for blood flow with CABG. However,
treatments may affect more than simply diseased segments.
Treatments potentially change blood flow and pressure in an entire
coronary artery tree. Therefore, the present disclosure is further
directed to a new approach for modeling geometric changes and
boundary condition changes secondary to changes in blood flow or
pressure resulting from treatments for arterial disease. For
example, the present disclosure may include updating geometric
model and boundary conditions (created from pre-treatment data) to
account for new post-treatment flow and pressure. In other words,
the present disclosure may include changing geometry and boundary
conditions in a blood flow simulation to model post-treatment state
of arteries due to predicted changes in blood flow and pressure
from models originally created using image data obtained prior to
treatment.
[0025] In a broader sense, the present disclosure is directed to a
new approach for systems and methods for modeling changes in
patient-specific blood vessel geometry and boundary conditions
based on changes in blood flow or pressure.
[0026] Referring now to the figures, FIG. 1 depicts a block diagram
of an exemplary system and network for modeling changes in
patient-specific blood vessel geometry and boundary conditions.
Specifically, FIG. 1 depicts a plurality of physicians 102 and
third party providers 104, any of whom may be connected to an
electronic network 100, such as the Internet, through one or more
computers, servers, and/or handheld mobile devices. Physicians 102
and/or third party providers 104 may create or otherwise obtain
images of one or more patients' cardiac and/or vascular systems.
The physicians 102 and/or third party providers 104 may also obtain
any combination of patient-specific information, such as age,
medical history, blood pressure, blood viscosity, etc. Physicians
102 and/or third party providers 104 may transmit the
cardiac/vascular images and/or patient-specific information to
server systems 106 over the electronic network 100. Server systems
106 may include storage devices for storing images and data
received from physicians 102 and/or third party providers 104.
Server systems 106 may also include processing devices for
processing images and data stored in the storage devices.
[0027] FIG. 2 is a block diagram of an exemplary method 200 of
changing geometry and boundary conditions in a blood flow
simulation to model a second state of a patient different from a
first state of the patient (e.g., from the state in which the
patient was imaged), according to an exemplary embodiment. The
second state, for example, may be (1) a resting state, free of the
administration of drugs used during imaging, (2) a hyperemic state
of arteries, free of drugs used to increase blood flow, (3) a
post-treatment state, or (4) any other desired state.
[0028] In one embodiment, step 201 may include constructing a
patient-specific anatomic model. In one embodiment, the model may
be from two-dimensional imaging modalities (e.g., coronary
angiography, biplane angiography, etc.) or three-dimensional
imaging modalities (e.g., 3-D rotational angiography, coronary
computed tomographic angiograph (cCTA), magnetic resonance
angiography (MRA)). Step 201 may further include directly
segmenting image data and creating a patient-specific
three-dimensional anatomic model of the patient's arteries.
Alternately or in addition, step 201 may involve modifying a
previously-constructed "generic" model, customizing the model for a
particular patient, and creating a patient-specific model. In yet
another embodiment, step 201 may include providing, receiving,
and/or loading a patient-specific anatomic model of a patient into
a computer. For example, the model may be from an electronic
storage device (e.g., a hard drive, network drive, etc.). In one
embodiment, the model may represent a first, baseline state of a
patient.
[0029] In any or all of the embodiments of step 201, the
patient-specific anatomic model may include information related to
arteries of interest, including the length of each segment,
diameter along the length of a segment (or any other geometric
description of the segment), branching patterns, presence of
disease, characteristics of disease (including composition of
atherosclerotic plaques), etc. A representation of the
patient-specific model may be defined by a surface enclosing a
three-dimensional volume, a one-dimensional model where the
centerline of the vessels is defined together with cross-sectional
area information along the length, and/or an implicit
representation of a vessel surface.
[0030] In one embodiment, step 203 may include defining physiologic
conditions associated with blood flow and pressure that reflect a
patient's condition at the time that imaging was taken. Conditions
at the time of imaging may make up a "first (physiological) state"
for a patient. For example, a patient may be administered beta
blockers to lower his heart rate and/or sublingual nitrates to
dilate his coronary arteries in order to improve image quality.
Step 203 of determining physiologic conditions may include
determining and/or assigning aortic pressure conditions and
resistance of coronary artery microcirculation based on a patient's
intake of beta blockers and/or nitrates.
[0031] In one embodiment, step 205 may include creating a
biomechanical model of a vessel wall, for example, generating a
biomechanical model for each segment of artery extracted in the
patient-specific anatomic model of step 201. In one embodiment, the
vessel wall model may be based on one-dimensional elastic or
viscoelastic models of blood vessels. Such models may include
models that typically relate pressure to vessel cross-sectional
area along the length of a vessel. Exemplary models are described
in Olufsen et al. (Olufsen M S. "Structured tree outflow condition
for blood flow in larger systemic arteries." Am J Physiol Heart
Circ Physiol 276:H257-H268, 1999.), Wan et al. (J. Wan, B. N.
Steele, S. A. Spicer, S. Strohband, G. R. Feijoo, T. J. R. Hughes,
C. A. Taylor (2002) "A One-dimensional Finite Element Method for
Simulation-Based Medical Planning for Cardiovascular Disease."
Computer Methods in Biomechanics and Biomedical Engineering. Vol.
5, No. 3, pp. 195-206.), and Raghu et al. (R. Raghu, I. E.
Vignon-Clementel, C. A. Figueroa, C. A. Taylor (2011) "Comparative
Study of Viscoelastic Arterial Wall Models in Nonlinear
One-dimensional Finite Element Simulations of Blood Flow." Journal
of Biomechanical Engineering, Vol. 133, No. 8, pp 081003.).
Alternately, biomechanical models of vessel wall may represent the
vessel wall as a surface with spatially-varying thickness and
material properties, for example, as described in Figueroa et al.
(C. A. Figueroa, I. E. Vignon-Clementel, K. C. Jansen, T. J. R.
Hughes, C. A. Taylor (2006) "A Coupled Momentum Method For Modeling
Blood Flow In Three-Dimensional Deformable Arteries." Computer
Methods in Applied Mechanics and Engineering, Vol. 195, Issues
41-43, pp. 5685-5706.) or in Figueroa et al. (C. A. Figueroa, S.
Baek, C. A. Taylor, J. D. Humphrey (2009) "A Computational
Framework for Coupled Fluid-Solid Growth Modeling in Cardiovascular
Simulations." Computer Methods in Applied Mechanics and
Engineering, Vol. 198, No. 45-46, pp. 3583-3602.). Another example
of a biomechanical model may include a blood vessel as a
three-dimensional continuum model, as in Gee et al. (Gee M W,
Forster C, Wall W A (2010) "A computational strategy for
prestressing patient-specific biomechanical problems under finite
deformation." Int J Numer Methods Biomed Eng 26(1):52-72.), Gerbeau
et al. (Gerbeau J-F, Vidrascu M, Frey P (2005) "Fluid-structure
interaction in blood flows on geometries based on medical imaging."
Comput Struct 83(2-3):155-165.), or as in Kioussis et al. (Kiousis
D E, Gasser T C, Holzapfel G A. 2007. "A numerical model to study
the interaction of vascular stents with human atherosclerotic
lesions." Ann. Biomed. Eng. 35:1857-69.). Material properties of
vessel walls may be defined based on population averaged material
properties, imaging data, and/or data inferred by experimental
measurement of deformation of coronary arteries during a cardiac
cycle and solving an inverse optimization problem to estimate the
best constitutive fit consistent with data. Examples of
constitutive models include linear elastic, hyperelastic, linear
and nonlinear viscoelastic models including those discussed in Wan
et al. (J. Wan, B. N. Steele, S. A. Spicer, S. Strohband, G. R.
Feijoo, T. J. R. Hughes, C. A. Taylor (2002) A One-dimensional
Finite Element Method for Simulation-Based Medical Planning for
Cardiovascular Disease. Computer Methods in Biomechanics and
Biomedical Engineering. Vol. 5, No. 3, pp. 195-206), Raghu et al.
([R. Raghu, I. E. Vignon-Clementel, C. A. Figueroa, C. A. Taylor
(2011) Comparative Study of Viscoelastic Arterial Wall Models in
Nonlinear One-dimensional Finite Element Simulations of Blood Flow.
Journal of Biomechanical Engineering, Vol. 133, No. 8, pp 081003.)
and Taylor et al. (C. A. Taylor, J. D. Humphrey (2009) Open
Problems in Computational Vascular Biomechanics: Hemodynamics and
Arterial Wall Mechanics. Computer Methods in Applied Mechanics and
Engineering, 198, No. 45-46, pp. 3514-3523). These material models
may be purely phenomenological stress-strain relations or
phenomenological models that are based on the microstructure of
blood vessels, e.g. including data on collagen and elastin fiber
orientation derived from experimental data, see for example
Humphrey et al. (J. D. Humphrey, Cardiovascular Solid Mechanics:
Cells, Tissues, and Organs, Springer, New York, 2002.) and
Holzapfel et al. (G. A. Holzapfel, T. C. Gasser, R. W. Ogden, A new
constitutive framework for arterial wall mechanics and a
comparative study of material models, J. Elasticity (2000)
1-48).
[0032] In one embodiment, an elastic modulus of a vessel wall may
be roughly estimated from a Hounsfield unit (HU) of tissue
surrounding a lumen boundary. Thickness of a vessel wall may be
estimated from image data and/or approximated by a theoretical
relationship between vessel radius and wall thickness, e.g.,
assuming the thickness is 1/5.sup.th or 1/10.sup.th of the radius.
Vessel wall models may represent material behavior passively or may
include active behavior to model tension due to smooth muscle tone
in the vessel wall. The material properties may be affected by
pressure, flow, wall shear stress, wall tensile stress, and/or
vasoactive drugs that may alter tension in the vessel wall (e.g.,
by inducing smooth muscle cell contraction or relaxation).
[0033] In one embodiment, steps 201-205 of determining a
patient-specific geometrical model, a physiologic model, and a
biomechanical model may all pertain to a "first state." In some
embodiments, such a first-state model may represent the patient's
conditions when imaging was performed.
[0034] In one embodiment, step 207 may include defining physiologic
conditions, boundary conditions, and/or material properties of a
patient in a second state, other than the first state. For example,
physiologic conditions and boundary conditions of a patient under
hyperemic conditions may be defined using a method described in
U.S. Pat. No. 8,315,812 issued Nov. 20, 2012, the entire disclosure
of which is hereby incorporated by reference in its entirety. The
physiologic conditions and boundary conditions of a patient after
treatment may be defined using the method described in U.S. Pat.
No. 8,249,815 issued Aug. 21, 2012, the entire disclosure of which
is hereby incorporated in reference in its entirety. In one
embodiment, changes in elastic properties of a blood vessel may be
modified for a second state based on an expected response to
medications (e.g., nitrates). For example, if nitrates were used
during imaging of a patient, a second state may include determining
vasoactive response of arteries in response to a "removal" of
nitrates. The second state may thus include vasoconstriction of
arteries relative to the first state, which may closer model a
patient's anatomy and/or physiology under resting conditions. In
some embodiments, step 207 may include determining changes in
properties based on data in literature. For example, an expected
response of nitrates known in literature, is an increase in
diameters of 0% to 30%, depending on the size of a vessel and
whether it is healthy or diseased. In another embodiment, if image
data is available for a population of patients with and without
nitrates, changes in vessel size due to administration of nitrates
may be determined using machine learning methods. The data may then
be used to update vessel properties for the second state of the
patient. FIG. 5, described further herein, provides further detail
on machine learning methods for determining changes in geometry
with respect to various states.
[0035] In one embodiment, step 209 may include generating an
anatomic model of the second state, based on flow and pressure
conditions of the patient in a second state. In one embodiment,
step 209 may include updating and/or revising a patient-specific
first-state model (e.g., the patient-specific anatomic model from
step 201). For example, step 209 may include simulating blood flow
and pressure of the patient in the second state, using the
patient-specific anatomic model and biomechanical model of the
patient in the first state. In other words, step 209 may include
simulating blood flow and pressure in the first-state model, along
with boundary conditions and/or material properties associated with
the second-state model. Further detail regarding step 209 is
provided in FIGS. 3 and 4.
[0036] In one embodiment, step 211 may include performing
simulations using a model reflecting a patient's second state. For
example, step 211 may include performing a simulation of blood flow
and pressure using the second-state model. Furthermore, step 213
may include providing and/or outputting results of the simulation
in the form of a report via a computer output device.
[0037] FIG. 3 is a block diagram of an exemplary method 300 of
determining a second-state model of conditions, according to an
exemplary embodiment. In one embodiment, step 301 may include
determining various available models and/or patient conditions. For
example, a second-state model may include (i) an original baseline
or resting state of blood vessels (e.g., arteries) prior to
administration of drugs (e.g., to improve image data), (ii) a
hyperemic state subsequent to administration of a drug to increase
blood flow (e.g., adenosine, papaverine, ATP, Regadenoson, etc.),
(iii) a simulated exercise state, (iv) a post-treatment state, etc.
In one embodiment, step 303 may include determining which of the
available models is of interest. For example, step 303 may include
selecting one or more of the available models as a second-state
model based on user selection, inferences from input associated
with the first-state model, patient information, etc.
[0038] In one embodiment, step 305 may include determining
conditions associated with the selected model. For example, in
response to a hyperemic state, physiologic condition changes may
include: aortic pressure decreases, heart rate increases, vascular
microcirculatory resistance decreases, healthy arteries dilating in
response to flow, stenosis or segments of arteries downstream of
disease reducing in size in response to pressure changes, etc. In
another example, a response to a simulated exercise state may
include the following physiologic condition changes: cardiac output
increases, aortic pressure increases, heart rate increases,
vascular microcirculatory resistance decreases, healthy arteries'
dilation in response to flow, stenosis or segments of arteries
downstream of disease reducing in size in response to pressure
changes, etc. For a post-treatment state, for example subsequent to
treatment including angioplasty and stenting or bypass surgery,
local blood pressure and flow along an arterial tree may be altered
for resting conditions and high flow conditions (e.g., hyperemia,
exercise, etc.).
[0039] In one embodiment, step 307 may include computing forces on
blood vessel walls for the second state. For example, step 307 may
include simulating blood flow and pressure in the first-state model
and using results of the simulation to modify physiologic boundary
conditions to represent those of the second state. In one
embodiment, step 307 may be performed using, for example, (i) a
reduced order model (e.g., a lumped-parameter or one-dimensional
wave propagation model), (ii) a three-dimensional finite element,
finite volume, lattice Boltzman, level set, immersed boundary, or
particle-based method to solve 3-D equations of blood flow and
pressure, or (iii) a fluid-structure interaction method to solve
for blood flow, pressure, and vessel wall motion.
[0040] FIG. 4 is a block diagram of an exemplary method 400 of
determining an updated geometric model based on the second-state
conditions, according to an exemplary embodiment. In other words,
method 400 may be directed at determining a geometric model based
on a biomechanical model of a patient's arteries (e.g., a model
from method 300). In one embodiment, step 401 may include
determining a relationship between geometry and biomechanical
properties. For example, step 401 may include determining one or
more one-dimensional elastic and/or viscoelastic models of blood
vessels relating pressure to vessel cross-sectional area along the
length of a vessel. In such a case, a representative pressure
diameter curve may be calibrated to match pre-treatment
pressure-diameter values at different centerline points. Then, a
new lumen diameter may be estimated by probing the diameter of the
calibrated curve at the new pressure. Alternatively, step 401 may
include solving stress-equilibrium equations for a computational
model of the vessel wall, with the pressure difference between the
first state and the second state acting in the inner wall and a
zero traction boundary condition acting on the outer surface of the
vessel.
[0041] In one embodiment, step 403 may include solving the models
and/or equations from step 401 to determine geometry based on the
biomechanical data. For example, step 403 may include solving for
geometry correction iteratively along with computational fluid
dynamics (CFD) (e.g., using predictor-corrector methods). Such a
method of determining the geometry may be possible because changes
in geometry affect flow rate and blood pressure. Alternatively,
geometry may be solved for in a coupled manner using an arbitrary
Lagrangian-Eulerian framework. In another example, if experimental
data relating changes in geometry to different physiologic
conditions is available, machine learning methods may be used to
model how cross-sectional area of vessels change locally, given a
change in pressure and surrounding geometry. In some cases,
training data from other patients may be used to inform the model
to predict area changes from pressures computed in method 300.
[0042] In one embodiment, step 405 may include determining a
deformation that may be computed to update the geometrical model.
For example, step 405 may include determining a minimal deformation
that creates a segmentation with desired cross-sectional area at
each location, that may then be computed to update a geometry. For
example, the flow domain and vessel walls may be represented by an
explicit mesh. This explicit mesh may be modified by a variety of
elastic deformation techniques. Alternatively, a flow domain may
have an implicit representation deformed by a speed function using
a level set method. A level set method may permit tracking shapes
by building a surface from two-dimensional boundaries of shapes,
where the shapes may include level "slices" of the surface. For
example, a speed function may be used to change a representation
for a level set method defined by computed desired cross sectional
areas along the centerline. The speed function for a level set
method may include terms to control the curvature of the implicit
surface as it is modified from the first state to the second state.
Step 407 may include producing a patient-specific anatomic model to
represent a patient in second state conditions. Producing the
second-state model may include updating a first-state geometric
model. For example, step 407 may include deforming an implicit
representation. Step 407 may further include determining whether to
mesh the implicit representation with other representations or use
the implicit representation directly. Step 409 may include using
either a mesh of multiple implicit representations or a single
implicit representation of modified boundary conditions for
display, calculations of CFD equations, or a combination thereof.
For example, step 409 may include performing a simulation of blood
flow and pressure using the second-state patient-specific anatomic
model and/or biomechanical model. In a further example, step 409
may include providing information based on the simulation to a
user, for instance, through a report or display via a computer
output device.
[0043] In one embodiment, method 400 may include further modeling
geometric changes based on post-treatment states (e.g.,
angioplasty, stenting, or bypass surgery). For example, deforming
the mesh for an anatomic model in step 407 may include accounting
for a geometry of a stent or a geometry post-angioplasty. Then, a
simulation of step 409 may include simulating blood flow and
pressure through the anatomic models built from physiologic state
boundary conditions and geometry, as well as treatment-related
geometry. As a further step, results from the simulations may be
output or displayed. For example, such output may include a
treatment recommendation, where several simulations may be run to
simulate various treatment options.
[0044] FIG. 5 is a block diagram of an exemplary method 500, such
as machine learning methods, of determining geometry responses to
different physiologic conditions, according to an exemplary
embodiment. In one embodiment, step 501 may include determining
information of a population of patients (e.g., patient age, gender,
physical conditions, height, weight, diet, family medical history,
etc.). Step 503 may include determining image and/or experimental
data associated with the population of patients. For example, the
image and/or experimental data may characterize the population of
patients as a group. Alternately, image and/or experimental data
may include data respective to each patient in the population of
patients.
[0045] In one embodiment, step 505 may include determining or
calculating a value of interest associated with the image and/or
experimental data. For example, a value of interest may be a
measurement (e.g., material properties of a vessel wall) and/or a
change in a measurement (e.g., changes in vessel size due to
administration of a nitrate). In any case, step 505 may include
computing, for each patient in a population of patients, the value
of interest. Step 505 may further include averaging the values for
an entire population of patients. Step 507 may then include
predicting a change in geometry based on the values given by the
population of patients. For example, step 507 may include using the
values from step 505 to model how cross-sectional area of a vessel
changes locally, given change in pressure and surrounding geometry.
Step 507 may then help predict area changes from pressures computed
by biomechanical modeling based on physiologic conditions.
[0046] Various embodiments of the present disclosure relate to
medical modeling and related methods, specifically, modeling
changes in patient-specific anatomic models. For example, the
present disclosure includes calculating blood flow and pressure in
patient-specific arterial models updated to reflect geometric and
boundary condition changes. In some embodiments, the changes arise
from a state change subsequent a state of a patient in which
imaging was performed. Some instances of applications for such
modeling include (i) resting, exercise, or hyperemic conditions
using image data obtained subsequent administration of nitrates
and/or beta blockers and/or (ii) post-treatment conditions using
image data obtained prior to treatment. The present disclosure
describes the systems and methods directed to coronary arteries,
but the disclosure may also apply to simulations of blood flow and
pressure in any arterial tree including but not limited to the
carotid, cerebral, renal, and lower extremity arteries.
[0047] Other embodiments of the invention will be apparent to those
skilled in the art from consideration of the specification and
practice of the invention disclosed herein. It is intended that the
specification and examples be considered as exemplary only, with a
true scope and spirit of the invention being indicated by the
following claims.
* * * * *