U.S. patent application number 17/399704 was filed with the patent office on 2021-12-02 for systems and methods for patient-specific imaging and modeling of drug delivery.
The applicant listed for this patent is HeartFlow, Inc.. Invention is credited to Gilwoo CHOI, Leo GRADY, Stanley C. HUNLEY, Charles A. TAYLOR.
Application Number | 20210375401 17/399704 |
Document ID | / |
Family ID | 1000005767971 |
Filed Date | 2021-12-02 |
United States Patent
Application |
20210375401 |
Kind Code |
A1 |
CHOI; Gilwoo ; et
al. |
December 2, 2021 |
SYSTEMS AND METHODS FOR PATIENT-SPECIFIC IMAGING AND MODELING OF
DRUG DELIVERY
Abstract
Systems and methods are disclosed for providing personalized
chemotherapy and drug delivery using computational fluid dynamics
and medical imaging with machine learning from a vascular
anatomical model. One method includes receiving a patient-specific
anatomical model of at least one vessel of the patient and a target
tissue where a drug is to be supplied; receiving patient-specific
information defining the administration of a drug; deriving
patient-specific data from the patient specific anatomical model
and/or the patient; determining one or more blood flow
characteristics in a vascular network leading to the one or more
locations in the target tissue where drug delivery data will be
estimated or measured, using the patient-specific anatomical model
and the patient-specific data; and computing drug delivery data at
the one or more locations in the target tissue using
transportation, spatial, and/or temporal distribution of the drug
particles.
Inventors: |
CHOI; Gilwoo; (Mountain
View, CA) ; GRADY; Leo; (Darien, CT) ; TAYLOR;
Charles A.; (Atherton, CA) ; HUNLEY; Stanley C.;
(San Jose, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
HeartFlow, Inc. |
Redwood City |
CA |
US |
|
|
Family ID: |
1000005767971 |
Appl. No.: |
17/399704 |
Filed: |
August 11, 2021 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15267671 |
Sep 16, 2016 |
11120893 |
|
|
17399704 |
|
|
|
|
62219490 |
Sep 16, 2015 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G16H 50/50 20180101;
G16C 20/30 20190201; G16H 20/10 20180101; A61B 5/4848 20130101 |
International
Class: |
G16C 20/30 20060101
G16C020/30; A61B 5/00 20060101 A61B005/00; G16H 50/50 20060101
G16H050/50; G16H 20/10 20060101 G16H020/10 |
Claims
1-20. (canceled)
21. A computer-implemented method of estimating drug delivery at a
target tissue, the method comprising: receiving one or more
patient-specific models of (1) a target tissue where a drug is
delivered, and (2) of a vascular network supplying blood to the
target tissue from each of a plurality of individuals with known
drug delivery data at one or more locations of the target tissue;
receiving information on drug administration for each of the
plurality of individuals with known drug delivery data from whom
patient-specific models were received; deriving one or more patient
specific data from the patient-specific models from whom
patient-specific models were received, the patient-specific data
including one or more physiological conditions; receiving drug
delivery data for one or more locations of the target tissue for
each of the plurality of individuals with known drug delivery data
from whom patient-specific models were received; calculating
feature vectors including information regarding the one or more
locations on the target tissue with known drug delivery data,
information on drug administration and the patient-specific data,
for each of the plurality of individuals with known drug delivery
data from whom patient-specific models were received; associating
the feature vectors with the received drug delivery data at one or
more locations of the target tissue, for each of the plurality of
individuals with known drug delivery data from whom
patient-specific models were received; and training a system that
can predict drug delivery data at one or more locations on a
patient-specific model of a target tissue and of a vascular network
supplying blood to the target tissue, from one or more
patient-specific data and information on drug administration, using
the associated feature vectors.
22. The computer-implemented method of claim 21, further
comprising: receiving a patient-specific model of the target tissue
where drug is delivered and of the vascular network supplying blood
to the target tissue of a patient seeking analysis; receiving
information on drug administration in the patient seeking analysis;
receiving data on one or more locations in the target tissue of the
patient seeking analysis where drug delivery data will be estimated
or measured; deriving one or more patient specific data from the
patient-specific model and/or the patient seeking analysis; forming
feature vectors comprising of information regarding the one or more
locations on the target tissue where drug delivery data will be
estimated or measured, information on drug administration and the
patient-specific data; applying the trained system to estimate the
drug delivery data at one or more locations on the target tissue of
the patient-specific model of the patient seeking analysis, using
the feature vectors; and outputting the estimated or measured drug
delivery data to an electronic storage medium or display.
23. The computer-implemented method of claim 22, wherein the
estimated or measured drug delivery data within the electronic
storage medium includes data extracted from one or more scanning
modalities, including, but not limited to, MR, FDG-PET, stress
echo/MRI contractile reserve, multidetector CT, dual energy CT,
.mu.CT, or .mu.MR.
24. The computer-implemented method of claim 21, wherein training a
system that can predict drug delivery data includes using, one or
more of: a support vector machine (SVM), a multi-layer perceptron
(MLP), a multivariate regression (MVR), deep learning, random
forests, k-nearest neighbors, Bayes networks, and a weighted linear
or logistic regression.
25. The computer-implemented method of claim 21, further
comprising: developing a patient-specific fluid dynamics model of
blood flow in the patient-specific models, using patient-specific
data, for each of the plurality of individuals with known drug
delivery data at one or more locations of the target tissue;
calculating a transportation, spatial, and/or temporal distribution
of drug particles in the fluid dynamics model of blood flow, using
the information on drug administration, for each of a plurality of
individuals with known drug delivery data at one or more locations
of the target tissue; and incorporating information related to a
transportation, spatial, and/or temporal distribution of the drug
particles into the feature vectors, for each of a plurality of
individuals with known drug delivery data at one or more locations
of the target tissue.
26. The computer-implemented method of claim 21, wherein the
information on drug administration includes drug administration
amount, drug concentration, administration location, administration
frequency, route of drug administration, administration time, type
of therapy, or a combination thereof.
27. The computer-implemented method of claim 21, wherein the one or
more patient specific data includes: a vascular anatomical image
characteristic; a target tissue image characteristic; an estimated
perfusion territory in the target tissue; an estimated blood supply
to the target tissue; estimated blood flow data; a patient
characteristic; a disease burden characteristic; an
electromechanical measurement; the transportation, spatial, and/or
temporal distribution of drug particles in the vascular network; or
a combination thereof.
28. The computer-implemented method of claim 21, wherein the one or
more drug delivery data includes, one or more of: an estimate of an
amount of drug delivered to the one or more locations in the target
tissue; an estimate of a concentration of drug particles delivered
to the one or more locations in the target tissue; a circulatory
destination probability of drug particles released at the drug
delivery location to the target tissue, the circulatory destination
probability based on a ratio of the amount of drug particles
reaching the target tissue with respect to a total number of
released drug particles; a transportation, spatial, and/or temporal
distribution of the drug; an estimate blood flow data; or a
combination thereof.
29. The computer-implemented method of claim 21, wherein the target
tissue is, one or more of: a tissue or organ affected with tumorous
growth, including, one or more of, a brain, breast, prostate,
cervix, lung, skin, colon, or stomach; a tissue or organ affected
by a stenosis within the vascular and/or a microvascular network;
or a tissue or organ affected by a thrombosis within the vascular
and/or a microvascular network.
30. The computer-implemented method of claim 22, further including:
receiving one or more desired drug delivery data at one or more
locations in the target tissue; assessing effectiveness of drug
delivery by comparing one or more actual drug delivery data at the
one or more locations in the target tissue with the desired drug
delivery data at the one or more locations in the target tissue;
and reconfiguring the drug administration so that one or more
actual drug delivery data at the one or more locations in the
target tissue matches or is within range of the desired drug
delivery data at the one or more locations in the target
tissue.
31. The computer-implemented method of claim 30, wherein the
reconfiguring the drug administration includes systematically
adjusting one or more of a drug amount, a drug concentration, a
drug administration location, a drug administration frequency, a
route of drug administration, an administration time, a type of
therapy, or a combination thereof, in order to increase the
effectiveness of the drug delivery.
32. The computer-implemented method of claim 22, further including:
receiving, at different time points, one or more medical images of
patients; extracting patient specific data from the one or more
medical images; assessing effectiveness of drug delivery by
comparing the patient specific data extracted from the two images;
and reconfiguring the drug administration in order to increase the
effectiveness of drug delivery.
33. The computer-implemented method of claim 32, wherein the
effectiveness of drug delivery is assessed by comparing the one or
more medical images and determining a degree to which a tumor or
lesion has regressed.
34. The computer-implemented method of claim 32, wherein the
reconfiguring the drug administration includes systematically
adjusting one or more of a drug amount, a drug concentration, a
drug administration location, a drug administration frequency, a
route of drug administration, an administration time, a type of
therapy, or a combination thereof, in order to increase the
effectiveness of the drug delivery.
35. A system for estimating drug delivery at a target tissue, the
system comprising: a data storage device storing instructions for
estimating drug delivery at a target tissue; and a processor
configured to execute the instructions to perform a method
including: receiving one or more patient-specific models of (1) a
target tissue where a drug is delivered, and (2) of a vascular
network supplying blood to the target tissue from each of a
plurality of individuals with known drug delivery data at one or
more locations of the target tissue; receiving information on drug
administration for each of the plurality of individuals with known
drug delivery data from whom patient-specific models were received;
deriving one or more patient specific data from the
patient-specific models from whom patient-specific models were
received, the patient-specific data including one or more
physiological conditions; receiving drug delivery data for one or
more locations of the target tissue for each of the plurality of
individuals with known drug delivery data from whom
patient-specific models were received; calculating feature vectors
including information regarding the one or more locations on the
target tissue with known drug delivery data, information on drug
administration and the patient-specific data, for each of the
plurality of individuals with known drug delivery data from whom
patient-specific models were received; associating the feature
vectors with the received drug delivery data at one or more
locations of the target tissue, for each of the plurality of
individuals with known drug delivery data from whom
patient-specific models were received; and training a system that
can predict drug delivery data at one or more locations on a
patient-specific model of a target tissue and of a vascular network
supplying blood to the target tissue, from one or more
patient-specific data and information on drug administration, using
the associated feature vectors.
36. The system of claim 35, further comprising: receiving a
patient-specific model of the target tissue where drug is delivered
and of the vascular network supplying blood to the target tissue of
a patient seeking analysis; receiving information on drug
administration in the patient seeking analysis; receiving data on
one or more locations in the target tissue of the patient seeking
analysis where drug delivery data will be estimated or measured;
deriving one or more patient specific data from the
patient-specific model and/or the patient seeking analysis; forming
feature vectors comprising of information regarding the one or more
locations on the target tissue where drug delivery data will be
estimated or measured, information on drug administration and the
patient-specific data; applying the trained system to estimate the
drug delivery data at one or more locations on the target tissue of
the patient-specific model of the patient seeking analysis, using
the feature vectors; and outputting the estimated drug delivery
data to an electronic storage medium or a display.
37. The system of claim 35, further comprising: developing a
patient-specific fluid dynamics model of blood flow in the
patient-specific models, using patient-specific data, for each of
the plurality of individuals with known drug delivery data at one
or more locations of the target tissue; determining the
transportation, spatial, and/or temporal distribution of the drug
particles in the fluid dynamics model of blood flow, using the
information on drug administration, for each of a plurality of
individuals with known drug delivery data at one or more locations
of the target tissue; and incorporating information related to the
transportation, spatial, and/or temporal distribution of the drug
particles into the feature vectors, for each of a plurality of
individuals with known drug delivery data at one or more locations
of the target tissue.
38. A non-transitory computer readable medium for use on a computer
system containing computer-executable programming instructions for
estimating drug delivery at a target tissue, the method comprising:
receiving one or more patient-specific models of (1) a target
tissue where a drug is delivered, and (2) of a vascular network
supplying blood to the target tissue from each of a plurality of
individuals with known drug delivery data at one or more locations
of the target tissue; receiving information on drug administration
for each of the plurality of individuals with known drug delivery
data from whom patient-specific models were received; deriving one
or more patient specific data from the patient-specific models from
whom patient-specific models were received, the patient-specific
data including one or more physiological conditions; receiving drug
delivery data for one or more locations of the target tissue for
each of the plurality of individuals with known drug delivery data
from whom patient-specific models were received; calculating
feature vectors including information regarding the one or more
locations on the target tissue with known drug delivery data,
information on drug administration and the patient-specific data,
for each of the plurality of individuals with known drug delivery
data from whom patient-specific models were received; associating
the feature vectors with the received drug delivery data at one or
more locations of the target tissue, for each of the plurality of
individuals with known drug delivery data from whom
patient-specific models were received; and training a system that
can predict drug delivery data at one or more locations on a
patient-specific model of a target tissue and of a vascular network
supplying blood to the target tissue, from one or more
patient-specific data and information on drug administration, using
the associated feature vectors.
39. The non-transitory computer readable medium of claim 38,
further comprising: receiving a patient-specific model of the
target tissue where drug is delivered and of the vascular network
supplying blood to the target tissue of a patient seeking analysis;
receiving information on drug administration in the patient seeking
analysis receiving data on one or more locations in the target
tissue of the patient seeking analysis where drug delivery data
will be estimated or measured; deriving one or more patient
specific data from the patient-specific model and/or the patient
seeking analysis; forming feature vectors comprising of information
regarding the one or more locations on the target tissue where drug
delivery data will be estimated or measured, information on drug
administration and the patient-specific data; applying the trained
system to estimate the drug delivery data at one or more locations
on the target tissue of the patient-specific model of the patient
seeking analysis, using the feature vectors; and outputting the
estimated drug delivery data to an electronic storage medium or a
display.
40. The non-transitory computer readable medium of claim 38,
further comprising developing a patient-specific fluid dynamics
model of blood flow in the patient-specific models, using
patient-specific data, for each of the plurality of individuals
with known drug delivery data at one or more locations of the
target tissue; determining the transportation, spatial, and/or
temporal distribution of the drug particles in the fluid dynamics
model of blood flow, using the information on drug administration,
for each of a plurality of individuals with known drug delivery
data at one or more locations of the target tissue; and
incorporating information related to the transportation, spatial,
and/or temporal distribution of the drug particles into the feature
vectors, for each of a plurality of individuals with known drug
delivery data at one or more locations of the target tissue.
Description
RELATED APPLICATION(S)
[0001] This application claims priority to U.S. Provisional
Application No. 62/219,490 filed Sep. 16, 2015, the entire
disclosure of which is hereby incorporated herein by reference in
its entirety.
FIELD OF THE DISCLOSURE
[0002] Various embodiments of the present disclosure relate
generally to drug delivery assessment, treatment planning, and
related methods. More specifically, particular embodiments of the
present disclosure relate to systems and methods for
patient-specific imaging and modeling of drug delivery, e.g., for
chemotherapy.
BACKGROUND
[0003] Cancer affects millions of people worldwide and is one of
the most common causes of death. Chemotherapy is one form of cancer
treatment in which patients are given one or more drugs as part of
a standardized regimen. The efficacy of chemotherapy may be
influenced by the effectiveness of drug delivery to target organs
in patients. The effectiveness of drug delivery may be shaped by
factors including, but not limited to, the drug amount, drug
concentration, administration location, frequency of dosage, type
of therapy, and/or patient characteristics. For example,
determining the optimal dose of drug agent may be important since
over-dosing can cause serious side effects due to drug toxicity,
and under-dosing can lead to reduced effectiveness of therapy.
[0004] Current methods of drug delivery, such as those using body
surface area (BSA), are inaccurate because they do not account for
inter-patient variations. For example, there is a fourfold to
tenfold variation in cytotoxic drug clearance between individuals
due to differing activity of drug elimination processes related to
genetic and environmental factors. Thus, there is a desire for a
drug delivery system and method that is personalized, in order to
account for inter-patient variation, and accurate, in order to
minimize toxicities and help to improve treatment outcomes of
chemotherapy.
[0005] Furthermore, having an effective personalized and accurate
drug delivery system and method may also improve treatments of
other ailments. Coronary artery disease is a common ailment that
may cause blood vessels supplying blood to the heart to develop
lesions, such as a stenosis (abnormal narrowing of a blood vessel).
The presence or absence of stenosis, thrombosis and other
circulatory conditions can affect blood flow characteristics along
with drug delivery patterns. One of the treatments for coronary
artery disease, percutaneous coronary intervention, involves
treating the stenotic (narrowed) coronary arteries of the heart.
While percutaneous coronary intervention runs the risk of
generating a reappearance of stenosis (restenosis), an effective
delivery of drug agents for percutaneous coronary intervention can
help reduce restenosis rates. Thus, there is a desire for a system
and method to assess the effectiveness of drug delivery by
evaluating convection, diffusion, and/or metabolism rates
throughout the circulatory system based on the blood flow
characteristics.
[0006] The foregoing general description and the following detailed
description are exemplary and explanatory only and are not
restrictive of the disclosure.
SUMMARY
[0007] According to certain aspects of the present disclosure,
systems and methods are disclosed for providing personalized
chemotherapy and drug delivery by using computational fluid
dynamics and/or machine learning from a vascular anatomical
model.
[0008] One method includes: receiving a patient-specific anatomical
model of at least one vessel of the patient and a target tissue
where a drug is to be supplied; receiving patient-specific
information defining the administration of a drug; identifying one
or more locations in the target tissue where drug delivery data
will be estimated or measured; deriving patient-specific data from
the patient specific anatomical model and/or the patient;
determining one or more blood flow characteristics in a vascular
network leading to the one or more locations in the target tissue
where drug delivery data will be estimated or measured, using the
patient-specific anatomical model and the patient-specific data;
determining the transportation, spatial, and/or temporal
distribution of the drug particles in one or more locations in the
vascular network using the patient-specific information defining
the administration of the drug; and computing drug delivery data at
the one or more locations in the target tissue using the
transportation, spatial, and/or temporal distribution of the drug
particles.
[0009] In accordance with another embodiment, a system is disclosed
for estimating drug delivery at a target tissue. The system
comprises: a data storage device storing instructions for
estimating drug delivery at a target tissue; and a processor
configured for: receiving a patient-specific anatomical model of at
least one vessel of the patient and a target tissue where a drug is
to be supplied; receiving patient-specific information defining the
administration of a drug; identifying one or more locations in the
target tissue where drug delivery data will be estimated or
measured; deriving patient-specific data from the patient specific
anatomical model and/or the patient; determining one or more blood
flow characteristics in a vascular network leading to the one or
more locations in the target tissue where drug delivery data will
be estimated or measured, using the patient-specific anatomical
model and the patient-specific data; determining the
transportation, spatial, and/or temporal distribution of the drug
particles in one or more locations in the vascular network using
the patient-specific information defining the administration of the
drug; and computing drug delivery data at the one or more locations
in the target tissue using the transportation, spatial, and/or
temporal distribution of the drug particles.
[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 estimating drug
delivery at a target tissue is provided. The method includes:
receiving a patient-specific anatomical model of at least one
vessel of the patient and a target tissue where a drug is to be
supplied; receiving patient-specific information defining the
administration of a drug; identifying one or more locations in the
target tissue where drug delivery data will be estimated or
measured; deriving patient-specific data from the patient specific
anatomical model and/or the patient; determining one or more blood
flow characteristics in a vascular network leading to the one or
more locations in the target tissue where drug delivery data will
be estimated or measured, using the patient-specific anatomical
model and the patient-specific data; determining the
transportation, spatial, and/or temporal distribution of the drug
particles in one or more locations in the vascular network using
the patient-specific information defining the administration of the
drug; and computing drug delivery data at the one or more locations
in the target tissue using the transportation, spatial, and/or
temporal distribution of the drug particles.
[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 on 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 detailed
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 providing personalized chemotherapy and drug delivery,
according to an exemplary embodiment of the present disclosure.
[0015] FIG. 2 is a block diagram of an exemplary method of
estimating drug delivery at a target tissue.
[0016] FIG. 3 is a block diagram of an exemplary method of
determining drug delivery data at one or more locations in the
target tissue using patient-specific data using computational fluid
dynamics.
[0017] FIG. 4 is a block diagram of an exemplary method of
determining drug delivery data at one or more locations in the
target tissue by training a machine learning algorithm using
patient-specific data.
[0018] FIG. 5 is a block diagram of an exemplary method of using
the system and method described in method 200 to regulate drug
administration until the actual drug delivery data matches, or
falls within the range of, a desired drug delivery data.
[0019] FIG. 6 is a block diagram of an exemplary method of using
the system and method described in method 200 for simulating
changes in drug delivery towards a target tissue by assessing the
effectiveness of the drug delivery.
DESCRIPTION OF THE EMBODIMENTS
[0020] Reference will now be made in detail to the exemplary
embodiments of the disclosure, 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.
[0021] The efficacy of chemotherapy may be influenced by the
effectiveness of drug delivery to target organs in patients. For
the purposes of the disclosure: "patient" may refer to any
individual or person for whom the effectiveness of drug delivery is
being assessed, or any individual or person associated with the
drug delivery assessment of one or more individuals. Determining
the optimal dose of drug agent may be important; however,
inter-patient variation in drug handling and a lack of accurate
methods may pose challenges for determining the optimal dose. A
personalized approach by monitoring drug levels in blood plasma and
adjusting dose can minimize toxicities and help to improve
treatment outcomes of chemotherapy. Moreover, effective delivery of
drug agents for percutaneous coronary intervention can help reduce
restenosis rates.
[0022] Given the potentially wide scope of this problem and
utility, assessing the effectiveness of drug delivery can help
clinicians determine treatment strategies for patients by
evaluating drug delivery patterns along with convection, diffusion,
and/or metabolism rates throughout the circulatory system based on
the flow characteristics. The present disclosure describes systems
and methods for providing personalized chemotherapy and drug
delivery modeling using computational fluid dynamics and medical
imaging with machine learning from a vascular/anatomical model.
While the following embodiments may be directed to brain tumor and
coronary lesions, the same system and method could be applied to
creating patient-specific models of chemotherapy and drug delivery
in other oncological diseases, including breast cancer, prostate
cancer, liver cancer, colon cancer, lung cancer, and/or cervical
cancer. Furthermore, the framework could be extended to assess
patient-specific models of implantable drug delivery devices,
including nanoparticles and hydrogels. In some cases, a "particle"
may refer to a small unit of drug (e.g., molecule) that may be
released into the patient's blood stream or transported to the
target tissue.
[0023] The steps described in the methods may be performed in any
order, or in conjunction with any other step. It is also
contemplated that one or more of the steps may be omitted for
performing the methods described in the present disclosure. In
general, FIG. 1 provides depicts an overview of a system and
network of the current disclosure: FIG. 2 illustrates a general
embodiment of a method for estimating drug delivery in the target
tissue; FIG. 3 and FIG. 4 illustrate step 210 within method 200
disclosed in FIG. 2, in more detail; and FIGS. 5 and 6 expand the
general embodiment of FIG. 2 into a personalized chemotherapy and
drug delivery system and method.
[0024] Referring now to the figures, FIG. 1 depicts a block diagram
of an exemplary system 100 and network for providing personalized
chemotherapy and drug delivery, according to an exemplary
embodiment. Specifically, FIG. 1 depicts a plurality of physicians
102 and third party providers 104, any of whom may be connected to
an electronic network 101, 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' anatomy. The physicians 102 and/or
third party providers 104 may also obtain any combination of
patient-specific data, such as medical image characteristics,
anatomical characteristics, perfusion territories, blood supply to
the target tissue, blood flow characteristics, patient
characteristics, disease burden characteristics, and/or
electromechanical measurements. Physicians 102 and/or third party
providers 104 may transmit the patient-specific data to server
systems 106 over the electronic network 101. 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. For purposes of disclosure,
"electronic storage devices" or "electronic storage media" may
include, but are not limited to, a hard drive, network drive, cloud
drive, mobile phone, tablet, or the like that may or may not be
affixed to a display screen.
[0025] FIG. 2 depicts a general embodiment of an exemplary method
200 for estimating drug delivery at a target tissue. One or more
steps of method 200 may be performed by a processor of server
systems 106. In one embodiment, step 202 may include receiving a
patient specific vascular anatomical model in an electronic storage
medium of the server system 106. Specifically, receiving the
patient-specific anatomical model may include either generating the
patient-specific anatomical model at the server system 106, or
receiving one over an electronic network (e.g., electronic network
100). The patient specific vascular anatomical model may include,
but is not limited to, the patient's cerebrovascular system,
cardiovascular system, and/or the vasculature perfusing the
breasts, prostate, liver, colon, lung, and/or cervix. In one
embodiment, the vascular anatomical model may be derived from
images of the person acquired using one or more available imaging
or scanning modalities (e.g., computed tomography (CT) scans,
magnetic resonance (MR) imaging, micro-computed tomography (.mu.CT)
scans, micro-magnetic resonance (.mu.MR) imaging, dual energy
computed tomography scans, ultrasound imaging, single photon
emission computed tomography (SPECT) scans, or positron emission
tomography (PET) scans). The vascular anatomical model may be
obtained through segmentation of an imaging study, including, but
not limited to, images obtained from one or more said available
imaging or scanning modalities.
[0026] Step 204 may include receiving a patient specific model of
the target tissue in an electronic storage medium. For purposes of
disclosure, a "target tissue" may refer to the tissue and/or organ
in which the blood supply and/or drug delivery data may be
estimated. In one embodiment, the target tissue may be found in an
organ afflicted with cancerous growth, including, but not limited
to the brain, one or more breasts, the prostate, liver, colon,
lung, and/or cervix. In one embodiment, the target tissue may be
found in vasculature affected with thrombosis, including but not
limited to the coronary, aortic, or cerebrovascular systems,
peripheral vasculature perfusing one or more muscles, renal
vasculature supplying the kidney, and/or visceral vasculature
supplying the bowels, liver, and/or spleen. In one embodiment, the
patient specific model of the target tissue may be obtained through
segmentation of an imaging study, including, but not limited to,
images obtained from one or more said available imaging or scanning
modalities.
[0027] Step 206 may include receiving patient-specific drug
administration information. In one embodiment, drug administration
information may include one or more drug administration locations
and the respective drug amounts inserted at the one or more drug
administration locations. In another embodiment, drug
administration information may include the administered drug
concentration, administration frequency, administration time, type
of therapy, and/or one or more routes of drug administration, with
said routes including, but not limited to, oral administration,
intravenous administration, or direct administration into the tumor
and/or lesion.
[0028] Step 208 may include receiving or calculating one or more
elements of patient-specific data. In one embodiment, the one or
more patient-specific data elements may be extracted from the
patient specific vascular anatomical model and/or patient specific
model of a target tissue. In one embodiment, the patient-specific
data may include, but are not limited to an estimate of one or more
anatomical characteristics of the vessel model and/or target tissue
model, for example, size (volume, surface area, mass, etc.), shape,
tortuosity, length, thickness, number and length of branches,
network topology, etc. In one embodiment, the patient-specific data
may include, but are not limited to an estimate of the supplied
blood to each area of the target tissue under one or more
physiological states, and/or one or more blood flow
characteristics, for example, fractional flow reserve (FFR), flow
magnitude and direction, etc. The blood flow characteristics may be
determined through several means, for instance, invasive
measurement (e.g., through invasive FFR, thrombolysis in myocardial
infarction (TIMI), microspheres, etc.), calculation using a blood
flow simulation (e.g., a 3D or 1D fluid simulation model,
calculation, transluminal attenuation flow encoding (TAFE), etc.),
calculation using imaging characteristics (e.g., transluminal
arterial gradient (TAG), corrected coronary opacification (CCO),
etc.) and/or calculation using a machine learning estimation of
blood supply based on anatomical or imaging features.
[0029] The patient-specific data may include, but are not limited
to an estimate of the perfusion or diffusion territories of the
target tissue related to the vascular model. This estimate may be
determined, for instance, using a nearest-neighbor (Voronoi)
approach to assigning locations in the target tissue to the closest
supplying vessel in the vascular model, a microvascular estimation
technique calculation from an anatomical model, including
constrained constructive optimization.
[0030] Furthermore, the patient-specific data may include, but are
not limited to an estimate of the perfusion or diffusion
territories of the drug delivery device related to the release of
particles into the blood flow. The estimate of the perfusion or
diffusion territories may be determined by an exemplary method of
using a thermodynamic model and mass transport models, including
energy conservation or Fick's Laws of Diffusion, to simulate drug
release assuming ionic equilibrium. In one embodiment, the estimate
of the perfusion or diffusion territories may be determined by an
exemplary method of using Multiphase Mixture Theory to simulate
drug release driven by mechanical, chemical, or electrochemical
potentials in response to shear, compression, or expansion
flow.
[0031] Furthermore, the patient-specific data may include drug
delivery device characteristics, including but not limited to,
device size, particle properties (e.g., size, concentration,
molecular weight, etc.), hydrogel polymer material properties
(e.g., elastic modulus, cross-link density, solubility, porosity,
matrix swelling, etc.), and/or operating parameters (e.g.,
temperature, solvent quality, pH, charge density, etc.), etc.
[0032] Furthermore, the patient-specific data may include, but are
not limited to, medical images (e.g., a CT, MR, .mu.CT, .mu.MR,
dual energy CT, ultrasound, PET, SPECT, etc.) in one or more
physiological states (e.g., rest, stress, etc.) of the target
tissue and/or vessels represented by the vascular model. Image
characteristics of the target tissue or vessels may be received or
calculated in one or more locations, including, but not limited to,
local average intensities at one or more resolutions, differences
of average intensities (e.g., calculated using wavelet bases such
as Haar), texture characteristics (e.g., Haralick texture
features), any standard image features, including histograms of
gradients, SIFT, steerable filters, etc.
[0033] Furthermore, the patient-specific data may include, but are
not limited to, patient characteristics, such as age, gender,
smoking history, height, weight, body surface area (BSA), diabetic
status, hypertensive status, ethnicity, family history, and/or
genetic information.
[0034] Furthermore, the patient-specific data may include, but are
not limited to, vascular or target tissue disease characteristics,
including tumor size, degree of malignancy, location of tumor in
the brain, tumor blood flow, oxygen transport, vascular endothelial
growth factor distribution, extracellular pH, interstitial fluid
pressure, interstitial fluid velocity, vascular permeability,
tissue transport properties, angiogenic parameters (e.g., tumor
blood volume), plaque burden, presence of adverse plaque
characteristics (e.g., spotty calcification, low attenuation
plaque, napkin ring sign, positive remodeling, thrombus formation,
etc.), calcium score (patient-level or vessel-level), perfusion
information, ejection fraction, wall motion, wall thickness, wall
morphology, wall histology, etc.
[0035] Furthermore, the patient-specific data may include, but are
not limited to electromechanical measurements, including ECG
measurements or invasive EP measurements.
[0036] Step 210 may include determining one or more drug delivery
data at one or more locations in the target tissue using one or
more of the patient-specific data. In one embodiment, the drug
delivery data may include an estimate of drug delivery at one or
more locations in the target tissue. The determination of drug
delivery data may be performed by computational simulations of
tumor vasculature hemodynamics (e.g., FIG. 3) and/or training a
machine learning algorithm using a database of patients with known
drug delivery (e.g. FIG. 4).
[0037] Step 212 may include outputting the estimated drug delivery
data to an electronic storage medium and/or a display medium. In
one embodiment, the drug delivery estimates may be displayed in
greyscale or color, in 2D or 3D, overlaid on the anatomical model
of the target tissue, and/or overlaid on an image of the target
tissue.
[0038] FIG. 3 depicts an exemplary embodiment of method 300 for
performing step 210 of FIG. 2, which may include determining drug
delivery data at one or more locations in the target tissue by
performing computational simulations of tumor vasculature
hemodynamics on patient-specific data. In one embodiment, method
300 may include using Navier-Stokes equation, computational fluid
dynamics (or approximations thereof), and or
reaction-advection-diffusion equations to determine blood flow
characteristics, and then using drug administration information and
blood flow characteristics to estimate the transportation, spatial,
and/or temporal distribution of the drug. For purposes of
disclosure, "drug administration information" may refer to details
regarding the administration of a drug or therapy to a patient,
including, but not limited to, administered drug amount,
administered drug concentration, administration location,
administration frequency, route of drug administration,
administration time, type of therapy, etc. One or more steps of
method 300 may be performed by a processor of server systems
106.
[0039] In one embodiment, the drug delivery data in the one or more
locations in the target tissue may be estimated or calculated from
the transportation, spatial, and/or temporal distribution of the
drug. In another embodiment, the transportation, spatial, and/or
temporal distribution of the drug may be used to train a machine
learning algorithm for estimating or calculating the drug delivery
data in the one or more locations in the target tissue. The method
of FIG. 3 may be performed by server systems 106, based on
patient-specific data received from physicians 102 and/or third
party providers 104 over the electronic network 100. For purposes
of disclosure, "drug delivery data" may refer to an estimate of the
amount of drug delivered to the one or more locations in the target
tissue, an estimate of the concentration of drug particles
delivered to the one or more locations in the target tissue; the
circulatory destination probability of the drug particles released
at the drug delivery location to the target tissue, which may be
defined as the ratio of the amount of drug particles reaching the
target tissue with respect to the total number of released drug
particles, the transportation, spatial, and/or temporal
distribution of the drug, an estimate of the blood flow data; or a
combination thereof.
[0040] Step 302 may include computing the blood flow
characteristics in the patient specific vascular anatomy. In one
embodiment, computing the blood flow data may include solving the
Navier-Stokes equation, or a modified Navier-Stokes equation with
Darcy's law term for a flow through a porous medium, numerically
under the patient's physiologic conditions (e.g., hyperemic or rest
state, interstitial fluid pressure, tumor growth state, etc.). In
one embodiment, the blood flow data may include the velocity field
of blood flow in the heart, the coronary, cerebral, carotid
vasculature, the aortic arch, or any other vascular network. In one
embodiment, the blood flow data may include the velocity field of
venous circulation and/or micro-circulation, as well as arterial
circulation.
[0041] Step 304 may include inserting the drug amounts in the
received one or more drug administration locations. In one
embodiment, the drug amounts may be administered virtually so as to
simulate drug delivery.
[0042] Step 306 may include determining the transportation,
spatial, and/or temporal distribution of the drug particles. In one
embodiment, determining drug particle distribution may utilize an
ordinary differential equation of a form characteristic of
x'(t)=u(x,t); x(t.sub.0)=x.sub.0, using an appropriate numerical
method, where u(x,t) may be the velocity field and x(t) may be the
location of the particle at time t. The size and amount of
particles may be determined as a prescribed dosage of administered
drug. In another embodiment, determining drug particle distribution
may utilize an equation describing drug transport, which may be
described as
.differential. C F .differential. t + v F .gradient. C F =
.gradient. D .times. .gradient. C F , ##EQU00001##
where C.sub.F may be the drug concentration in the fluid, t may be
time, v.sub.F may be fluid velocity, and D may be the diffusion
tensor with the effective diffusivity factors of the drug in the
fluid. In one embodiment, varying mechanical and transport
properties can be assigned to each of the three layers (e.g.,
intima, media, adventitia) in the vessel wall. Non-Newtonian
rheological properties may also be considered for tumor
microvasculature hemodynamics.
[0043] Step 308 may include determining the drug delivery data in
the one or more locations in the target tissue. In one embodiment,
the drug delivery data may be determined from the transportation,
spatial, and/or temporal distribution of drug particles, as
determined in step 306. In one embodiment, the drug delivery data
may be determined from training a machine learning algorithm that
utilizes the transportation, spatial, and/or temporal distribution
of drug particles (e.g., method 400). In one embodiment, the drug
delivery data may include the circulatory destination probability
of drug molecules in the one or more locations. The circulatory
destination probability may be computed from the ratio of the
amount of particles reaching a location in the target tissue with
respect to the total number of released particles. In one
embodiment, step 308 may be performed by a processor.
[0044] Step 310 may include outputting the estimated drug delivery
data to an electronic storage medium and/or display medium. In one
embodiment, the drug delivery data may include the transportation,
spatial, and/or temporal distribution of the drug. In one
embodiment, the drug delivery estimates may be displayed in
greyscale or color, in 2D or 3D, overlaid on the anatomical model
of the target tissue, and/or overlaid on an image of the target
tissue.
[0045] FIG. 4 depicts an exemplary embodiment of method 400 for
performing step 210 of FIG. 2, which may include determining drug
delivery data at one or more locations in the target tissue by
training a machine learning algorithm on patient-specific data.
Alternately, method 400 may be performed subsequent to method 300
as a means to complete step 210 of FIG. 2, which may include using
computational simulations of tumor vasculature hemodynamics and
training a machine learning algorithm on patient-specific data in
order to determine drug delivery data.
[0046] In one embodiment, method 400 may include determining drug
delivery data at one or more locations in the target tissue by
training a machine learning algorithm on patient-specific data. The
method of FIG. 4 may be performed by server systems 106, based on
patient-specific data received from physicians 102 and/or third
party providers 104 over the electronic network 100. One or more
steps of method 400 may be performed by a processor.
[0047] In one embodiment, step 402 may include assembling a
database containing one or more of the patient-specific data at one
or more locations with the measured drug delivery data at those
locations. For example, step 402 may include assembling a database
of patients with known patient-specific data and known drug
delivery data. The one or more patient-specific data may include a
numerical description of physiological or phenotypic parameters of
the patient and/or a description of the local geometry and
biophysical characteristics at one or more locations. In one
embodiment, the one or more patient-specific data may include blood
flow data, transportation, spatial, and/or temporal distribution of
drug particles, and/or circulatory destination probability of drug
particles obtained from method 300. The measured or known drug
delivery data at one or more locations may include, for example,
one or more combinations of MR image, fluorodeoxyglucose positron
emission tomography (FDG-PET) image, stress echo/MRI contractile
reserve, multidetector CT, dual energy CT, .mu.CT, .mu.MR, etc.
[0048] Step 404 may include training a machine algorithm to map the
patient-specific data for one or more locations to the drug
delivery data at those locations. In one embodiment, the
patient-specific data and drug delivery data may be obtained from a
database of patients with known patient-specific data and known
drug delivery data. The one or more patient-specific data may
include a numerical description of physiological or phenotypic
parameters of the patient and/or a description of the local
geometry and biophysical characteristics at one or more locations.
Furthermore, one or more patient-specific data may include blood
flow data, transportation, spatial, and/or temporal distribution of
drug particles, and/or circulatory destination probability of drug
particles obtained from method 300. The machine learning algorithm
may take many forms, for example, one or more algorithms
implementing a multilayer perceptron, deep learning, support vector
machines, random forests, k-nearest neighbors, Bayes networks,
etc.
[0049] Step 406 may include applying the trained machine learning
algorithm to the new patient-specific data to estimate the drug
delivery data at one or more locations in the target tissue. In one
embodiment, the machine learning algorithm would be trained with a
database of patients with known patient-specific data and known
drug delivery data, and will be applied to the patient-specific
data of a new patient to estimate the drug delivery data at one or
more locations of the new patient's body.
[0050] In one embodiment, step 408 may include outputting, to an
electronic storage medium and/or display screen, the estimated drug
delivery data at one or more locations in the target tissue. The
drug delivery data may include the amount of drug delivered to one
or more locations on the target tissue. The output drug delivery
data may be displayed in greyscale or color in 2D or 3D, overlaid
on the anatomical model of the target tissue, and/or overlaid on an
image of the target tissue. In the output model, one or more
locations in the target tissue may be associated with a circulatory
destination probability of drug particles, which may be defined as
the ratio of the amount of drug particles reaching the target
tissue with respect to the total number of released drug particles.
In one embodiment, one or more locations in the target tissue may
be associated with the transportation, spatial, and/or temporal
distribution of the drug delivered.
[0051] FIG. 5 depicts an exemplary embodiment of method 500 for
using the system and method described in method 200 to regulate
drug administration until the amount of drug delivered matches, or
falls within the range of, a desired drug delivery amount. One or
more steps of method 500 may be performed using a processor of
server systems 106.
[0052] In one embodiment, step 502 may include receiving
information on the desired drug delivery data for one or more
locations in the target tissue. For example, step 502 may include
receiving information on the desired amount of drug to be delivered
to one or more locations on a tumor or lesion.
[0053] Step 504 may include receiving patient-specific drug
administration information. For example, drug administration
information may include one or more drug administration locations
and the respective drug amounts inserted at the one or more drug
administration locations. In another example, drug administration
information may also include one or more of the administered drug
concentration, administration frequency, administration time, type
of therapy, and/or one or more routes of drug administration, with
said routes including, but not limited to, oral administration,
intravenous administration, or direct administration into the tumor
and/or lesion.
[0054] Step 506 may include determining the drug delivery data at
one or more locations on the target tissue by using computed
simulations of tumor vasculature hemodynamics (e.g., FIG. 3) and/or
by applying a trained machine learning algorithm (e.g., FIG. 4).
For example, step 506 may include determining the amount of drug
delivered to the one or more locations on the target tissue. In one
embodiment, step 506 may be performed by a processor.
[0055] In one embodiment, subsequent to step 506, step 508 may
include determining whether the drug delivery data at one or more
locations on the target tissue matches, or falls within the range
of, the desired drug delivery data at those locations. For example,
step 508 may include determining whether the amount of drug
delivered to a location on the target tissue matches the desired
amount of drug to be delivered to that location. Step 508 may be
performed by a processor.
[0056] If, subsequent to step 508, the actual drug delivery data
does not match or does not fall within the range of the desired
drug delivery data for one or more locations on a target tissue,
then, in one embodiment, step 510 may include adjusting the drug
administration accordingly, and repeating step 504. For example, if
the amount of drug delivered is greater than the desired amount of
drug to be delivered, then the administered amount may be decreased
or maintained. In another example, if the amount of drug delivered
is less than the desired amount of drug to be delivered, then the
administered amount may be increased or maintained. In one
embodiment, if the amount of drug delivered does not match or does
not fall within the range of the desired amount of drug to be
delivered, then factors other than the administration amount may be
adjusted. These said factors may include, but are not limited to,
the one or more locations of the drug administration, the drug
concentration, the route of drug administration, the type of
therapy, and/or the frequency of drug insertion, etc.
[0057] If, subsequent to step 508, the amount of drug delivered
matches or falls within the range of the desired amount of drug to
be delivered, then, in one embodiment, step 512 may include
outputting drug delivery data to an electronic storage medium
and/or display medium. In one embodiment, the drug delivery data
may include the amount of drug delivered to one or more locations
on the target tissue. The output drug delivery data may be
displayed in greyscale or color in 2D or 3D, overlaid on the
anatomical model of the target tissue, and/or overlaid on an image
of the target tissue.
[0058] FIG. 6 depicts an exemplary embodiment of method 600 for
using the system and method described in method 200 for simulating
changes in tumor blood flow by assessing effectiveness of the drug
delivery or therapy. One or more steps of method 600 may be
performed using a processor of server systems 106.
[0059] Step 602 may include receiving one or more medical images at
a given time, t1. The one or more medical images may be from one or
more available scanning modalities. In one embodiment, the one or
more medical images may be obtained using segmentation of an
imaging study, including, but not limited to, images obtained from
one or more said available imaging or scanning modalities.
[0060] Step 604 may include receiving one or more medical images at
a given time, t2. The one or more medical images may be from one or
more available scanning modalities. In one embodiment, the one or
more medical images may be obtained using segmentation of an
imaging study, including, but not limited to, images obtained from
one or more said available imaging or scanning modalities. The
segmentation of the images may be performed by a processor.
[0061] Step 606 may include comparing the tumor, lesion,
patient-specific data, and/or drug delivery data extracted from the
one or more medical images from different times. For example, the
comparison may include determining whether intensity gradients
between medical images from different times have a difference that
is within a predetermined threshold.
[0062] Step 608 may include assessing the effectiveness of the
current drug delivery system, subsequent to the comparison in step
606. In one embodiment, an assessment on the effectiveness of the
current drug delivery system may include determining the status
change of the tumor or lesion. For example, the status change of
the tumor or lesion may be classified into one or more levels of
progression and regression, and the effectiveness of the drug
delivery system may be correlated with the level of regression in
the tumor or lesion. The status change may be determined by
comparing the images received at step 602 and step 604. Determining
the status change of the tumor or lesion may also be aided by
patient-specific data and/or drug delivery data.
[0063] If, subsequent to step 608, the current drug delivery system
is deemed to be insufficiently effective, and/or the status of the
tumor or lesion has progressed, then, in one embodiment, step 610
may include using the system described in method 200 to increase or
maintain the drug administration amount, and then repeating step
602. In one embodiment, step 610 may include adjusting factors
other than the drug administration amount. These said factors may
include, but are not limited to, the one or more locations of the
drug administration, the drug concentration, the route of drug
administration, the type of therapy, and/or the frequency of drug
insertion, etc.
[0064] If, subsequent to step 608, the current drug delivery system
is deemed to be effective, and/or the status of the tumor or lesion
has regressed, then, in one embodiment, step 612 may include using
the system described in method 200 to maintain or decrease the drug
administration amount, and then repeating step 602. In one
embodiment, step 612 may include adjusting factors other than the
drug administration amount. These said factors may include, but are
not limited to, the one or more locations of the drug
administration, the drug concentration, the route of drug
administration, the type of therapy, and/or the frequency of drug
insertion, etc.
[0065] In one embodiment, subsequent to step 608, step 614 may
include outputting the one or more medical images, tumor or lesion
status, and/or drug delivery data to an electronic storage medium
and/or display. In one embodiment, the one or more medical images,
tumor or lesion status, and/or drug delivery data may be displayed
in greyscale or color, in 2D or 3D, overlaid on the anatomical
model of the target tissue, and/or overlaid on an image of the
target tissue.
[0066] 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.
* * * * *