Patent Application
20210027466
The present invention relates generally to representing a patient at a particular physiological state and more particularly to mapping patient data from one physiological state to another physiological state.
For clinical diagnosis, therapy planning, and prognosis, it is of interest to not only measure or compute data for a patient during a particular physiological state, but also to compare data for a patient at two or more different physiological states. For example, it is of interest to compare the role of a stenosis (in terms of pressure drop) when a patient is under different physiological conditions. Currently, the comparison of patient data at different physiological states can be achieved by either directly measuring the data of interest at each physiological state or by simulating each physiological state using appropriate boundary conditions for the model. However, these current solutions have drawbacks. The repeated measurement of the same data at different physiological states is not only costly, but requires an additional clinical procedure which may expose the patient to higher risk. In addition, in some instances, the measurement of data at different physiological states may not be practical since the primary purpose of the comparison would be to plan a therapy a priori and not just to assess it after it is administered. Similarly, the simulation of patient data for each physiological state using biophysical models may be time-consuming, resource-intensive, and prone to inaccuracies due to the difficulty in obtaining correct boundary conditions that represent a particular physiological state.
In accordance with an embodiment, systems and methods for determining a quantity of interest of a patient comprise receiving patient data of the patient at a first physiological state. A value of a quantity of interest of the patient at the first physiological state is determined based on the patient data. The quantity of interest represents a medical characteristic of the patient. Features are extracted from the patient data, wherein the features which are extracted are based on the quantity of interest to be determined for the patient at a second physiological state. The value of the quantity of interest of the patient at the first physiological state is mapped to a value of the quantity of interest of the patient at the second physiological state based on the extracted features.
In one embodiment, the quantity of interest of the patient at the first physiological state is mapped to the quantity of interest of the patient at the second physiological state without using data of the patient at the second physiological state. The quantity of interest of the patient at the first physiological state may be a same quantity of interest as the quantity of interest of the patient at the second physiological state. The quantity of interest of the patient at the first physiological state may alternatively be different from the quantity of interest of the patient at the second physiological state.
The quantity of interest of the patient at the first physiological state may be mapped to the quantity of interest of the patient at the second physiological state by applying a trained mapping function to the value of the quantity of interest of the patient at the first state. The trained mapping function may represent 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. The trained mapping function may be determined in an offline step. The trained mapping function may be a machine-learning based mapping function trained based on training data comprising quantities of interest of the set of patient at the first physiological state and corresponding quantities of interest of the set of patients at the second physiological state. The training data may comprise simulated quantities of interest of the set of patients at the first physiological state and simulated corresponding quantities of interest at the second physiological state.
In some embodiments, systems and methods are provided for determining fractional flow reserve (FFR) for a coronary stenosis of a patient at a hyperemia state. Patient data of the patient at a rest state is received. A value of a pressure drop over the coronary stenosis of the patient at the rest state is calculated based on the patient data. Features are extracted from the patient data. The value of the pressure drop over the coronary stenosis of the patient at the rest state is mapped to a value of the pressure drop over the coronary stenosis of the patient at the hyperemia state based on the extracted features. The FFR for the coronary stenosis of the patient outputted based on the pressure drop over the coronary stenosis of the patient at the hyperemia state.
These and other advantages of the invention will be apparent to those of ordinary skill in the art by reference to the following detailed description and the accompanying drawings.
The present invention relates to mapping patient data from one physiological state to another physiological state. Embodiments of the present invention are described herein to give a visual understanding of the method for mapping the patient data. A digital image is often composed of digital representations of one or more objects (or shapes). The digital representation of an object is often described herein in terms of identifying and manipulating the objects. Such manipulations are virtual manipulations accomplished in the memory or other circuitry/hardware of a computer system. Accordingly, is to be understood that embodiments of the present invention may be performed within a computer system using data stored within the computer system.
Referring to
Features 204 that are extracted from patient data at State 1 may depend on the QoI that is to be determined for the patient at State 2. Features 204 may include, e.g., anatomical features (e.g., geometry of a vessel), functional features (e.g., vital statistics, blood pressure, heart rate, oxygen saturation), or any other pertinent data. Anatomical features may be extracted from, e.g., medical image data of a patient at State 1, such as medical image data from magnetic resonance imaging (MRI), computed tomography (CT), X-ray angiography, ultrasound, or any other suitable imaging modality. For example, anatomical features such as a length of an anatomical structure may be determined by processing the medical image data to extract features 204. Functional features may be, e.g., measured invasively or non-invasively. Such measurements may be determined directly from the patient data to extract features 204. The particular features 204 that are extracted may be learned by mapping function 208 using training data 210.
For example, in the case of a cardiac disease patient, it is of interest to determine cardiac output as the QoI of a patient at State 2 (e.g., an exertion state). Cardiac output is the amount of blood that a patient's heart pumps out. Features 204 that are extracted from patient data at State 1 (e.g., a rest state) are based on the QoI of cardiac output to be determined for a patent at State 2. Features at State 1204 for the QoI of cardiac output may include, e.g., the blood pressure at various chambers of the heart and the blood pressure in the aorta, which may be measured invasively or non-invasively as part of the patient data.
A value of QoI at State 1206 may also be extracted from the patient data at State 1. The value of QoI at State 1206 may be extracted from patient data at State 1 in a similar manner as features 204. For example, the patient data at State 1 may include medical image data, which may be processed to determine a value of the QoI of cardiac output of a patient at State 1. Cardiac output can be extracted from medical images by segmenting the heart chambers or by direct flow measurement using Doppler ultrasound, phase contrast MRI, or any other suitable technique. In another example, patient data at State 1 may include blood flow measurements (e.g., measured or simulated) to provide a value of the QoI of cardiac output of a patient at State 1.
In one embodiment, QoI at State 1 corresponds to a same quantity of interest desired at State 2. For instance, in the example of the cardiac disease patient above, where the quantity of interest at State 2 is the cardiac output at State 2, QoI at State 1206 is the cardiac output at State 1. In other embodiments, QoI at State 1206 may be used to determine a different QoI at State 2214. QoI at State 1206 may include a plurality of quantities of interest where there are multiple quantities of interest desired at State 2.
Mapping function 208 receives input 202 (comprising features at State 1204 and QoI at State 1206) to determine a value of QoI at State 2214 as output 212. The goal of mapping function 208 is to compute a value of the QoI (e.g., the cardiac output) if the patient were at some other physiological state (i.e., State 2), which is distinct from the physiological state from which measurements/computations were performed (i.e., State 1). For example, mapping function 208 may be used to compute a value of the QoI of cardiac output during treadmill exercises based on measurements/computations of the patient at rest without subjecting the patient to the treadmill exercises, and thereby without making any new measurements. In another example, mapping function 208 may be used to compute the QoI of cardiac output when the patient is under a drug induced hyperemia condition without actually administering the drug to the patient.
Mapping function 208 may be a trained mathematical transformation that maps QoI at State 1206 to QoI at State 2214. It should be understood that the mapping relies on patient data from State 1, and not on any measurements/computations of the patient at State 2. However, in some embodiments, where patient data of the patient at Stage 2 is available, this patient data of the patient at Stage 2 may also be used by the trained mapping function 208 to determine QoI at State 2214. Mapping function 208 may be represented as function ƒ in equation (1) below.
QoI(State 2)=ƒ(QoI(State 1),Features(State 1)) (1)
In equation (1), QoI(State 2) represents the quantity of interest at State 2214, which is represented as a function of the quantity of interest at State 1206 and features extracted from patient data at State 1204. As can be seen from equation (1), QoI at State 2214 depends only on measurements/computations from the patient at State 1.
The function ƒ in equation (1) may be any suitable function that can represent relationships between a QoI at State 1 and the QoI at State 2. For example, function ƒ may be determined using a machine learning based method, a parameter tuning approach, an optimization approach, a statistical approach, a data-driven approach, a control system based approach, etc.
In one embodiment, mapping function 208 is trained to learn relationships between the QoI at State 1 and the QoI at State 2 using training data 210 in an offline step, which may be performed at a previous point in time. Mapping function 208 learns which features from the patient data are relevant for mapping the QoI at State 1 and the QoI at State 2 from training data 210. The relationships learned between the QoI at State 1 and the QoI at State 2 may be for all patients, a particular subgroup of patients, or distinct for each individual (i.e., patient specific). Training data 210 may include, e.g., a patient database for which the QoI at State 1 and the corresponding QoI at State 2 are available. Training data 210 may include historical patient data (e.g., for all patients, a particular group of patients, a distinct patient), synthetic patient data (e.g., generated using simulations without any patient specific data), patient data obtained from bench-top experiments, etc.
For example, in one embodiment, mapping function 208 may employ a parameter estimation approach, which estimates parameters of the function ƒ (e.g., QoI at State 1 and features at Stage 1) for a QoI at State 2 using test data where corresponding QoI at State 2 is available.
In another example, mapping function 208 may include a machine learning based algorithm. Machine learning based algorithms include a training phase and a testing phase. The training phase represents the offline portion of the process, where function ƒ is determined by fitting input data (e.g., QoI at State 1 and patient data at State 1) to a ground truth (i.e., QoI at State 2) using training data. The training data includes the QoI at State 1 and a corresponding QoI at State 2 for a set of patients. Mapping function 208 may be determined using any machine learning based algorithm, such as, e.g., linear or non-linear regression, neural networks, support-vector machines, etc. The testing phase refers to the use of the trained machine learning based algorithm for an unseen case (i.e., a new case that was not used for training). In the testing phase, QoI at State 1206 and features at State 1204 are used by the trained machine learning based algorithm to compute QoI at State 2214.
In an additional example, mapping function 208 may include a control systems based approach. In this approach, mapping function 208 is represented as a transfer function which maps QoI at State 1206 (i.e., input) to the QoI at State 2214 (i.e., output), using features of patient data at State 1204 to model the system that is to be controlled. The function ƒ is obtained by designing a controller (using any control design algorithm) that maps the input to the output.
Mapping function 208 receives features of the patient at state 1204 and QoI at State 1206 as input 202 and provides QoI at State 2214 as output 212. It should be understood that the embodiments described herein are not limited to State 1 and State 2, but may also be applied for more than two states by, e.g., concatenating the mathematical transformation in mapping function 208 or by performing the mapping function 208 in a repeated manner. Additionally, multiple QoI at State 1 may be available and may each be mapped by mapping function 208 to a same or different mapping to their respective QoI at Stage 2. Advantageously, system 200 determines a quantity of interest of a patient at State 2 based on data of the patient at State 1 and without relying on measurements or computations of the patient at State 2. System 200 thus provides a user with fast access to information that can impact a diagnosis or a therapy decision.
At step 304, a value of a quantity of interest of the patient at the first physiological state is determined based on the patient data. In one embodiment, the quantity of interest represents a medical characteristic of the patient. The value of the quantity of interest of the patient at the first physiological state may be determined directly from the patient data or may be computed or calculated using the patient data. For example, the value of the quantity of interest of the patient at the first physiological state may include blood flow measurements that can be measured or simulated to provide the value of the quantity of interest of the patient at the first physiological state. In one embodiment, the quantity of interest of the patient at the first physiological state corresponds to a same quantity of interest to be determined for the patient at the second physiological state. In other embodiments, the quantity of interest of the patient at the first physiological state may be used to determine a different quantity of interest of the patient at the second physiological state.
At step 306, features are extracted from the patient data. The features which are extracted are based on the quantity of interest to be determined for the patent at the second physiological state. The features may include anatomical features and functional features. The anatomical features may be extracted from, e.g., medical image data of the patient. The functional features may be measured (e.g., invasively or non-invasively). For example, for a quantity of interest of cardiac output, the features extracted from the patient data may include the blood pressure at various chambers of the heart and the blood pressure of the aorta. The features in this example may be extracted by processing medical image data of the patient or determined directly from the patient data as measurements or simulation results. In another example, for a quantity of interest of pressure drop in an artery, the features extracted from the patient data may include the anatomical features such as the geometry of the vessel and functional features such as blood pressure, heart rate, oxygen saturation and other vital statistics of the patient. In a further example, for a quantity of interest of pressure drop before and after stenting, the features may include anatomical aspects of the vessel and properties of the stent (e.g., length, diameter).
At step 308, the quantity of interest of the patient at the first physiological state is mapped to the quantity of interest of the patient at the second physiological state based on the extracted features. The mapping may be performed by applying a trained mapping function on the extracted features of the patient at the first physiological state and the quantity of interest of the patient at the first physiological state to determine the quantity of interest of the patient at the second physiological state. The trained mapping function may be any function that can represent the relationships between the quantity of interest at the first state and the quantity of interest at the second state for a set of patients. For example, the trained mapping function may be trained using a machine learning based method, a parameter tuning approach, an optimization approach, a statistical approach, a data-driven approach, a control system based approach, etc. The trained mapping function may be trained in an offline step based on training data, such as, e.g., historical patient data, synthetic patient data, patient data obtained from bench-top experiments, etc.
In one embodiment, where there are multiple quantities of interest to be determined for a patient at the second state, steps 304 and 306 may be repeated for each quantity of interest.
The conventional approach to determine the pressure drop during hyperemia would be to either directly measure the pressure drop during hyperemia or use a computational fluid dynamics (CFD) algorithm that simulates a hyperemia (by employing appropriate boundary conditions in the CFD algorithm and using the results of the CFD computation to determine the pressure drop at hyperemia). However, both of these conventional approaches have drawbacks. The measurement process is invasive, costly, and inherently risky for the patient since it is an additional clinical procedure. The CFD based computation is non-invasive, but may be time intensive and inaccurate due to the difficulty in obtaining the correct boundary conditions that represent the hyperemia state for a particular patient.
Referring to
At step 404, a value of a pressure drop across a stenosis for the patient at the rest state is determined. The pressure drop may be any value indicating the drop in pressure across a, e.g., stenosis. For example, the pressure drop, ΔP, may be represented as a difference between pressure distal to the stenosis, Pd, and pressure proximal to the stenosis, Pa, such that ΔP=Pd−Pa. Pressure drop, ΔP, may also be represented as a ratio between Pd and Pa, such that ΔP=Pd/Pa. Other representations are also contemplated. Distal and Proximal pressure may be extracted from medical image data or determined directly from patient data as measurements or simulation results. For example, the pressure may be measured invasively by, e.g., inserting a pressure wire into the artery. The pressure may also be measured by computing a CFD algorithm. Rest state measurements can be measured more easily, safely, cheaply, and quickly than hyperemic measurements since they do not require the patient to be in a hyperemic state.
At step 406, features are extracted from the patient data. The extracted features are based on the QoI (i.e., pressure drop across a stenosis) to be determined for the patient at the hyperemia state. In this example, the features may include anatomical features, such as features representing the geometry of the vessel, (e.g., length of the stenosis, radius of the vessel at/before/after the stenosis). The anatomical features may be extracted by processing medical image data to determine the geometry of the vessel. The features may also include function features such as vital statistics (e.g., blood pressure, heart rate, and oxygen saturation) of the patient at the rest state. The functional features may be determined as measurements or simulation results directly from the patient data.
At step 408, the pressure drop across the stenosis for the patient at the rest state is mapped to a pressure drop across the stenosis for the patient at the hyperemia state based on the features extracted at step 406. The mapping determines the pressure drop across the stenosis for the patient at the hyperemia state based on the pressure drop across the stenosis for the patient at the rest state, without subjecting the patient to hyperemia, or performing any measurements or biophysical computations at hyperemia.
The mapping may be generally represented as equation (2) to determine the change in pressure ΔP at hyperemia.
ΔP(hyperemia)=ƒ(ΔP(rest),Features(rest)) (2)
The mapping function ƒ may be determined using, e.g., machine learning, an optimization approach, a statistical approach, a data-driven approach, a control system based approach, or any other suitable function that can represent relationships between the pressure drop across the stenosis for a patient at the rest state and the pressure drop across the stenosis for a patient at the hyperemia state. The mapping function ƒ may be trained in an offline step using training data. The training data may include pressure drops across a stenosis for a set of patients at a rest state and at a hyperemia state. The pressure drops across a stenosis may be computed based on, e.g., actual blood flow measurements for the set of patients at a rest state and at a hyperemia state, simulated blood flow measurements for the set of patients at a rest state and at a hyperemia state, or a combination of both. The set of patients used in the training may be all patients having pressure drop data available at a rest state and at a hyperemia state, a particular subclass of patients (e.g., based on demographics, medical history, family history, etc.), or may be patient specific (e.g., using data from that specific patient).
In one embodiment, the pressure drop across the stenosis for the patient at the hyperemia state may be determined by first determining, at the rest state, a pressure distal to the stenosis, Pd(rest), and a pressure proximate to the stenosis, Pa(rest), and then mapping Pd(rest) and Pa(rest) individually to the hyperemia state values, Pd(hyperemia) and Pa(hyperemia), respectively. The pressure drop across the stenosis can then be calculated for a patient at the hyperemia state based on Pd(hyperemia) and Pa(hyperemia). In another embodiment, the pressure drop across the stenosis for the patient at the hyperemia state may be determined by first determining a pressure drop across the stenosis for the patient at the rest state (e.g., ΔP=Pd/Pa), and then mapping the pressure drop across the stenosis for the patient at the rest state ΔP(rest) to the pressure drop across the stenosis for the patient at the hyperemia state ΔP(hyperemia), as shown in equation (3).
P
d
/P
a(hyperemia)=ƒ(Pd/Pa(rest),Features(rest) (3)
At step 410, the FFR of the stenosis for the patient at the hyperemia state is outputted. The FFR of a stenosis is a representation of the pressure drop across the stenosis for a patient at hyperemia. The FFR for a stenosis may be represented as in equation (4).
FFR=Pd/Pa(hyperemia) (4)
In some embodiments, the FFR of a stenosis for a patient at a hyperemia state may be determined using a different rest state metric. For example, the instantaneous wave-free ratio (iFR) of a patient at a rest state may be mapped to the FFR of the stenosis for the patient at a hyperemia state based on features extracted from patient data of a patient at the rest state. The mapping may be represented as equation (5).
FFR(hyperemia)=ƒ(iFR(rest),Features(rest)) (5)
In another example, a quantity of interest of a patient in a post-intervention state may be determined without actually performing the intervention or explicitly measuring/computing the QoI in the post-intervention state. Conventional techniques for predicting the effect of an intervention on a quantity of interest rely on first performing a virtual intervention (e.g., modifying an anatomical model of an organ to simulate the effect of the intervention on the anatomy) and then computing the physiological metrics in the post-intervention state by performing biophysical computations.
In accordance with one embodiment, a QoI of a patient in the post-intervention state may be determined based on the QoI of the patient in the pre-intervention state and features of the patient in the pre-intervention state. The intervention may be, e.g., a stenting procedure whereby the stenosis in a blood vessel is treated by inserting a stent and dilating it in order to ease constriction. This allows for blood to flow more easily, thereby alleviating the pressure drop. In order to ascertain whether the stenting procedure was successful and to plan in case of multiple stenosises, one would benefit from the knowledge of the potential post-intervention pressure drop even before performing the intervention. The mapping of the pressure drop of a patient in a pre-stenting state can be mapped to determine the pressure drop of a patient in a post-stenting state based on features of the patient data in the pre-stenting state. The features of the patient data in the pre-stenting state may include anatomical aspects of the vessel, properties of the stent (e.g., length of the stent, diameter of the stent), etc. The mapping may be represented in equation (6).
ΔP(post−stent)=ƒ(ΔP(pre−stent),Features(pre−stent)) (6)
Other examples of interventions may include invasive, minimally invasive, or non-invasive interventions that may utilize a medical device (e.g., implanted or used to perform the intervention). For example, the intervention may include a stent therapy, a flow diverter therapy, an ablation therapy, an electrical simulation therapy, a vascular surgery, etc. The interventions may also include pharmaceutical interventions. Examples of pharmaceutical interventions include blood pressure lowering drugs, a vasodilator or vasoconstriction drug, etc.
It should be understood that the embodiments discussed herein may be applied to determine any quantity of interest at any state. For example, for cardiac resynchronization therapy (CRT), quantities of interest may include, e.g., electrocardiography (ECG) (e.g., QT wave duration) and activation time, which may be determined for a post-implant stage (i.e., after the CRT device is implanted in the heart) based on quantities of interest at a pre-implant stage and features extracted from patient data acquired at the pre-implant stage, where the features may include features based on cardiac anatomy, flow, mechanics and electrophysiology, CRT device parameters, etc. In another example, for aortic valve stenting, quantities of interest may include, e.g., trans-stenotic pressure drop, para-valvular leakage, and peak velocity, which may be determined for a post-implant stage (i.e., after the stent is implanted in the aortic valve) based on quantities of interest at a pre-implant stage and features extracted from patient data acquired at the pre-implant stage, where the features may include, e.g., valve, aorta, and outflow tract anatomy, blood pressure, and stent size.
It should further be understood that the embodiments discussed herein are not limited to the medical field, but may be employed to determine any quantity of interest for any subject at a second state based on data of the subject at a first state.
Systems, apparatuses, and methods described herein may be implemented using digital circuitry, or using one or more computers using well-known computer processors, memory units, storage devices, computer software, and other components. Typically, a computer includes a processor for executing instructions and one or more memories for storing instructions and data. A computer may also include, or be coupled to, one or more mass storage devices, such as one or more magnetic disks, internal hard disks and removable disks, magneto-optical disks, optical disks, etc.
Systems, apparatus, and methods described herein may be implemented using computers operating in a client-server relationship. Typically, in such a system, the client computers are located remotely from the server computer and interact via a network. The client-server relationship may be defined and controlled by computer programs running on the respective client and server computers.
Systems, apparatus, and methods described herein may be implemented within a network-based cloud computing system. In such a network-based cloud computing system, a server or another processor that is connected to a network communicates with one or more client computers via a network. A client computer may communicate with the server via a network browser application residing and operating on the client computer, for example. A client computer may store data on the server and access the data via the network. A client computer may transmit requests for data, or requests for online services, to the server via the network. The server may perform requested services and provide data to the client computer(s). The server may also transmit data adapted to cause a client computer to perform a specified function, e.g., to perform a calculation, to display specified data on a screen, etc. For example, the server may transmit a request adapted to cause a client computer to perform one or more of the method steps described herein, including one or more of the steps of
Systems, apparatus, and methods described herein may be implemented using a computer program product tangibly embodied in an information carrier, e.g., in a non-transitory machine-readable storage device, for execution by a programmable processor; and the method steps described herein, including one or more of the steps of
A high-level block diagram 500 of an example computer that may be used to implement systems, apparatus, and methods described herein is depicted in
Processor 504 may include both general and special purpose microprocessors, and may be the sole processor or one of multiple processors of computer 502. Processor 504 may include one or more central processing units (CPUs), for example. Processor 504, data storage device 512, and/or memory 510 may include, be supplemented by, or incorporated in, one or more application-specific integrated circuits (ASICs) and/or one or more field programmable gate arrays (FPGAs).
Data storage device 512 and memory 510 each include a tangible non-transitory computer readable storage medium. Data storage device 512, and memory 510, may each include high-speed random access memory, such as dynamic random access memory (DRAM), static random access memory (SRAM), double data rate synchronous dynamic random access memory (DDR RAM), or other random access solid state memory devices, and may include non-volatile memory, such as one or more magnetic disk storage devices such as internal hard disks and removable disks, magneto-optical disk storage devices, optical disk storage devices, flash memory devices, semiconductor memory devices, such as erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM), digital versatile disc read-only memory (DVD-ROM) disks, or other non-volatile solid state storage devices.
Input/output devices 508 may include peripherals, such as a printer, scanner, display screen, etc. For example, input/output devices 508 may include a display device such as a cathode ray tube (CRT) or liquid crystal display (LCD) monitor for displaying information to the user, a keyboard, and a pointing device such as a mouse or a trackball by which the user can provide input to computer 502.
Any or all of the systems and apparatus discussed herein, including elements of system 200 of
One skilled in the art will recognize that an implementation of an actual computer or computer system may have other structures and may contain other components as well, and that
The foregoing Detailed Description is to be understood as being in every respect illustrative and exemplary, but not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed Description, but rather from the claims as interpreted according to the full breadth permitted by the patent laws. It is to be understood that the embodiments shown and described herein are only illustrative of the principles of the present invention and that various modifications may be implemented by those skilled in the art without departing from the scope and spirit of the invention. Those skilled in the art could implement various other feature combinations without departing from the scope and spirit of the invention.
This application is a divisional of U.S. patent application Ser. No. 14/599,678, filed Jan. 19, 2015, which claims the benefit of U.S. Provisional Application No. 62/004,294, filed May 29, 2014, the disclosures of which are herein incorporated by reference in their entirety.
| Number | Date | Country | |
|---|---|---|---|
| 62004294 | May 2014 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 14599678 | Jan 2015 | US |
| Child | 17070993 | US |
