Systems and Methods for Personalized Cardiovascular Analysis

Abstract

A system and method for performing personalized cardiovascular analysis is provided, comprising: using a modeling and simulation computing device to build a patient-specific model; using the device to store the patient-specific model in a database; receiving, at the device, remote monitoring data from at least one remote monitoring data source; and receiving, at the device, clinical data from at least one clinical data source. The method further comprises using the device to update the patient-specific model with the remote monitoring data and the clinical data; using the device to run at least one simulation against the updated patient-specific model; and outputting, from the device, at least one output based on the at least one simulation.

Description

[Technical Field]

[0001] (CROSS-REFERENCE TO RELATED APPLICATIONS) This application claims priority to U.S. Provisional Application No. 62 / 951,312, filed December 20, 2019, which is incorporated herein by reference in its entirety.

[0002] The present disclosure relates to monitoring patients with cardiovascular problems, and more particularly, the present disclosure relates to performing patient-specific cardiovascular analyses. [Background technology]

[0003] Ventricular assist systems (VAS) include ventricular assist devices (VADs), such as implantable blood pumps, used for both short-term (i.e., days, months) and long-term (i.e., years or lifelong) applications when a patient's heart cannot provide adequate circulation (commonly referred to as heart failure or congestive heart failure). Patients with heart failure may use a VAS while awaiting a heart transplant or as long-term destination therapy. In another example, a patient may use a VAS while recovering from cardiac surgery. Thus, a VAS can compensate for (i.e., provide partial support for) a weak heart or effectively replace the function of the natural heart. A VAS can be implanted within the patient and powered by a source either internal or external to the patient.

[0004] Understanding and managing hemodynamics in patients with heart failure is relatively challenging, and additional complications arise in patients with VADs who also have one or more underlying conditions, such as right ventricular (RV) dysfunction, valvular disorders, or arrhythmias.

[0005] One focus of next-generation VAD development is adding system enhancements that help physicians optimize treatment to improve adverse event profiles and overall patient quality of life. Examples of enhancements include the use of improved sensing capabilities (e.g., flow waveform sensing, pressure waveform sensing, accelerometers, etc.) combined with closed-loop control algorithms (e.g., pulse and physiological control). These enhancements are utilized in so-called "smart VADs."

[0006] However, these new capabilities increase the complexity and sophistication of patient management, especially when patients have additional devices such as pacemakers or pulmonary artery (PA) pressure sensors. Therefore, physicians need to be educated and trained on how to manage complex patient / device system interactions and understand the implications of new diagnostic tools as they emerge. Furthermore, there is a need to simplify and integrate cardiac assist device interfaces, leverage remote monitoring to record patient / device data over long periods outside of the clinic, and develop algorithms that can process and interpret device data to facilitate optimization of patient care. Therefore, there is a need for an integrated remote monitoring infrastructure. Summary of the Invention [Means for solving the problem]

[0007] In one embodiment, the present disclosure relates to a method for performing personalized cardiovascular analysis, the method including: constructing the patient-specific model using a modeling and simulation computing device; storing the patient-specific model in a database using the modeling and simulation computing device; receiving remote monitoring data from at least one remote monitoring data source using the modeling and simulation computing device; and receiving clinical data from at least one clinical data source using the modeling and simulation computing device. The method further includes updating the patient-specific model using the remote monitoring data and the clinical data using the modeling and simulation computing device; running at least one simulation using the modeling and simulation computing device against the updated patient-specific model; and outputting at least one output from the modeling and simulation computing device based on the at least one simulation.

[0008] In another embodiment, the present disclosure relates to a computing device for performing personalized cardiovascular analysis, the computing device including a memory device and a processor communicatively coupled to the memory device, the processor configured to: construct a patient-specific model, store the patient-specific model in the memory device, receive remote monitoring data from at least one remote monitoring data source, receive clinical data from at least one clinical data source, update the patient-specific model using the remote monitoring data and the clinical data, run at least one simulation on the updated patient-specific model, and output an output based on the at least one simulation.

[0009] In yet another embodiment, the present disclosure relates to a non-transitory computer-readable medium having computer-executable instructions that, when executed by a processor of a computing device, cause the processor of the computing device to: build a patient-specific model, store the patient-specific model in a database, receive remote monitoring data from at least one remote monitoring data source, receive clinical data from at least one clinical data source, update the patient-specific model using the remote monitoring data and the clinical data, perform at least one simulation on the updated patient-specific model, and output an output based on the at least one simulation.

[0010] These and other aspects, features, details, utilities, and advantages of the present disclosure will become apparent from a reading of the following description and claims, and a review of the accompanying drawings. [Brief explanation of the drawings]

[0011] [Figure 1] 1 is a schematic diagram of an embodiment of a remote monitoring and simulation system. [Figure 2] FIG. 1 is a block diagram of one embodiment of a computing device that may be used to implement the systems and methods described herein. [Figure 3] Block diagram of a method for performing personalized cardiovascular analysis.

[0012] Corresponding reference characters indicate corresponding parts throughout the several views. DETAILED DESCRIPTION OF THE INVENTION

[0013] The present disclosure provides a system and method for performing personalized cardiovascular analysis. The method includes: constructing a patient-specific model using a modeling and simulation computing device; storing the patient-specific model in a database using the modeling and simulation computing device; receiving remote monitoring data from at least one remote monitoring data source using the modeling and simulation computing device; and receiving clinical data from at least one clinical data source using the modeling and simulation computing device. The method further includes updating the patient-specific model using the remote monitoring data and the clinical data using the modeling and simulation computing device; running at least one simulation using the modeling and simulation computing device against the updated patient-specific model; and outputting at least one output from the modeling and simulation computing device based on the at least one simulation.

[0014] Embodiments described herein provide a cloud-based remote monitoring and simulation system that stores and maintains a cardiovascular model with patient-specific values. The system uses machine learning and optimization techniques to continuously update the model's simulation parameters using actual patient data collected from clinical and remote monitoring data sources. The more data collected by the system, the more accurate the model becomes. Because the model replicates a specific patient, it is sometimes referred to as a "digital clone" of that patient.

[0015] Referring to the drawings, in which like reference numbers are used to identify identical components in various views, FIG. 1 illustrates one embodiment of a remote monitoring and simulation system 100. The remote monitoring and simulation system 100 includes a modeling and simulation computing device 102. In an exemplary embodiment, the modeling and simulation computing device 102 is a cloud-based server system. Alternatively, the modeling and simulation computing device 102 may be any computing device suitable for implementing the systems and methods described herein.

[0016] The modeling and simulation computing device 102 is communicatively connected to multiple remote monitoring data sources 104 and multiple clinical data sources 106. To build and update a patient-specific model for a particular patient, the modeling and simulation computing device 102 collects remote monitoring data related to the patient from the remote monitoring data sources 104 and clinical data related to the patient from the clinical data sources 106, as described herein. In some embodiments, the data is time-stamped as it is received by the modeling and simulation computing device 102.

[0017] The patient-related remote monitoring data sources 104 include, for example, a patient monitoring device 110, a heart failure monitor 112, a CRT (cardiac resynchronization therapy) device 114, and a VAD 116. In some embodiments, the patient monitoring device 110 first collects remote monitoring data from other devices (e.g., the heart failure monitor 112, the CRT device 114, and the VAD 116) and relays the collected remote monitoring data to the modeling and simulation computing device 102. Alternatively, each device may independently transmit remote monitoring data to the modeling and simulation computing device 102.

[0018] The remote monitoring data collected from the remote monitoring data sources 104 may include patient hemodynamic data and device status data. For example, the collected data may include a pump flow waveform, a left ventricular (LV) pressure waveform, and an aortic pressure waveform (e.g., from the VAD 116). The collected remote monitoring data may also, or alternatively, include a pulmonary artery (PA) pressure waveform (e.g., from the heart failure monitor 112), CRT data (e.g., from the CRT device 114), and implantable cardiac monitor data. Those skilled in the art will appreciate that other types of remote monitoring data may also be collected.

[0019] Clinical data sources 106 may include, for example, physician computing devices, electronic medical record systems, etc. Additionally, patient clinical data may include in-hospital measurements such as, for example, right catheter measurements, echocardiographic data, blood pressure measurements, etc.

[0020] 2 illustrates one embodiment of a computing device 200 that may be used to implement the systems and methods described herein. Computing device 200 may be used, for example, to implement modeling and simulation computing device 102 (shown in FIG. 1). Computing device 200 includes at least one memory device 210 and a processor 215 coupled to memory device 210 for executing instructions. In some embodiments, executable instructions are stored in memory device 210. In this embodiment, computing device 200 performs one or more operations described herein by programming processor 215. For example, processor 215 may be programmed by encoding operations as one or more executable instructions and by providing the executable instructions to memory device 210.

[0021] Processor 215 includes one or more processing units (e.g., in a multi-core configuration). Additionally, processor 215 may be implemented using one or more heterogeneous processor systems in which a main processor resides on a single chip with secondary processors. In another illustrative example, processor 215 may be a symmetric multiprocessor system including multiple processors of the same type. Additionally, processor 215 may be implemented using any suitable programmable circuitry, including one or more system and microcontrollers, microprocessors, reduced instruction set circuits (RISC), application specific integrated circuits (ASICs), programmable logic circuits, field programmable gate arrays (FPGAs), and any other circuitry capable of performing the functions described herein.

[0022] In this embodiment, memory device 210 is one or more devices that allow for the storage and retrieval of information, such as executable instructions and / or other data. Memory device 210 includes one or more computer-readable media, such as, but not limited to, dynamic random access memory (DRAM), static random access memory (SRAM), solid-state disks, and / or hard disks. Memory device 210 is configured to store, but is not limited to, application source code, application object code, subject source code portions, subject object code portions, configuration data, execution events, and / or any other type of data.

[0023] In this embodiment, computing device 200 includes a presentation interface 220 connected to processor 215. Presentation interface 220 presents information to user 225. For example, presentation interface 220 includes a display adapter (not shown) connected to a display device such as a cathode ray tube, a liquid crystal display (LCD), an organic light-emitting diode (OLED) display, and / or an "electronic ink" display. In some embodiments, presentation interface 220 includes one or more display devices. Input signals and / or filtered signals processed using embodiments described herein are displayed on presentation interface 220.

[0024] In this embodiment, computing device 200 includes a user input interface 235. User input interface 235 is coupled to processor 215 and receives input from user 225. User input interface 235 may include, for example, a keyboard, a pointing device, a mouse, a stylus, a touch-sensitive panel (e.g., a touchpad or touchscreen), a gyroscope, an accelerometer, a position detector, and / or an audio user input interface. A single component, such as a touchscreen, may function as both the display device for presentation interface 220 and as user input interface 235.

[0025] In this embodiment, computing device 200 includes a communication interface 240 coupled to processor 215. Communication interface 240 communicates with one or more remote devices. To communicate with remote devices, communication interface 240 may include, for example, a wired network adapter, a wireless network adapter, and / or a mobile communication adapter.

[0026] 1, using the remote monitoring data collected from the remote monitoring data sources 104 and the clinical data collected from the clinical data sources 106, the modeling and simulation computing device 102 builds and updates a patient-specific model. The patient-specific model is stored, for example, in the memory device 210 (shown in FIG. 2).

[0027] The modeling and simulation computing device 102 then runs one or more simulations against the patient-specific model to facilitate optimization of the patient's treatment, with the output of the simulations being displayed, for example, on a presentation interface 220 (shown in FIG. 2).

[0028] For example, before a patient is implanted with a VAD, a simulation can be performed to simulate how the patient will respond to VAD therapy. In this scenario, because a VAD has not yet been implanted, a patient model is constructed using data from the heart failure monitor 112 and / or CRT device 114 and data from in-hospital measurements. These results of such a simulation can help a clinician decide whether to implant a VAD and the type of VAD to implant (e.g., an LVAD, a Bi-VAD, etc.).

[0029] After implantation of the VAD, simulations are performed with the patient-specific model to simulate the operation of the VAD. For example, the effects of adjusting pump speed and / or pulse type are simulated. In another example, the effects of adjusting medication and / or fluid volume are simulated.

[0030] Additionally, in some embodiments, the modeling and simulation computing device 102 analyzes the patient-specific model (e.g., using machine learning and / or other artificial intelligence techniques), and the output of these analyses is displayed, for example, in presentation interface 220 (shown in FIG. 2).

[0031] For example, as a result of analyzing the patient-specific model, the modeling and simulation computing device 102 recommends changes in VAD parameters (e.g., pump speed) and / or medications to achieve a desired hemodynamic outcome (e.g., a desired LV pressure profile). In another example, the modeling and simulation computing device 102 tracks changes in hemodynamic characteristics over time to identify patterns and / or generate alerts. Tracked hemodynamic characteristics include LV / RV systolic function, pulmonary and systemic vascular resistance, vascular compliance, fluid / volume status, valve function, hematocrit, etc. In one example, the modeling and simulation computing device 102 tracks cardiac recovery (e.g., ventricular working capacity) and optimizes the recovery protocol on a patient-specific basis (e.g., by recommending medication amount / timing, pump weaning, pulse type, etc.).

[0032] Further, based on the analysis of the patient-specific model, the modeling and simulation computing device 102 generates recommended device settings. For example, the modeling and simulation computing device 102 generates recommended settings for pump speeds, physiological control settings, pulse types, and / or pacemaker settings. In some embodiments, the modeling and simulation computing device 102 controls one or more of the remote monitoring data sources 104. For example, the modeling and simulation computing device 102 generates and sends control signals to the CRT device 114 and / or the VAD 116 to instruct the CRT device 114 and / or the VAD 116 to adjust their settings.

[0033] In the embodiments described herein, the patient-specific model is an online numerical model that continuously updates to match the patient's current condition based on data received from remote monitoring data sources 104 and clinical data sources 106. The more data that is available and incorporated into the model, the more accurate the simulations and analyses performed by the modeling and simulation computing device 102 will be.

[0034] Patient-specific models are, for example, high-fidelity lumped-parameter numerical models that simulate the human circulatory system. Additionally, the models include features that allow the models to better match real patients, such as physiological feedback mechanisms and nonlinear outflow graft dynamics.

[0035] For example, in one embodiment, the model includes approximately 75 parameters that define the behavior of the entire circulatory system. By varying these parameters, the model can reproduce the behavior of most patients (except those with severe autoregulation disorders).

[0036] Given the characteristics and hemodynamic variability of VAD patients, in some embodiments, the model is tailored to match a particular patient. Specifically, machine learning and / or other artificial intelligence techniques are used to systematically vary key model parameters until the simulated model replicates actual hemodynamic and pump parameter data (e.g., ventricular dimensions, PA pressures, etc.) from clinical data.

[0037] In some embodiments, the modeling and simulation computing device 102 also builds a database of generalized models. For example, each generalized model is built using corresponding clinical study data. The generated database therefore includes multiple de-identified models. These models are searchable based on general patient characteristics, such as gender, race, weight, BMI, cardiac index, INTERMACS classification, etc. This searchable database allows users to search for de-identified models based on the characteristics of a patient of interest and use the de-identified models as a "starting point" for a patient-specific model. In particular, the more similar the de-identified models are to the patient of interest, the faster they will converge to an accurate patient-specific model. Furthermore, in some embodiments, the modeling and simulation computing device 102 receives existing clinical and / or demographic data associated with the patient and automatically selects a de-identified model from the database based on the existing clinical and / or demographic data.

[0038] The modeling and simulation computing device 102 allows a user (e.g., a clinician) to clearly visualize and understand the mechanisms behind various patient / device interactions. Furthermore, by running simulations, a user can "virtually" experiment by changing pump speeds, medications, fluid volumes, etc., and observe the quantified expected hemodynamic response. Furthermore, the modeling and simulation computing device 102 allows a user to simulate various events (e.g., dehydration, arrhythmia, exercise, acute hypertension, etc.) to observe the expected results and aid in the future recognition of such events.

[0039] The modeling and simulation computing device 102 also uses machine learning and / or other artificial intelligence techniques to continuously update key model parameters (e.g., ventricular elastance curves, valve resistance, aortic compliance, etc.) to ensure that the simulation results are consistent with or closely track clinical data. That is, the patient-specific models estimate various patient parameters without direct physical measurements, which is very useful for tracking physiological changes over time in ventricular function, valve leakage, etc.

[0040] To illustrate the effectiveness of the systems and methods described herein, an exemplary case study is described below. Specifically, the following case study illustrates how the systems and methods described herein can be used to assist in the treatment of a dilated cardiomyopathy (DCM) heart failure patient throughout the heart failure continuum. Those skilled in the art will appreciate that similar techniques can be applied to other heart failure patients. This case study is purely fictitious and is intended to illustrate how the systems and methods described herein can be implemented.

[0041] In this example, assume that a patient gradually develops an underlying electrophysiological disorder (e.g., left bundle branch block) and LV function deteriorates. The patient is initially asymptomatic and unaware of the problem. However, over time, the electrical conduction problems worsen, ventricular pump function is impaired, and the patient occasionally exhibits mild heart failure (HF) symptoms (i.e., Stage II).

[0042] Despite activation of the renin-angiotensin (RAS) feedback system, the patient is unable to maintain adequate cardiac output and arterial pressure. Chronically elevated ventricular pressure further dilates the LV, further impairing pump function. The patient is now symptomatic enough to consult a cardiologist (i.e., stage III). The cardiologist performs a complete hemodynamic workup, measuring an ejection fraction of 30% and identifying an underlying left bundle branch block disorder.

[0043] At this point, the patient is implanted with a pacemaker and a heart failure monitor (e.g., a PA pressure sensor) and assigned a typical HF medication regimen. The patient is then registered for an account with a remote monitoring system (which may be the same as or separate from the modeling and simulation computing device 102), and the patient is provided with the necessary equipment to enable remote monitoring of the device. Further, in accordance with the systems and methods described herein, the clinician initializes a patient-specific model for the patient using the modeling and simulation computing device 102. For example, the modeling and simulation computing device 102 automatically selects initial parameters for the model based on the patient's existing clinical and / or demographic data.

[0044] The pacemaker initially improves the heart's pumping ability, mitigating the patient's apparent HF system back to Stage II. However, the previous chronic LV dilation physically damages the heart, resulting in persistently high ventricular filling pressures. Over the next several months / years, the LV continues to dilate despite successful pacing. The heart failure monitor tracks the higher and drifting PA pressure, and these results are communicated to the remote monitoring system and displayed to the clinician. Furthermore, the patient-specific model is updated based on data received from the pacemaker and heart failure monitor, and now has a 65% confidence level of agreement with the actual patient. The modeling and simulation computing device 102 also reports a decline in LV systolic function based on the model's analysis.

[0045] The clinician feels generally OK but decides to return the patient to the clinic for a complete hemodynamic workup. The results show severely dilated LV and poor systemic hemodynamics. These in-clinic results are provided to the modeling and simulation computing device 102, and the confidence level of agreement increases to 75 percent. The clinician adjusts medications and sends the patient home for remote observation.

[0046] Over the next few months, HF symptoms worsen to stage IV, rendering the patient unable to perform daily tasks. Patient-specific models suggest that LV systolic function is very poor and steadily deteriorating. Eventually, the patient returns to the clinic for another hemodynamic workup. The results are poor, with an ejection fraction of 20% and a low cardiac index. The patient appears to be a good candidate for recovery (young, only a few years into HF), but a poor candidate for transplant (blood type O, weight >100 kg).

[0047] Using the modeling and simulation computing device 102, the clinician simulates how the patient will respond to VAD therapy (the confidence level of agreement is currently 80 percent). The patient-specific model shows a good hemodynamic response to VAD support and strong RV function. Therefore, it is decided to implant a VAD, which can communicate data to the modeling and simulation computing device 102.

[0048] Within a few weeks of VAD implantation, the patient is healthy and ambulatory without HF symptoms (i.e., Stage I). The patient-specific model, which now also incorporates extensive data from the VAD, now achieves a 95% confidence level of fit. After three months, analysis of the patient-specific model by the modeling and simulation computing device 102 now shows improved LV systolic function and remodeling. The clinician reviews the recorded data and allows the modeling and simulation computing device 102 to instruct the VAD to initiate a recovery protocol. In this recovery protocol, the VAD periodically reduces pump speed to automatically "train" the LV.

[0049] After an additional two months, analysis of the patient-specific model by the modeling and simulation computing device 102 consistently showed strong LV function, and a weaning algorithm was implemented to reduce the VAD pump speed to a relatively low level so that VAD support was minimized. Using the modeling and simulation computing device 102, the clinician simulated VAD removal, and the simulation results indicated a high likelihood of recovery. Therefore, the decision was made to wean the patient from the VAD. Two years after weaning from the VAD, the patient is showing stable, sustained recovery. The patient-specific model maintained by the modeling and simulation computing device 102 remains active, using available data from the pacemaker and heart failure monitor.

[0050] This case study therefore illustrates the benefits realized using the systems and methods described herein.

[0051] 3 is a block diagram of a method 300 for performing personalized cardiovascular analysis. The method 300 may be implemented, for example, in the modeling and simulation computing device 102 (shown in FIG. 1).

[0052] The method 300 includes building 302 a patient-specific model. The method 300 further includes storing 304 the patient-specific model in a database. The method 300 further includes receiving 306 remote monitoring data from at least one remote monitoring data source and receiving 308 clinical data from at least one clinical data source. The method 300 further includes updating 310 the patient-specific model using the remote monitoring data and the clinical data. The method 300 further includes running 312 at least one simulation on the updated patient-specific model and outputting 314 at least one output based on the at least one simulation.

[0053] The systems and methods described herein include a build step of constructing a patient-specific model using a modeling and simulation computing device, a store step of storing the patient-specific model in a database using the modeling and simulation computing device, a remote monitoring data receiving step of receiving remote monitoring data from at least one remote monitoring data source at the modeling and simulation computing device, and a clinical data receiving step of receiving clinical data from at least one clinical data source at the modeling and simulation computing device. The systems and methods further include an update step of updating the patient-specific model with the remote monitoring data and the clinical data using the modeling and simulation computing device, an execution step of running at least one simulation on the updated patient-specific model using the modeling and simulation computing device, and an output step of outputting at least one output from the modeling and simulation computing device based on the at least one simulation.

[0054] While particular embodiments of the present disclosure have been described above with a certain degree of particularity, those skilled in the art may make numerous modifications to the disclosed embodiments without departing from the spirit or scope of the present disclosure. All directional references (e.g., top, bottom, upward, downward, left, right, leftward, rightward, up, down, upward, downward, vertical, horizontal, clockwise, and counterclockwise) are used for identification purposes only to aid in understanding the present disclosure and are not intended to create limitations with respect to the location, orientation, or use of the present disclosure. Joint references (e.g., attached, connected, coupled, etc.) should be interpreted broadly and include intermediate members between connections of elements as well as relative movement between elements. As such, joint references do not necessarily imply that two elements are directly connected and fixed with respect to each other. All matter contained in the above description or shown in the accompanying drawings is intended to be interpreted as illustrative only and not limiting. Changes in detail or structure may be made without departing from the spirit of the disclosure, as defined by the appended claims.

[0055] When introducing elements of the disclosure or preferred embodiments thereof, the articles "a," "an," "the," and "said" are intended to mean the presence of one or more elements. The terms "including," "comprising," and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements.

[0056] Since various changes may be made in the above configurations without departing from the scope of the present disclosure, it is intended that all matter contained in the above description or shown in the accompanying drawings be interpreted as illustrative and not in a limiting sense.

Claims

1. 1. A computer-implemented method for performing personalized cardiovascular analysis, comprising: constructing said patient-specific model using a modeling and simulation computing device; storing the patient-specific model in a database using the modeling and simulation computing device; receiving, at the modeling and simulation computing device, remote monitoring data from at least one remote monitoring data source; receiving, at said modeling and simulation computing device, clinical data from at least one clinical data source; updating the patient-specific model using the remote monitoring data and the clinical data using the modeling and simulation computing device; running at least one simulation against the updated patient-specific model using the modeling and simulation computing device; and outputting, from said modeling and simulation computing device, at least one output based on said at least one simulation.

2. an analyzing step of analyzing the updated patient-specific model using the modeling and simulation computing device; 10. The method of claim 1, further comprising: outputting at least one additional output from the modeling and simulation computing device based on the analysis.

3. 3. The method of claim 2, wherein the additional output step includes displaying a recommended adjustment of one of a device parameter and a medication regimen.

4. The method of claim 2 , wherein the analyzing step comprises analyzing the updated patient-specific model using machine learning.

5. 3. The method of claim 2, wherein the additional output step includes transmitting a control signal to an implantable device.

6. 10. The method of claim 1, wherein the constructing step comprises generating a database including a plurality of de-identified patient models; and selecting one of the plurality of de-identified patient models as the patient-specific model.

7. 10. The method of claim 1, wherein the performing step includes simulating operation of a ventricular assist device.

8. 1. A computing device for performing personalized cardiovascular analysis, comprising: a memory device; a processor communicatively connected to the memory device; The processor: Build patient-specific models storing the patient-specific model in the memory device; receiving remote monitoring data from at least one remote monitoring data source; receiving clinical data from at least one clinical data source; updating the patient-specific model using the remote monitoring data and the clinical data; running at least one simulation on the updated patient-specific model; A computing device that outputs an output based on the at least one simulation.

9. The computing device of claim 8 , wherein the processor analyzes the updated patient-specific model and outputs at least one additional output based on the analysis.

10. The computing device of claim 9 , wherein the at least one additional output includes a recommended adjustment of one of a device parameter and a medication regimen.

11. 10. The computing device of claim 9, wherein to analyze the updated patient-specific model, the processor is configured to analyze the updated patient-specific model using machine learning.

12. The computing device of claim 9 , wherein the at least one additional output comprises a control signal sent to an implantable device.

13. 10. The computing device of claim 8, wherein to construct the patient-specific model, the processor is configured to generate a database including a plurality of de-identified patient models and select one of the plurality of de-identified patient models as the patient-specific model.

14. The computing device of claim 8 , wherein to perform the at least one simulation, the processor is configured to simulate operation of a ventricular assist device.

15. A non-transitory computer-readable medium having computer-executable instructions that, when executed by a processor of a computing device, cause the processor of the computing device to: Building a patient-specific model; storing the patient-specific model in a database; receiving remote monitoring data from at least one remote monitoring data source; receiving clinical data from at least one clinical data source; updating the patient-specific model using the remote monitoring data and the clinical data; and running at least one simulation on the updated patient-specific model; and outputting an output based on the at least one simulation.

16. 16. The non-transitory computer-readable medium of claim 15, wherein the computer-executable instructions further cause the processor to analyze the updated patient-specific model and output at least one additional output based on the analysis.

17. 17. The non-transitory computer-readable medium of claim 16, wherein the at least one additional output includes a recommended adjustment of one of a device parameter and a medication regimen.

18. 17. The non-transitory computer-readable medium of claim 16, wherein the computer-executable instructions cause the processor to analyze the updated patient-specific model using machine learning.

19. 17. The non-transitory computer-readable medium of claim 16, wherein the at least one additional output comprises a control signal transmitted to an implantable device.

20. 16. The non-transitory computer-readable medium of claim 15, wherein the computer-executable instructions, to perform the at least one simulation, cause the processor to simulate operation of a ventricular assist device.

Images