Clinical decision-making support in drug therapy for patients with acute heart failure

Abstract

Embodiments disclosed herein establish relationships between changes in cardiovascular indicators and drug combinations. These relationships are established by tracking causal chains described by linear relationships, from drug combinations to cardiovascular parameters and from cardiovascular parameters to cardiovascular indicators, by measuring the direction and sensitivity of hemodynamic changes caused by changes in cardiovascular parameters (resulting from drug administration). Some embodiments also model myocardial oxygen consumption, which cannot be directly measured, and establish relationships between hemodynamic changes in myocardial oxygen consumption and changes in one or more cardiovascular parameters. Because these embodiments qualitatively and quantitatively establish causal chains from drugs to cardiovascular indicators, they support clinical decision-making, thereby reducing the level of mere guesswork and uncertainty in determining the optimal drug combination for patients with acute heart failure.

Description

Technical Field

[0001] Cross - Reference to Related Applications This disclosure relates to U.S. Provisional Application No. 63 / 493,609, filed on March 31, 2023, entitled "Assisting Clinical Decision - Making in Drug Therapy for Acute Heart Failure Patients", claims priority from the said application, and the entire application is incorporated herein by reference.

[0002] This disclosure also relates to U.S. Provisional Application No. 63 / 346,143, filed on May 26, 2022, entitled "Optimizing Drug Combinations for Treating Acute Heart Failure", and the entire application is incorporated herein by reference.

[0003] This disclosure relates to systems, methods, and devices for assisting clinical decision - making to determine optimal combinations of drugs to reach target cardiovascular metrics in patients with acute heart failure.

Background Art

[0004] Acute heart failure is caused by different factors and thus generally requires combination drug therapy (also called pharmacotherapy) using multiple drugs. Some examples of drugs for treating acute heart failure include positive inotropes such as dobutamine, vasopressors such as norepinephrine, vasodilators such as sodium nitroprusside, fluids such as dextran, and diuretics such as furosemide. Each of these drugs can treat different specific aspects of acute heart failure. For more effective treatment of acute heart failure, generally combinations of these drugs are required.

[0005] Furthermore, acute heart failure has an adverse effect on multiple organs of the body, including the heart. In particular, acute heart failure changes the cardiovascular metrics of the heart, and these changes cause adverse effects on other organs. Therefore, in a clinical setting, it is desirable to stably maintain the cardiovascular metrics within a safe margin so that problems arising from acute heart failure do not chain to other organs.

[0006] However, the technical challenges are complexity and many uncertainties. As described above, there are different drugs, each affecting one aspect of the heart and producing its specific side effects. For example, even when only two drugs are administered, the effects, both positive and negative, are unknown and difficult to predict. Therefore, clinical decision-making is made by trial and error and mere speculation, and is inherently inaccurate. Such an undesirable situation creates unnecessary suffering for the patient. The health outcome is never desirable, and the mortality rate of heart failure patients remains unnecessarily high.

[0007] Therefore, significant improvements in systems, methods, and devices for assisting clinical decision-making regarding the optimal drug combination to achieve the desired level of cardiovascular metrics in acute heart failure patients are thereby desired. SUMMARY OF THE INVENTION

[0008] In some embodiments, a system may be provided. The system may include a non-temporary storage medium for storing computer program instructions and one or more processors configured to execute computer program instructions to trigger an action. The action may include receiving a plurality of current cardiovascular indicators of a patient having acute heart failure and executing a clinical decision-making assistance module based on the plurality of current cardiovascular indicators to generate a recommended drug combination to reach a plurality of target cardiovascular indicators. The clinical decision-making assistance module may be based on a drug library that linearly models pathways from a plurality of current cardiovascular parameters associated with a plurality of current cardiovascular indicators to a plurality of target cardiovascular parameters associated with a plurality of target cardiovascular indicators, and a mapping between the plurality of target cardiovascular indicators and the plurality of target cardiovascular parameters, the mapping including the time transient directional response of one or more cardiovascular indicators due to changes in one or more cardiovascular parameters caused by the administration of one or more drugs. The action may further include outputting a recommended drug combination.

[0009] In some embodiments, a computer-aided method may be provided. The method may include: a computing system receiving multiple current cardiovascular indicators from a client device of a patient having acute heart failure; and the computing system executing a clinical decision support module based on the multiple current cardiovascular indicators to generate a recommended drug combination to reach multiple target cardiovascular indicators. The clinical decision support module may be based on a drug library that linearly models pathways from multiple current cardiovascular parameters associated with multiple current cardiovascular indicators to multiple target cardiovascular parameters associated with multiple target cardiovascular indicators; and a mapping between the multiple target cardiovascular indicators and multiple target cardiovascular parameters, which includes the temporal transient response of one or more cardiovascular indicators due to changes in one or more cardiovascular parameters caused by the administration of one or more drugs. The method may further include the computing system outputting the recommended drug combination to the client device.

[0010] In some embodiments, a non-temporary storage medium is provided for storing computer program instructions. The computer program instructions, when executed by one or more processors, may trigger an action, which may include receiving a plurality of current cardiovascular indicators of a patient having acute heart failure, and executing a clinical decision support module based on the plurality of current cardiovascular indicators to generate a recommended combination of drugs to reach a plurality of target cardiovascular indicators. The clinical decision support module may be based on a drug library that linearly models pathways from a plurality of current cardiovascular parameters associated with a plurality of current cardiovascular indicators to a plurality of target cardiovascular parameters associated with a plurality of target cardiovascular indicators, and a mapping between the plurality of target cardiovascular indicators and the plurality of target cardiovascular parameters, which includes the temporal transient response of one or more cardiovascular indicators due to changes in one or more cardiovascular parameters caused by the administration of one or more drugs. The action may further include outputting a recommended combination of drugs. [Brief explanation of the drawing]

[0011] [Figure 1] This disclosure illustrates an exemplary computing environment for supporting clinical decision-making in pharmacotherapy for patients with acute heart failure, based on exemplary embodiments of this disclosure. [Figure 2] An exemplary analytical model is shown according to an exemplary embodiment of the present disclosure. [Figure 3] A flowchart of an exemplary method based on exemplary embodiments of this disclosure is shown. [Figure 4] The exemplary embodiments of this disclosure show exemplary graphs that visualize the hemodynamic direction and predicted response of real-world drugs based on analytical models. [Figure 5A] An exemplary graph visualizing a single drug infusion simulation based on exemplary embodiments of this disclosure is shown. [Figure 5B] An exemplary graph visualizing a single drug infusion simulation based on exemplary embodiments of this disclosure is shown. [Figure 5C] An exemplary graph visualizing a single drug infusion simulation based on exemplary embodiments of this disclosure is shown. [Figure 5D] An exemplary graph visualizing a single drug infusion simulation based on exemplary embodiments of this disclosure is shown. [Figure 5E] An exemplary graph visualizing a single drug infusion simulation based on exemplary embodiments of this disclosure is shown. [Figure 6] An exemplary graph visualizing the simulation of three acute heart failure scenarios based on exemplary embodiments of this disclosure is shown. [Figure 7] An exemplary graph illustrating a simulation using the Forrester classification, according to an exemplary embodiment of this disclosure, is shown. [Figure 8] A block diagram of an exemplary computing device that implements various features and processes according to exemplary embodiments of this disclosure is shown. [Modes for carrying out the invention]

[0012] For patients with acute heart failure, it is desirable to maintain cardiovascular indicators (e.g., left atrial pressure, cardiac output, mean arterial pressure, or myocardial oxygen consumption) within a stable range. If cardiovascular indicators are outside this range, adverse effects may occur on other organs and the heart itself. Drug combinations (e.g., inotropic agents, vasoconstrictors, vasodilators, fluids, or diuretics) are used to induce hemodynamic changes in cardiovascular indicators. Drug-induced hemodynamic changes can be modeled as the drug altering cardiovascular parameters (systemic vascular resistance, myocardial contractility, heart rate, or load blood volume), and the cardiovascular parameters similarly altering cardiovascular indicators.

[0013] Embodiments disclosed herein support clinical decision-making regarding drug combinations administered to reach a target range for cardiovascular indicators. For example, some embodiments establish relationships between changes in cardiovascular indicators (also known as hemodynamic changes) and drug combinations. These relationships are established by tracking causal chains described by linear relationships, from drug combinations to cardiovascular parameters and from cardiovascular parameters to cardiovascular indicators, by measuring the direction and sensitivity of hemodynamic changes caused by changes in cardiovascular parameters (resulting from drug administration). Some embodiments also model myocardial oxygen consumption, which cannot be directly measured, and establish relationships between hemodynamic changes in myocardial oxygen consumption and changes in one or more cardiovascular parameters. Because these embodiments qualitatively and quantitatively establish causal chains from drugs to cardiovascular indicators, they support clinical decision-making and thereby reduce the level of mere guesswork and uncertainty in determining the optimal drug combination for patients with acute heart failure.

[0014] Figure 1 shows an exemplary computing environment 100 for assisting clinical decision-making in pharmacotherapy for patients with acute heart failure, according to an exemplary embodiment of the present disclosure. As illustrated, the computing environment 100 may be based on a client-server model in which a server 102 is connected to a plurality of clients 106a-106d (commonly referred to as client 106, or collectively referred to as client 106) via a network 104. However, it should be understood that the client-server model is for illustrative purposes only and for ease of explanation and should not be considered limiting. Accordingly, any type of computing environment performing the functions disclosed herein should be considered within the scope of the disclosure. Furthermore, the individual components of computing environment 100 are merely illustrative, and computing environments with alternative, additional, or fewer components should be considered within the scope of the disclosure.

[0015] The computing environment 100 can generally be in a clinical setting to assess clinical decision-making regarding patients with acute heart failure. In some exemplary use cases, a server 102 may store different software modules 108 that can be accessed by a client 106 using a network 104. The client 106 itself may have a standalone application (not shown) for accessing the software modules 108. Alternatively, the client 106 may access the software modules through, for example, a browser application.

[0016] The hardware of the server 102 that stores the software module 108 may include any type of computing device. For example, server 102 may include, but is not limited to, server computers, desktop computers, laptop computers, tablet computers, and smartphones. Server 102 may not necessarily be in a single location and may be implemented by a network of computers. Furthermore, server 102 may not necessarily be jointly installed within the clinical setting itself and may be hosted by a third-party cloud computing provider. Therefore, any type of server 102 should be considered to be within the scope of this disclosure.

[0017] As described above, client 106 may access server 102 through network 104. Network 104 may include any combination of one or more packet-switched networks (e.g., IP-based networks) and one or more circuit-switched networks (e.g., cellular telephone networks). Some non-limiting examples of network 104 include local area networks, metropolitan area networks, wide area networks such as the Internet, etc. Similarly, non-limiting examples of client 106 may include desktop terminals (e.g., desktop terminal 106a), laptop computers (e.g., laptop computer 106b), tablet computers (e.g., tablet computer 106c), smartphones (e.g., smartphone 106d), etc. Any type of computing device that allows access to server 102 through network 104 should be considered within the scope of this disclosure. Furthermore, the functions described in this disclosure may be distributed in any manner; that is, the functions of server 102 may be performed by one or more clients 106, and vice versa.

[0018] As described above, server 102 may include multiple software modules. Figure 1 shows some non-limiting and exemplary software modules: cardiovascular index data input module 110, optimal dosage calculation module 112, recommended dosage output module 114, model development module 116, simulation module 118, and experimental data acquisition module 120. It should be understood that this modularization of the functions of server 102 described is merely for the sake of clarity and should not be considered limiting. Therefore, any alternative modularization of any kind should be considered within the scope of this disclosure.

[0019] The cardiovascular index data input module 110 may receive cardiovascular index data from the client 106. The received cardiovascular index data may include the patient's current cardiovascular index and target cardiovascular index. For example, the cardiovascular index data input module 110 may receive one or more current measurements of left atrial pressure, cardiac output, mean arterial pressure, or myocardial oxygen consumption as current cardiovascular indexes. Similarly, as target cardiovascular indexes, the cardiovascular index data input module may receive one or more desired measurements of left atrial pressure, cardiac output, mean arterial pressure, or myocardial oxygen consumption. The cardiovascular index data input module 110 may receive cardiovascular index data (current and / or target) from one or more clients 106 via the network 104. In some embodiments, the cardiovascular index data input module 110 may support real-time data input, in which case the server 102 processes the received cardiovascular index data in real time. In other embodiments, the cardiovascular index data input module 110 may support batch processing, in which case individual cardiovascular index data may be batched (e.g., buffered or stored), and processing may be performed on the batched data together (e.g., during off-peak hours on server 102). Thus, the cardiovascular index data input module 110 can manage the reception of data from client 106 for any type of processing.

[0020] The optimal dosage calculation module 112 can calculate the optimal dosage for received cardiovascular index data. For example, the optimal dosage calculation module 112 may use the received current cardiovascular index data to compare it to a target and / or received desired cardiovascular index and call one or more analysis models disclosed herein to calculate the optimal dosage of a drug. The optimal dosage may be based, for example, on both the direction and sensitivity of the changes to cardiovascular index caused by the drug (it should be understood that embodiments disclosed herein model the direction and sensitivity of this change when a drug changes a measurable cardiovascular parameter, and the cardiovascular parameter similarly changes a cardiovascular index). Thus, the optimal dosage calculation module 112 may consider different effects to generate a recommended combination, i.e., different effects such as a first drug causing a change in one direction for a cardiovascular index and a second drug causing a change in the opposite direction. In general, the optimal dosage calculation module 112 can calculate the optimal dosage by utilizing one or more analysis models for specific cardiovascular indexes received by the cardiovascular index data input module 110.

[0021] The recommended dosage output module 114 may provide the clinician with a recommended dosage. Specifically, the recommended dosage output module 114 may take the optimal dosage calculated by the optimal dosage calculation module 112, format the calculated optimal dosage into a target format, and provide it to the clinician. For example, if a clinician provides cardiovascular index data using a desktop terminal 106a to seek advice on dosage, the recommended dosage output module 114 may return the recommended dosage (i.e., the optimal dosage calculated by the optimal dosage calculation module 112) to the desktop terminal 106a. However, in some embodiments, the clinician may provide cardiovascular index data using a desktop terminal 106a, but may also attempt to receive the recommended dosage on a smartphone 106d. In this situation, the recommended dosage output module 114 may format the recommended dosage into a format compatible with the smartphone 106d and provide the formatted recommended dosage to the smartphone 106d. Furthermore, recommended dosages may be provided in any format, such as displayed in an application window, a browser window, email, text message, and / or any other format.

[0022] Software modules 110, 112, and 114 generally interface with clinicians, while software modules 116, 118, and 120 may generally interface with model developers. That is, model developers may use one or more clients 106 to access these software modules to develop, modify, and / or improve the analytical models described throughout this disclosure.

[0023] The model development module 116 may enable a model developer to develop one or more analysis modules. For example, the model development module 116 may provide a model developer with an interface, such as a graphical user interface, for defining one or more analysis models. Furthermore, the model development module 116 may enable a model developer to upload and / or port predefined analysis modules to the server 102. Thus, the model development module 116 may, in general, provide support for any kind of computing environment for developing the analysis models described throughout this disclosure.

[0024] The simulation module 118 may enable model developers to simulate analysis models (for example, defined / developed / ported using the model development module 116). The simulation may include, for example, numerical simulations, in which case the collected numerical data may be used in the analysis model to observe the output. The simulation may produce, for example, numerical data, graphical data, and / or any other type of output data. The simulation module 118 may generally enable evaluation of the analysis model based on these simulations (for example, to determine whether the analysis model should be run as desired using the simulated scenarios).

[0025] The experimental data acquisition module 120 may acquire experimental data to be used in model development. For example, the experimental data may include measured cardiovascular parameters and / or indicators of animals. Such experimental data may be used by the simulation module 118 to simulate the analysis model. Other experimental data may include detailed experimental data, including both inputs and outputs, which may be used to compare real-world results with simulated results. In yet another example, the experimental data may include a continuous data stream of a patient being treated in a clinical setting, which may be used to continuously modify and / or improve the analysis model. Thus, the experimental data acquisition module 120 may receive any kind of real-world numerical data that may be used to develop, improve and / or modify the analysis modules disclosed throughout this disclosure.

[0026] After one or more analytical models have been developed and validated, a clinician may use a software module 108 in server 102 to assist in the clinical decision-making process for patients with acute heart failure. In exemplary operation, the clinician uses the interface of client 106 to input current cardiovascular indicators, such as the four-dimensional indicators described throughout this disclosure (1. left atrial pressure, 2. cardiac output, 3. mean arterial pressure, and 4. myocardial oxygen consumption). Alternatively, or additionally, the clinician may input current cardiovascular parameters. Client 106 may then transmit the input indicators (and / or parameters) to server 102 via network 104. Server 102 may deploy the software module 108 to calculate a recommended dosage for the current cardiovascular indicator relative to a target cardiovascular indicator (in some embodiments, the clinician may provide the target cardiovascular indicator along with the current cardiovascular indicator). In particular, a cardiovascular indicator data input module 110 may receive and process the input cardiovascular indicators (current and / or target). The optimal dosage calculation module 112 may then develop one or more analytical models based on the input cardiovascular indicators to generate an optimal dosage. The recommended dosage output module 114 may then return the calculated optimal dosage to the clinician who requested it as the recommended dosage. One or more of the model development module 116, simulation module 118, or experimental data acquisition module 120 may, if any, refine one or more of the analytical models using the cycles and feedback described above.

[0027] Figure 2 shows an exemplary analysis model 200 according to an exemplary embodiment of the present disclosure. The exemplary analysis model 200 may be used by the software module 108 described in Figure 1. For example, a model development module 116 may be used to define and / or port the analysis model 200, a simulation module 118 may be used to perform a numerical simulation of the analysis model 200, and an experimental data acquisition module 120 may be used to receive experimental data for validating, modifying, and / or improving the analysis model. Furthermore, a cardiovascular index data input module 110 may receive cardiovascular index data used by the analysis model 200, an optimal dose calculation module 112 may use the analysis model 200 to calculate the optimal dose, and a recommended dose output module 114 may output the optimal dose calculated by the analysis model 200 in a format compatible with the target device / platform. It should be further understood that the analysis model 200 is merely an example and should not be considered limiting, and analysis models having additional, alternative, or fewer steps and / or components should be considered within the scope of the present disclosure. As illustrated, the exemplary analysis model 200 divides the causal chain from drug input 210 to cardiovascular index 212 as a linear drug infusion model 202 to generate cardiovascular parameters 214 from drug input 210, and as a hemodynamic model 204 to generate cardiovascular index 212 from cardiovascular parameters 214. In other words, the drug infusion model 202 and the hemodynamic model 204 divide the causal chain from the drug input 210 to cardiovascular parameters 214, y0=h(x0), labeled as the initial state 206, to the final state 208, labeled as the final state, y0=h(x0), which represents the cardiovascular parameters 214, y0=h(x0), before drug infusion. f =h(x f ) incorporates changes to ). Taking into account all the complexities described throughout this disclosure, the objective is to find the optimal drug input 210 (combination of different drugs) that produces a desired (also called target) final state 208 corresponding to a desired (also called target) cardiovascular index 212.

[0028] Drug input 210 is vector

number

[0029] Cardiovascular parameters 214 include, for example, systemic vascular resistance R s (measured in mmHg seconds / ml), myocardial contractility E es This may include (measured in mmHg / ml), heart rate HR (measured in beats / min), and overload blood volume SBV (measured in ml). Systemic vascular resistance can be defined as the resistance of the vascular system. The vascular system is used to produce blood pressure, for example, when blood vessels constrict. s This can increase. Myocardial contractility may refer to the ability of the heart muscle to contract. Heart rate may refer to the number of heartbeats per unit time, e.g., beats per minute. Loaded blood volume may be defined as any amount of blood exceeding the standard amount of blood required to fill the blood vessels in order to put pressure on them (i.e., the amount of blood without load). In the illustrated analysis model 200, the cardiovascular parameters 214 are,

[0030]

number

[0031] The cardiovascular index 212 can include, for example, the left atrial pressure P LA (measured in mmHg), the cardiac output CO (measured in L / min), the mean arterial pressure MAP (measured in mmHg), the myocardial oxygen consumption MVO2 (measured in mlO2 / min / 100g), etc. The left atrial pressure is generally the pressure generated during the contraction of the left atrium. The cardiac output is generally the amount of blood pumped out of the heart per unit time, which can be expressed as the product of the heart rate and the stroke volume. The mean arterial pressure is generally defined as the average arterial pressure over one cardiac cycle including both the systolic and diastolic phases. The myocardial oxygen consumption is generally an approximation of the amount of oxygen used by the heart, mainly during the contraction of the cardiovascular system. The cardiovascular index is a vector

[0032]

Number

[0033] As described above, the causal chain progresses from the drug input 210 to the cardiovascular parameter 214 and then to the cardiovascular index 212. For the first part of the chain, the cardiovascular parameter 214 is thus a vector function of the drug input, that is,

[0034]

number

[0035] In the second part of the causal chain, cardiovascular parameter 214 (i.e., changes therein) modulates cardiovascular index 212 (or causes hemodynamic changes in cardiovascular index 212). LA The analytical solutions for CO and MAP can be derived using the intersection of the Frank-Stirling curve (known in the art) and Guyton's formula for the venous return curve (known in the art), as follows:

[0036]

number

[0037] The remaining cardiovascular index 212 that can be analytically derived is MVO2. The metabolic demand of the heart itself, as indicated by the MVO2 measurement, is known to be an important indicator for preventing poor prognosis, readmission, cardiovascular adverse events, and / or death in patients with acute heart failure. Embodiments disclosed herein analytically derive, for example, MVO2 for the left ventricle as a function of other cardiovascular index 212 by using the following function, with respect to the mechanical energy generated by ventricular contraction, for example, left ventricular contraction. MVO2(x) = (A0PVA(x) + B0x2 + C0)x3 Here, A0, B0, and C0 are constant parameters shown in Table I below, where PVA(x) (measured at mmHgml / 100g / beat) is the normalized pressure-volume area of ​​the left ventricle per 100g.

[0038] Table I: Constant parameters of the MVO2 model [Table 1]

[0039] Assuming that the weight of the left ventricle is approximately 0.4% of total body weight (BW), the normalized PVA(x) is:

number

[0040] As the drug traverses the entire pathway from the initial state 206 to the final stage 208, the direction and sensitivity of the response to the drug input 210 may differ for each of the cardiovascular indices. For example, the direction of the response to MVO 2216 (generally affecting the heart) may be opposite to that of MAP 218 (generally affecting the body, brain, and kidneys), and P LA The direction of the reaction of (generally affecting the lungs) may be opposite to that of CO22 (generally affecting the body, brain, and kidneys). That is, for example, drug input 210 increases CO22, while P LA 220 can be reduced. Embodiments disclosed herein use an analytical model 200 to analyze different directions of changes in cardiovascular index 212 (i.e., hemodynamic changes) and their sensitivity thereto.

[0041] To determine the direction of hemodynamic changes, from basic calculus, these changes can be expressed as a quantification of small changes dy in cardiovascular indices 212, resulting from small changes dx in cardiovascular parameters 214. Let's represent the cardiovascular indices 212(y) as vector functions.

[0042]

number

[0043]

number

[0044] From the above, the Jacobian matrix (

number

number

[0045]

number

number

[0046] The above analytical derivation shows that the cardiovascular index 212 can be expressed in the form of cardiovascular parameters 214, and the hemodynamics (i.e., dy) of the cardiac index can be expressed in the form of changes in cardiac parameters. The cardiovascular parameters 214 can similarly be controlled by drug input based on the drug infusion model 202. In the case of the drug infusion model, the drug library B can be determined in one or more embodiments as described later.

[0047] For example, drug library B may be determined using animal experiments. In one experiment, dogs were anesthetized and bilateral carotid baroreceptors and vagus trunks were removed. A thoracotomy was performed, and the dogs were then given MAP from the right femoral artery, CO via an ultrasonic flowmeter around the ascending aorta, and P directly in the corresponding atria. LA and right atrial pressure (P RA Both the MAP and the HR were connected to a system for measuring HR. Before drug administration, baseline values ​​were measured for 1 minute. Then, a single drug was administered for 10 minutes, and the pharmacological effect was measured until a steady state was reached. Subsequently, a minimum washout time (i.e., the drug was removed from the dog's body) was ensured until the MAP stabilized at the value before drug administration. These procedures were repeated for other drugs. The drug doses administered in this animal study were dobutamine: 5.0 μg / kg / min, norepinephrine: 0.15 μg / kg / min, sodium nitroprusside: 5.0 μg / kg / min, and dextran: 5.0 ml / kg. It should be further noted that this animal study protocol was approved by the Animal Experiment Committee of the National Cerebral and Cardiovascular Center (NCVC) in Japan.

[0048] Using the above experiment, drug library B was developed. Each drug input u i (Each element of drug input 210) for each cardiovascular parameter x i The gains to changes in each of the 214 cardiovascular parameters were identified by fitting a first-order single-input single-output process model with dead time. In this setting, R s 60 (MAP-P RA ) / CO is calculated, E esThe gains were estimated by a cardiac dynamics-based estimation method, HR was measurable from the sensor, and SBV was the load blood volume estimated by the circulatory equilibrium framework. Aggregating these identified gains, drug library B was as follows (note that it is assumed that furosemide reduces only SBV):

[0049]

number

[0050] Figure 3 shows a flowchart of an exemplary method 300 based on an exemplary embodiment of the present disclosure. The exemplary method 300 can be performed using any part of the analysis model 200 shown in Figure 2, and any combination of components of the computing environment 100 shown in Figure 1. It should be understood that the steps of method 300 are merely examples and should not be considered limiting. Methods having additional, alternative, or fewer steps should be considered within the scope of the present disclosure.

[0051] Method 300 may begin in step 302. In step 302, server 102 may receive input of cardiovascular indicators. The clinician may input current cardiovascular parameters in place of, or in addition to, current cardiovascular indicators. For example, a hospital terminal desktop may be used by the clinician to input cardiovascular indicators and / or cardiovascular parameters. Alternatively, the clinician may input cardiovascular indicators and / or cardiovascular parameters on a smartphone or tablet computer. The cardiovascular indicators may include, for example, current cardiovascular indicators and / or target cardiovascular indicators.

[0052] In step 304, the server 102 may calculate drug combinations based on the input cardiovascular indices and / or cardiovascular parameters. In some embodiments, regardless of the modality of the input of the desired cardiovascular indices and / or cardiovascular parameters, step 304 may be performed to calculate drug combinations for the target cardiovascular indices. Drug combinations may be calculated based on the analytical models described throughout this disclosure.

[0053] In step 306, the server 102 may output drug combinations (e.g., on the requesting device) to support clinical decision-making. That is, clinicians can rely on tested and simulated models to aid decision-making and reduce reliance on mere guesswork.

[0054] As described above, one or more analytical models can be validated using animal experiments. One of the purposes of animal experiments is to verify the direction of hemodynamic changes dy as shown by the above formula. The proposed direction (i.e., the analytically derived direction) was compared with the real-world direction of hemodynamic changes in animal experiments such as those described above. The performance (real-world vs. analysis) was (P LA The cardiac index (CI) was evaluated in space. This is because it is a common index known as the Forrester classification in the treatment of acute heart failure, where CI is the cardiac index defined as CO / BSA (body surface area).

[0055] For this comparative analysis, the real-world response from animal experiments after drug injection was used for P LA_r and CI r Provided by: To evaluate direction (e.g., using angles), P LA_r and CI r Since they have different dimensions and data ranges, they are normalized. Given a time step k that increments every 30 seconds and measured cardiovascular parameters x[k], the normalized vector of the actual response at time k is

[0056]

number

number

number

[0057] In some embodiments, the proposed direction of hemodynamic change can be considered in two different versions. The first version may be the original total differential using the cardiovascular parameter x[k+1] one step ahead, as follows:

[0058]

number

[0059]

number

[0060] dy generated by using cardiovascular parameters one step ahead f [k] may be more accurate, but dy p The use of [k] (i.e., one step prior) may be more practical for exemplary applications that predict the direction of hemodynamics based on past information. Table II below shows the constant parameters used in animal experiments, adjusted for body weight.

[0061] Table II: Constant parameters used in the experiment [Table 2]

[0062] The following evaluation metrics were used to compare real-world directions with analytically derived directions.

[0063]

number

[0064] Figure 4 shows exemplary graphs 402–408 visualizing the direction of hemodynamic changes of real-world drugs and predicted responses based on analytical models according to exemplary embodiments of the present disclosure. In particular, graph 402 shows a comparison of real-world and predicted responses for dobutamine (DOB), graph 404 shows a comparison of real-world and predicted responses for norepinephrine (NE), graph 406 shows a comparison of real-world and predicted responses for sodium nitroprusside (SNP), and graph 408 shows a comparison of real-world and predicted responses for dextran (DEX). Graphs 402, 404, 406, and 408 all show the direction of hemodynamic changes every 30 seconds. As shown, method dy f [k] and dy r [k] can accurately predict the direction of hemodynamics at each time step k.

[0065] For angular error comparisons, Table III shows the results of angular error comparisons between real-world reactions and reactions based on analytical models (using total differentials). The values ​​shown are the mean and confidence interval (CI). The angular error θ is the average of four drug injection experiments. rf This becomes 18.85 degrees. In actual prediction tasks where future information is not available, the mean error θ rp This becomes 45.5 degrees.

[0066] Table III: Results of angular error in Experiment 1 (Reality vs. the present invention) [Table 3]

[0067] Further simulations were performed to evaluate the disclosed analytical system, including the drug infusion model x=f(u) with an identified drug library, and the hemodynamic analysis solution y=h(x)=h(f(u)) and the direction of its change dy. For example, three simulation studies were conducted to evaluate the analytical model in both qualitative and quantitative ways. The three simulations included 1) single drug infusion, 2) three acute heart failure scenarios, and 3) recommended drug therapy according to the Forrester classification. All three simulations are described in detail below.

[0068] 1. Single drug infusion: Simulation experiments of single drug infusion are generally conducted using single drug infusion u i This involves the direction of change in the four-dimensional hemodynamics y (i.e., P) of a single drug infusion. LA Visualize how it affects CO, MAP, MVO2. The single dose injected is set as follows: DOB = u1 = 2 μg / kg / min, NE = u2 = 0.15 μg / kg / min, SNP = u3 = 2 μg / kg / min, DEX = u4 = 75 ml, and FRO = u5 = 5 mg. Different combinations of cardiovascular parameters are shown for simulated acute heart failure patients. s =[1.0,3.0,5.0], E es It was formed with =[6.0,12.0], HR=[60,120], and SBV=[100:150:700].

[0069] Figures 5A–5E show exemplary graphs 502–520 visualizing a single drug infusion simulation based on exemplary embodiments of the present disclosure. As shown in graphs 502–520, cardiovascular indices (e.g., cardiovascular index 212) are divided into two spaces, (i) LAP-CI (Forrester classification) and (ii) MAP-MVO2. In each of graphs 502–520, each filled marker represents the initial cardiovascular index (consistent with the visualization shown in Figure 2) y0 of a different acute heart failure patient, the arrow indicates the direction of hemodynamic change dy, and the dashed line represents the time transient response estimated using a Bézier curve. Each unfilled marker represents the steady state y f The final result (which is consistent with the visualization shown in Figure 2) is shown below.

[0070] As shown in Graphs 502, 504, 506, and 508, cardiac stimulants (dobutamine, norepinephrine) improve CI and MAP but significantly increase MVO2. As shown in Graphs 510, 512, 518, and 520, sodium nitroprusside and furosemide improve P LA It reduces both CI and MVO2. As shown in Graphs 514 and 516, dextran reduces CI and P LA Increase both. LA While MAP and MVO2 can be measured in clinical settings, MVO2 cannot be measured directly. Therefore, the visualization provided by graphs 502-520 can offer meaningful insights for clinicians.

[0071] 2. Three Acute Heart Failure Scenarios: This simulation visualizes three specific clinical scenarios, generally within subsets II, III, and IV of the Forrester classification. Three representative patients with the following cardiovascular parameters were considered. (wet-warm)x II =[3.0,15,80,600] T (dry-cold)x III =[4.0,10,100,200] T (wet-cold)x IV =[8.0,8.0,100,450] T Subsequently, each drug infusion was simulated. The dosage was set to the same amount as mentioned in the single drug infusion simulation described above.

[0072] Figure 6 shows exemplary graphs 602–604 visualizing simulations of three acute heart failure scenarios based on exemplary embodiments of the present disclosure. In both exemplary graphs 602 and 604, the initial directional and time-transient responses are consistent with the expected values ​​of the general pharmacological effect. It is further shown that sensitivity may differ depending on the patient's condition, even with the same drug and the same dosage. For example, DOB administration with a dosage of 2.0 μg / kg / min may improve CI in all scenarios, and the change in MVO2 in subset IV is particularly significant compared to the other scenarios, as shown below. (wet-warm)x II =+8.88mlO2 / min / 100g (dry-cold)x III =+5.87mlO2 / min / 100g (wet-cold)x IV =+11.3mlO2 / min / 100g

[0073] The mechanism of this nonlinear behavior arises from the gradient of MVO2(x) shown in the Jacobian matrix above. This nonlinear behavior indicates that oxygen consumption is greater in the hearts of patients in subset IV than in other patients, for the same dose of DOB.

[0074] 3. Recommended drug therapy according to the Forrester classification: This third simulation shows the direction of the analytically predicted hemodynamic changes when the recommended drug therapy is applied to various patients with acute heart failure (P LA Designed to verify whether the CI is moving towards a desired range within the space. For simulated acute heart failure patients, different combinations of cardiovascular parameters are used in R s =[1.0:1.0:5.0], Ees The combination was formed with ∇[3.0:2.0:15.0], HR=[60:20:150], and SBV=[100:50:700]. Unrealistic acute heart failure patients with MAP(x)≦44mmHg or 170mmHg≦MAP(x) were excluded. Including all selections and exclusions, 963 different acute heart failure patients were simulated. To treat these simulated acute heart failure patients, drug combinations and dosages were selected based on Forrester classification guidelines. • Subset II:u II =[0.0,0.0,5.0,0.0,1.5] T • Subset III:u III =[5.0,0.15,0.0,50.0,0] T • Subset II:u Iv =[5.0,0.0,5.0,0.0,1.5] T Here, the indicator for evaluation is the extension in the direction predicted from the initial state (e.g., initial state 206 in Figure 2), subset I: 3.0 ≤ P LA ≤17.0 mmHg and 2.25 ≤ CI ≤ 4.5 L / min / m 2 This determines whether the target area falls within the normal range of the normal range.

[0075] Figure 7 shows an exemplary graph 700 visualizing a simulation using the Forrester classification according to an exemplary embodiment of the present disclosure. The graph shows that out of 963 heart failure patients, 779 patients successfully moved toward the target region (i.e., an 80.9% success rate). This high success rate was achieved despite the drug inputs being simply fixed for each subset. Graph 700 generally shows that the orientation of the analytically derived and simulated directions largely corresponds to the results expected in clinical guidelines. Thus, those skilled in the art will understand that the analytical model disclosed herein, including the identified drug library B and the corresponding analytical solutions, is reasonably correct.

[0076] Therefore, the embodiments disclosed herein describe how each drug undergoes cardiac metabolism (indicated by MVO2) and (CI, MAP, P LA This allows for simultaneous visualization of how the cardiovascular system is affected (as shown by...). For example, inotropic agents such as catecholamines maintain sufficient hemodynamic stability and prevent hyperperfusion or pulmonary congestion. However, because these drugs cause metabolic stress and arrhythmias in cardiac suppression, increasing the dosage may worsen the prognosis of patients with cardiogenic shock. This is just one example, and determining the optimal drug dosage in clinical practice is extremely difficult.

[0077] As a solution to this problem, embodiments disclosed herein provide systems, methods, and devices that assist clinicians in making clinical decisions regarding optimal drug combinations. For example, as described above, the systemic and cardiac effects of various drugs can be visualized. Furthermore, embodiments disclosed herein also reveal variations in sensitivity to drugs depending on the patient's cardiovascular parameters, even at the same dosage. Embodiments disclosed using gradient analysis (e.g., by using the Jacobian matrix described above) enable the quantification of immediate changes caused by drugs and provide support for real-time clinical decision-making.

[0078] Furthermore, as stated above, the drug combinations disclosed herein are merely examples and should not be considered limiting. That is, the embodiments disclosed herein may be applicable to any type of drug. For example, since early administration of beta-blockers has been reported to improve prognosis and prevent sudden death resulting from arrhythmias, the embodiments are applicable to at least beta-blockers. In addition, the use of pure bradycardia agents (ivabradine) in combination with conventional drugs has been reported to improve the performance of simultaneous four-dimensional hemodynamic control in dogs with acute heart failure. Therefore, the embodiments disclosed herein are equally applicable to any type of drug used to improve the prognosis of patients suffering from acute heart failure.

[0079] Figure 8 shows a block diagram of an exemplary computing device 800 that performs various features and processes according to exemplary embodiments of the present disclosure. For example, in some embodiments, the computing device 800 may function as a server 102 and a client 106, or as part of or in combination thereof. Furthermore, the computing device 800 may host and deploy the analysis model 200, either partially or entirely. The computing device 800 may also perform one or more steps of Method 300. The computing device 800 is implemented on any electronic device that runs a software application derived from compiled instructions, including but not limited to personal computers, servers, smartphones, media players, electronic tablets, game consoles, and email devices. In some embodiments, the computing device 800 includes one or more processors 802, one or more input devices 804, one or more display devices 806, one or more network interfaces 808, and one or more computer-readable media 812. Each of these components is coupled by a bus 810.

[0080] The display device 806 includes any display technology, including but not limited to liquid crystal display (LCD) or light-emitting diode (LED) technology. The processor 802 uses any processor technology, including but not limited to graphics processors and multicore processors. The input device 804 includes any known input device technology, including but not limited to keyboards (including virtual keyboards), mice, trackballs, and touch-sensitive pads or displays. The bus 810 includes any internal or external bus technology, including but not limited to ISA, EISA, PCI, PCI Express, USB, Serial ATA, or FireWire. The computer-readable medium 812 includes any non-temporary computer-readable medium that provides instructions to the processor 802 for execution, including but not limited to non-volatile storage media (e.g., optical disks, magnetic disks, flash drives, etc.) or volatile media (e.g., SDRAM, ROM, etc.).

[0081] The computer-readable medium 812 contains various instructions 814 for implementing an operating system (e.g., Mac OS®, Windows®, Linux). The operating system may be multi-user, multi-processing, multi-tasking, multi-threaded, real-time, etc. The operating system performs basic tasks including, but not limited to, recognizing input from input device 804, sending output to display device 806, tracking files and directories on computer-readable medium 812, controlling peripheral devices (e.g., disk drives, printers, etc.) that can be controlled directly or through an I / O controller, and managing traffic on bus 810. Network communication instructions 816 establish and maintain network connections (e.g., software for implementing communication protocols such as TCP / IP, HTTP, Ethernet, telephone, etc.).

[0082] The clinical decision assistant 818 includes instructions for performing the disclosed processes to assist in clinical decision-making, as described throughout this disclosure. The application 820 may include applications that use or perform the processes and / or other processes described herein. These processes may also be performed by an operating system.

[0083] The features described may be implemented in one or more computer programs that can be run on a programmable system which includes at least one programmable processor coupled to receive data and instructions from a data storage system, at least one input device, and at least one output device, and to send data and instructions to the data storage system, at least one input device, and at least one output device. A computer program is a set of instructions that can be used directly or indirectly in a computer to perform an activity or to produce a result. A computer program may be written in any form of programming language (e.g., Objective-C, Java), including compiled or interpreted languages, and it may be deployed in any form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. In one embodiment, this may include Python. Thus, a computer program is multilingual.

[0084] Processors suitable for executing instruction programs may include, for example, both general-purpose and dedicated microprocessors, as well as one of a single processor or multiple processors or cores in any type of computer. Generally, a processor may receive instructions and data from read-only memory or random-access memory, or both. Essential elements of a computer may include a processor for executing instructions, and one or more memories for storing instructions and data. Generally, a computer may also include one or more large storage devices for storing data files, or may be operably coupled to them for communication. Such devices include magnetic disks such as internal hard disks and removable disks, magneto-optical disks, and optical disks. Storage devices suitable for tangibly embodying computer program instructions and data may include, for example, semiconductor memory devices such as EPROMs, EEPROMs, and flash memory devices, magnetic disks such as internal hard disks and removable disks, magneto-optical disks, and all forms of non-volatile memory, including CD-ROMs and DVD-ROM disks. Processors and memories may be assisted by or incorporated into ASICs (Application-Specific Integrated Circuits).

[0085] To provide user interaction, the features may be implemented on a computer having a display device such as a CRT (cathode ray tube) or LCD (liquid crystal display) monitor for displaying information to the user, and a pointing device such as a keyboard and mouse or trackball that the user can use to provide input to the computer.

[0086] The features can be implemented in a computer system or any combination thereof, including backend components such as data servers, middleware components such as application servers or internet servers, or frontend components such as client computers having a graphical user interface or internet browser. The components of the system can be connected by any form or medium of digital data communication, such as a communication network. Examples of communication networks include, for example, telephone networks, LANs, WANs, and networks that form computers and the internet.

[0087] A computer system may include clients and servers. Clients and servers may generally be remote from each other and typically interact over a network. The relationship between a client and a server may arise from computer programs running on each computer that have a client-server relationship with each other.

[0088] One or more features or steps of the disclosed embodiments may be implemented using an API. The API may define one or more parameters that are passed between a calling application and other software code (e.g., an operating system, library routines, functions) that provides services, provides data, or performs actions or calculations.

[0089] An API may be implemented as one or more calls in program code that send or receive one or more parameters through a parameter list or other structure based on the calling convention defined in the API specification document. Parameters may be constants, keys, data structures, objects, object classes, variables, data types, pointers, arrays, lists, or other calls. API calls and parameters may be implemented in any programming language. The programming language may define the vocabulary and calling conventions that programmers employ to access the features supporting the API.

[0090] In some embodiments, an API call may report to the application the capabilities of the device running the application, such as input capabilities, output capabilities, processing capabilities, power supply capabilities, and communication capabilities.

[0091] Additional examples of embodiments of the methods and devices described herein are suggested in accordance with the structures and techniques described herein. Other non-limiting examples may be configured to operate separately or may be combined in any permutation or combination with any one or more of the other examples provided above or throughout this disclosure.

[0092] Those skilled in the art will understand that this disclosure can be embodied in other specific forms without departing from the spirit or essential characteristics of this disclosure. Therefore, the embodiments disclosed herein are intended to be illustrative and not limiting in all respects. The scope of this disclosure is indicated by the appended claims rather than the foregoing description, and all modifications within the meaning, scope, and equivalents thereof are intended to be contained therein.

[0093] Please note that the terms “including” and “comprising” should be interpreted as “including but not limited to.” Unless explicitly stated in the claim, the term “a” should be interpreted as “at least one,” and “the,” “said,” etc., should be interpreted as “the at least one,” “said at least one,” etc. Furthermore, the applicant intends that only claims containing the explicit phrase “means for” or “step for” should be interpreted under § 112(f) of the U.S. Patent Act. Claims that do not explicitly contain the phrase “means for” or “step for” should not be interpreted under § 112(f) of the U.S. Patent Act.

Claims

1. It is a system, A non-temporary storage medium for storing computer program instructions, The system comprises one or more processors configured to execute computer program instructions for causing an action, wherein the action is Receiving multiple current cardiovascular indicators from patients with acute heart failure, To generate a recommended drug combination for achieving multiple target cardiovascular indicators, the clinical decision support module is executed based on the multiple current cardiovascular indicators, wherein the clinical decision support module is: A drug library that linearly models pathways from multiple current cardiovascular parameters associated with the multiple current cardiovascular indicators to multiple target cardiovascular parameters associated with the multiple target cardiovascular indicators, and A mapping between the plurality of target cardiovascular indices and the plurality of target cardiovascular parameters, the mapping including the transient response of one or more cardiovascular indices due to changes in one or more cardiovascular parameters caused by the administration of one or more drugs, Based on the above, the clinical decision support module is executed, Outputting the aforementioned recommended drug combination, A system that includes this.

2. The system according to claim 1, wherein the plurality of current cardiovascular indicators and the plurality of target cardiovascular indicators include one or more of left atrial pressure, cardiac output, mean arterial pressure, or myocardial oxygen consumption.

3. The system according to claim 1, wherein the plurality of current cardiovascular parameters and the plurality of target cardiovascular parameters include one or more of systemic vascular resistance, myocardial contractility, heart rate, or load blood volume.

4. The system according to claim 1, wherein the one or more drugs include at least one of a cardiotonic, a vasoconstrictor, a vasodilator, a fluid, or a diuretic.

5. The system according to claim 1, wherein the plurality of current cardiovascular indicators and the plurality of target cardiovascular indicators include myocardial oxygen consumption.

6. The system according to claim 1, wherein the plurality of current cardiovascular indicators and the plurality of target cardiovascular indicators are expressed as a function of myocardial oxygen consumption as one or more cardiovascular indicators.

7. The system according to claim 1, wherein the mapping further includes the sensitivity of the time-transient response of the one or more cardiovascular indices to the changes in the one or more cardiovascular parameters caused by the administration of the one or more drugs.

8. The system according to claim 7, wherein the mapping includes a Jacobian matrix.

9. The system according to claim 8, wherein the columns of the Jacobian matrix indicate the sensitivity to a change in a single cardiovascular parameter that affects multiple cardiovascular indicators.

10. The system according to claim 8, wherein the rows of the Jacobian matrix represent the sensitivity of changes in multiple cardiovascular parameters that affect a single cardiovascular index.

11. A computer implementation method, The computing system receives multiple current cardiovascular indicators from a client device for a patient with acute heart failure, The computing system executes a clinical decision support module based on the multiple current cardiovascular indicators in order to generate a recommended drug combination to reach multiple target cardiovascular indicators, wherein the clinical decision support module A drug library that linearly models pathways from multiple current cardiovascular parameters associated with the multiple current cardiovascular indicators to multiple target cardiovascular parameters associated with the multiple target cardiovascular indicators, and A mapping between the plurality of target cardiovascular indices and the plurality of target cardiovascular parameters, the mapping including the transient response of one or more cardiovascular indices due to changes in one or more cardiovascular parameters caused by the administration of one or more drugs, Based on the above, the clinical decision support module is executed, The computing system outputs the recommended drug combination to the client device, A computer implementation method, including

12. The computer-assisted method according to claim 11, wherein the plurality of current cardiovascular indicators and the plurality of target cardiovascular indicators include one or more of left atrial pressure, cardiac output, mean arterial pressure, or myocardial oxygen consumption.

13. The computer-assisted method according to claim 11, wherein the plurality of current cardiovascular parameters and the plurality of target cardiovascular parameters include one or more of systemic vascular resistance, myocardial contractility, heart rate, or load blood volume.

14. The computer-assisted method according to claim 11, wherein the one or more drugs include at least one of a cardiotonic, a vasoconstrictor, a vasodilator, a fluid, or a diuretic.

15. The computer-aided method according to claim 11, wherein the plurality of current cardiovascular indicators and the plurality of target cardiovascular indicators include myocardial oxygen consumption.

16. The computer-aided method according to claim 11, wherein the plurality of current cardiovascular indicators and the plurality of target cardiovascular indicators are expressed as a function of myocardial oxygen consumption as a function of one or more cardiovascular indicators.

17. The computer-aided method according to claim 11, wherein the mapping further includes the sensitivity of the transient response of the one or more cardiovascular indices to the changes in the one or more cardiovascular parameters caused by the administration of the one or more drugs.

18. The computer implementation method according to claim 17, wherein the mapping includes a Jacobian matrix.

19. The computer-aided method according to claim 18, wherein the columns of the Jacobian matrix represent the sensitivity of a change in a single cardiovascular parameter that affects multiple cardiovascular indicators, and the rows of the Jacobian matrix represent the sensitivity of a change in multiple cardiovascular parameters that affect a single cardiovascular indicator.

20. A non-temporary storage medium for storing computer program instructions, which, when executed by one or more processors, Receiving multiple current cardiovascular indicators from patients with acute heart failure, To generate a recommended drug combination for achieving multiple target cardiovascular indicators, the clinical decision support module is executed based on the multiple current cardiovascular indicators, wherein the clinical decision support module is: A drug library that linearly models pathways from multiple current cardiovascular parameters associated with the multiple current cardiovascular indicators to multiple target cardiovascular parameters associated with the multiple target cardiovascular indicators, and A mapping between the plurality of target cardiovascular indices and the plurality of target cardiovascular parameters, the mapping including the transient response of one or more cardiovascular indices due to changes in one or more cardiovascular parameters caused by the administration of one or more drugs, Based on the above, the clinical decision support module is executed, Outputting the aforementioned recommended drug combination, A non-temporary storage medium that triggers actions including [specific actions].

Images