System and method for estimating the value of a target cardiac parameter

Abstract

A technique for determining one or more sets of values ​​for target cardiac parameters of a subject. The cardiovascular model produces at least one set of output values ​​(for at least one respective characteristic of cardiac function) by processing the model parameters. One or more sets of measurements corresponding to non-invasively measurable characteristics of cardiac function are processed with the corresponding sets of output values ​​to modify the model parameters. This modification continues until some predefined criteria are met such that the model more closely matches a true representation of the subject's cardiac function. One or more sets of values ​​are then derived from the outputs of the cardiovascular model and / or from the model parameters of the cardiovascular model.

Description

[Technical Field]

[0001] The present invention relates to the field of cardiac parameters, and in particular to estimating the value of a target cardiac parameter. [Background technology]

[0002] Cardiac parameters such as cardiac filling pressures are important measurements for both diagnostic (e.g., heart failure with preserved ejection fraction, HFpEF) and monitoring purposes, providing information about left and right cardiac function and cardiovascular and cardiopulmonary interactions. High cardiac filling pressures are associated with myocardial dysfunction and / or hypervolemia, and subjects with high cardiac filling pressures tend to have poor long-term survival.

[0003] Cardiac filling pressure measurements are typically obtained by left heart catheterization for left ventricular pressure or left ventricular filling pressure (LVFP) and right heart catheterization for pulmonary capillary wedge pressure (PCWP). The invasive nature of catheterization means that traditional techniques for obtaining these measurements are uncomfortable, expensive, time-consuming, and pose additional risks to the patient. Furthermore, scheduling the acquisition of these measurements in the heart failure diagnostic process is logistically difficult. Because of these drawbacks, cardiac filling pressures have historically been used in the diagnosis of heart failure only when there is diagnostic ambiguity.

[0004] Various studies have explored the relationship between noninvasive measurements and such cardiac filling pressure measurements. However, the reproducibility of identified relationships in the literature has proven elusive, thus calling into question the reliability, robustness, or validity of statistical models utilizing such relationships. Such existing statistical models are likely to be highly dependent on the patient population and user / measurement accuracy. Summary of the Invention [Problem to be solved by the invention]

[0005] Therefore, there is a need for improved methods for estimating the value of target cardiac parameters such as left ventricular filling pressure. [Means for solving the problem]

[0006] The invention is defined by the claims.

[0007] According to an example according to an aspect of the present invention, there is provided a processing system for estimating a set of one or more values ​​of target cardiac parameters of a subject.

[0008] The processing system includes: acquiring a model of the cardiovascular system using a plurality of model parameters to produce output data including a plurality of output value sets, each output value set associated with a different non-invasively measurable characteristic of cardiac function, wherein each output value set includes at least one value of the associated non-invasively measurable characteristic of cardiac function, wherein the plurality of model parameters include a target cardiac parameter and / or the output data includes a further output value set of the target cardiac parameter; and acquiring cardiac data including a plurality of measurement sets, wherein each measurement set is associated with a different output value set and includes one or more measurements of the non-invasively measurable characteristic associated with the associated output value set. the output data includes a set of one or more predetermined output values; iteratively modifying values ​​of a plurality of model parameters of the model until one or more predetermined criteria are met, thereby correcting differences between a plurality of sets of output values ​​of the output data and corresponding measurement sets of the cardiac data; and after the iterative modification, if the plurality of model parameters includes a target cardiac parameter, defining the set of one or more values ​​of the modified target cardiac parameter as a set of one or more estimated values ​​for the target cardiac parameter, or if the output data includes a further set of output values, defining the further set of output values ​​as a set of one or more estimated values ​​for the target cardiac parameter.

[0009] The inventors have recognized that a model of the cardiovascular system (CVS) can be used to non-invasively estimate one or more values ​​of target cardiac parameters. The model is personalized to a subject's cardiac function by fitting model parameters based on non-invasively obtained measurements of the subject. This avoids the need for cardiac catheterization, thereby reducing the risk, cost, and time required to obtain left ventricular filling pressure.

[0010] The CVS model is a biophysical model of (part of) the function of the heart and circulatory system. For a defined set of values ​​for each of a plurality of model parameters, the CVS model outputs a solution (i.e., output data) that includes a set of output values ​​for a plurality of non-invasively measured parameters. The values ​​of the plurality of model parameters are modified until the set of output values ​​matches (e.g., within a predetermined error or certainty) the set of measurements included in the subject's cardiac data.

[0011] A non-invasive measurement indicator of cardiac function is a parameter indicative of cardiac function that is measured using a non-invasive method (e.g., echocardiography, cardiac MRI, cuff pressure measurement, etc.) Techniques for obtaining such measurements non-invasively are well known and established in the art.

[0012] After the iterative refinement, the refined value set for each model parameter is thereby personalized or adapted to the subject under study, i.e., subject-specific, meaning that a target cardiac parameter value set is obtained, which is included in the output data provided by the CVS model and / or the model parameters of the CVS model.

[0013] The target cardiac parameter is left ventricular pressure. This embodiment provides a particularly useful scenario for using the proposed approach, as existing methods for determining left ventricular pressure rely on invasive measurements, and it would be preferable (for improved subject assessment) to generate an accurate measurement of this value without the need for surgical intervention.

[0014] The plurality of non-invasively measurable characteristics includes left ventricular volume and at least one peripheral pressure characteristic, which alone are sufficient to determine the values ​​of the model parameters that most closely correspond to the subject's cardiac data.

[0015] Left ventricular volume is measured, for example, on the basis of 2D echocardiography (using the method of disks), 3D echocardiography, or cardiac MRI.

[0016] Peripheral pressure is measured, for example, by performing cuff pressure measurements during diastole and systole (e.g., to measure pressure in the upper arm, finger, neck, toe, ankle, and / or leg) or using ultrasound (e.g., to measure aortic pressure).

[0017] The at least one peripheral pressure measurement includes a pressure measurement on the subject's arm, leg, wrist, neck, foot, finger, or toe. In one example, the peripheral pressure measurement is, for example, a brachial cuff pressure (e.g., measured on the subject's arm). In another example, the peripheral pressure measurement is a measurement taken on the subject's neck, for example, carotid artery pressure.

[0018] In some examples, peripheral pressure measurements are converted to aortic pressure using, for example, one or more transfer functions.

[0019] The plurality of non-invasively measurable indicators include at least one of aortic valve flow, mitral valve flow, and / or timing of cardiac cycle events.

[0020] These measurements are used to improve the accuracy of the estimate, for example, compared to left ventricular volume and at least one peripheral pressure measurement alone.

[0021] Aortic and / or mitral flow measurements are determined, for example, using cardiovascular MRI or based on Doppler measurements and valve opening area.

[0022] The timing of cardiac cycle events may include, for example, timing of mitral valve opening, timing of mitral valve closing, timing of aortic valve opening, timing of aortic valve closing, and / or timing of the onset of left ventricular contraction, such measurements being determined from the ECG data.

[0023] The processing system is configured to iteratively define a solution vector including a plurality of sets of output values ​​for a model of the cardiovascular system, define a measurement data vector including a plurality of measurement sets, define a cost function that quantifies a difference between the solution vector and the measurement data vector, determine a value of the cost function C, and modify the plurality of model parameters based on the value of the cost function until one or more predetermined criteria are met.

[0024] The values ​​of the model parameters at which the cost function is minimized are those at which the output of the CVS model most closely matches the subject's cardiac data.

[0025] The processing system is configured to iteratively determine the value of the cost function and modify the multiple model parameters by: setting an initial value set for each of the multiple model parameters; determining multiple output value sets for the cardiovascular model based on the current value sets for each of the multiple model parameters; calculating a value of a cost function based on the determined multiple output values; modifying the value set for each of the multiple model parameters based on the calculated value of the cost function; and iteratively repeating the steps of determining the multiple output value sets, calculating the value of the cost function, and modifying the value set for each of the multiple model parameters until one or more predetermined criteria are met. In other words, a simultaneous approach is used to minimize the cost function. This allows subject-specific values ​​to be efficiently identified. The values ​​of the model parameters are adjusted using, for example, a gradient-based minimization method (e.g., BFGS method) or a non-gradient-based method (e.g., particle swarm optimization or Bayesian optimization).

[0026] In some examples, the processing system is configured to iteratively modify values ​​of a plurality of model parameters of the model by performing the iterative steps of dividing a plurality of model parameters into a plurality of sets of one or more model parameters; setting an initial value set for each of the plurality of model parameters; and, for each set of one or more model parameters, sequentially determining one or more sets of output values ​​of the model of the cardiovascular system based on at least a current value set of the set of model parameters; calculating a value of a cost function based on the determined one or more sets of output values ​​and a measurement set associated with the one or more sets of output values; and adjusting the value set of each model parameter of the set of model parameters based on the calculated value of the cost function, wherein the iterative step is repeated until a predefined convergence condition is met.

[0027] The entire process of iteratively modifying each set of one or more model parameters in turn is repeated iteratively until one or more predetermined criteria are met.

[0028] The processing system may include: dividing the plurality of model parameters into a plurality of sets of one or more model parameters by dividing the plurality of model parameters into a first set of model parameters and a second set of model parameters; determining a plurality of output value sets for the model of the cardiovascular system based on the current value sets of each of the plurality of model parameters; calculating values ​​of a cost function based on the determined plurality of output value sets; adjusting the value set of each of the first set of model parameters based on the calculated values ​​of the cost function; and iteratively repeating the steps of determining the plurality of output value sets, calculating values ​​of the cost function, and adjusting the value set of each of the first model parameters until a first predetermined criterion is met. and modifying the value set of each of the second set of model parameters by: determining a plurality of output value sets of the model of the cardiovascular system based on the current value sets of each of the plurality of model parameters; calculating a value of a cost function based on the determined plurality of output values; adjusting the value set of each of the second set of model parameters based on the calculated value of the cost function; and iteratively repeating the steps of determining the plurality of output value sets, calculating the value of the cost function, and adjusting the value set of each of the second model parameters until a second predetermined criterion is met; and iteratively repeating the steps of modifying the value set of each of the first set of model parameters and modifying the value set of each of the second set of model parameters until one or more predetermined criteria are met.

[0029] In other words, a bilevel optimization technique is used to minimize the cost function, optimizing different groups of parameters independently of each other. The use of a bilevel optimization algorithm reduces the chance that the optimization process will get stuck in a local minimum on the cost function hypersurface and therefore will not identify the global minimum.

[0030] The plurality of model parameters includes a target cardiac parameter, and the processing system is configured to divide the plurality of model parameters into a first set of model parameters and a second set of model parameters by defining a first set of model parameters to include only the target cardiac parameter and defining a second set of model parameters to include the remaining model parameters, in this manner, the target cardiac parameter is optimized independently of the other model parameters.

[0031] The processing system is configured to perform a sensitivity analysis on the model of the cardiovascular system and divide the plurality of model parameters into a first set of model parameters and a second set of model parameters by dividing the plurality of model parameters into a first set and a second set based on the sensitivity analysis.

[0032] The results of the sensitivity analysis are used to distinguish model parameters that have a strong influence on the cost function from model parameters that have a weak influence, for example, a first set of model parameters includes model parameters that have a strong influence and a second set of model parameters includes model parameters that have a weak influence.

[0033] The one or more predetermined criteria include at least one of a determination that a difference in a set of values ​​for each of a plurality of model parameters between the current iteration and the immediately preceding iteration is less than a predetermined threshold, a determination that a difference in a value of a cost function between the current iteration and the immediately preceding iteration is less than a predetermined threshold, a determination that the number of iterations exceeds a predetermined threshold, a determination that the amount of time spent performing the iterative corrections exceeds a predetermined amount of time, and / or a determination that a difference between each of a plurality of sets of output values ​​for the model of the cardiovascular system and a corresponding measurement set is less than the uncertainty of the corresponding measurement set.

[0034] Each of these conditions is used to identify values ​​of several model parameters at which the cost function is thought to be minimized.

[0035] The processing system is configured to iteratively modify the plurality of model parameters by: defining a plurality of quantities of interest, the quantities of interest being sets of output values ​​of a model of the cardiovascular system corresponding to predetermined characteristics; identifying measurement sets within the cardiac data that correspond to the quantities of interest; and iteratively modifying the plurality of model parameters of the model until one or more predetermined criteria are met, thereby correcting for differences between the quantities of interest and the identified measurement sets.

[0036] Models of the cardiovascular system are zero-dimensional or one-dimensional models. These low-dimensional CVS models, such as the heart, perform well and have a relatively small number of model parameters, which allows them to be solved efficiently.

[0037] The processing system is further configured to, after the iterative modification, determine a plurality of sets of output values ​​of the model of the cardiovascular system based on the modified set of values ​​for each of the plurality of model parameters, and generate one or more cardiac function curves based on the determined plurality of sets of output values. The cardiac function curves, such as pressure, volume, and flow time traces and pressure-volume loops, provide additional clinically useful information.

[0038] A computer-implemented method for estimating left ventricular filling pressure in a subject is further proposed, the computer-implemented method comprising the steps of: acquiring a model of the cardiovascular system using a plurality of model parameters to produce output data comprising a plurality of output value sets, each output value set associated with a different non-invasively measurable characteristic of cardiac function, wherein each output value set comprises at least one value of the associated non-invasively measurable characteristic of cardiac function, the plurality of model parameters comprising a target cardiac parameter and / or the output data comprising a further output value set of the target cardiac parameter; and acquiring cardiac data comprising a plurality of measurement sets, each measurement set associated with a different output value set and comprising one or more of the non-invasively measurable characteristics associated with the associated output value set. The method includes steps of obtaining, including measurements, iteratively modifying values ​​of a plurality of model parameters of the model until one or more predetermined criteria are met, thereby correcting differences between a plurality of sets of output values ​​of the output data and corresponding sets of measurements of the cardiac data, and, if the plurality of model parameters include a target cardiac parameter after the iterative modification, defining a set of one or more values ​​of the modified target cardiac parameter as a set of one or more estimated values ​​for the target cardiac parameter, or if the output data includes a further set of output values, defining a further set of output values ​​as a set of one or more estimated values ​​for the target cardiac parameter.

[0039] The processing systems described herein may be adapted to perform any variation of the methods described herein, and vice versa.

[0040] It is further proposed a computer program product comprising code means which, when executed on a computer having a processing system, causes the processing system to perform all of the steps of the methods described herein.

[0041] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0042] For a better understanding of the present invention, and to show more clearly how the same may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings in which: [Brief explanation of the drawings]

[0043] [Figure 1] FIG. 1 illustrates a system for estimating left ventricular filling pressure in a subject, according to an embodiment of the present invention. [Figure 2] FIG. 1 shows a graphical representation of an exemplary model of the cardiovascular system for use in the present invention. [Figure 3] FIG. 10 illustrates an example cost function as a weighted sum of the areas between the curves for each of a plurality of non-invasively measurable indicators. [Figure 4] FIG. 10 shows the point-by-point difference between the measured cuff pressure and the output of the CVS model during systole and diastole. [Figure 5] FIG. 2 illustrates a method for iteratively determining a cost function value and modifying multiple model parameters according to one embodiment of the present invention. [Figure 6] FIG. 10 illustrates an alternative method for iteratively determining values ​​of a cost function CF and modifying multiple model parameters according to another embodiment of the present invention. [Figure 7] FIG. 10 illustrates exemplary results of a Sobol sensitivity analysis. [Figure 8] FIG. 7 illustrates substeps of step 630 of the method of FIG. 6. [Figure 9] FIG. 7 illustrates substeps of step 640 of the method of FIG. 6. [Figure 10] FIG. 10 illustrates an alternative method for iteratively determining values ​​of a cost function CF and modifying multiple model parameters according to another embodiment of the present invention. [Figure 11] FIG. 1 illustrates a computer-implemented method for estimating left ventricular filling pressure in a subject, according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0044] The present invention will now be described with reference to the figures.

[0045] It should be understood that the detailed description and specific examples, while indicating exemplary embodiments of the devices, systems, and methods, are for purposes of illustration only and are not intended to limit the scope of the invention. These and other features, aspects, and advantages of the devices, systems, and methods of the present invention will become better understood from the following description, the appended claims, and the accompanying drawings. It should be understood that the figures are merely schematic and not drawn to scale. It should also be understood that the same reference numerals are used throughout the figures to indicate the same or similar parts.

[0046] According to the inventive concept, a technique is proposed for determining one or more sets of values ​​for target cardiac parameters of a subject. A cardiovascular model produces at least one set of output values ​​(for at least one respective characteristic of cardiac function) by processing model parameters. One or more sets of measurements corresponding to non-invasively measurable characteristics of cardiac function are processed with the corresponding sets of output values ​​to modify the model parameters. This modification continues until some predetermined criteria are met so that the model more closely matches a true representation of the subject's cardiac function. One or more sets of values ​​are then derived from the outputs of the cardiovascular model and / or from the model parameters of the cardiovascular model.

[0047] Embodiments are based, at least in part, on the recognition that a cardiovascular model can be adapted to more closely mimic a subject's true cardiovascular system by modifying the model based solely on non-invasively measurable data, which effectively means that values ​​for certain cardiac parameters that previously could only be accurately measured invasively are predicted using a more accurate model than previously available models.

[0048] Exemplary embodiments may be utilized, for example, in a clinical setting where non-invasive measurements of a subject's cardiac function are obtained, such as in a clinic, hospital, or even in the subject's home environment.

[0049] 1 shows a system 100 for estimating left ventricular filling pressure in a subject, according to one embodiment of the present invention. System 100 includes a processing system 110 and a memory unit 120. The processing system is itself an embodiment of the present invention.

[0050] The processing system 110 retrieves a cardiovascular model 130 from the memory unit 120. A cardiovascular model (CVS model) is a biophysical model of (at least part of) the function of the heart and circulatory system. A plurality of model parameters are used to generate output data including a plurality of output value sets. The model parameters are defined input or intermediate variables or coefficients of the CVS model. Each output value set is associated with a different characteristic of cardiac function and includes one or more output values ​​representing (predicted) values ​​of that characteristic of cardiac function. Cardiovascular models are well known, and suitable CVS models for implementing the present invention will be apparent to those skilled in the art. An example of a suitable CVS model is described in more detail below.

[0051] More specifically, the CVS model 130 obtained by the processing system 110 is a model of the cardiovascular system that provides output data including a set of output values, each representing a different non-invasively measurable characteristic of cardiac function.

[0052] The plurality of model parameters of the CVS model includes at least a target cardiac parameter, and / or the output data further includes a set of further output values ​​of the target cardiac parameter, and thus the target cardiac parameter is represented in the input data, intermediate data, or output data of the CVS model.

[0053] Non-invasively measurable properties of cardiac function are properties that are indicative of cardiac function and whose values ​​are determined or derived (preferably directly) from non-invasively obtained measurements (e.g., echocardiography measurements, cardiac or cardiovascular MRI, peripheral pressure measurements, etc.). For example, non-invasively measurable properties include one or more of left ventricular volume, aortic pressure, cuff pressure, aortic valve flow, and / or mitral valve flow. Techniques for non-invasively measuring these properties are established in the art.

[0054] In some examples, the CVS model 130 is a low-dimensional model of the cardiovascular system (i.e., a zero-dimensional or one-dimensional CVS model). A low-dimensional CVS model efficiently models cardiac function using a relatively small number of model parameters (i.e., fewer than a high-dimensional model) and is more efficiently solved than a high-dimensional model. However, the present invention is not limited to the use of low-dimensional CVS models, and the use of higher-dimensional models of the cardiovascular system is also contemplated.

[0055] 2 provides a graphical representation of an exemplary model 230 of the cardiovascular system for use in the present invention. The CVS model 230 is a simple open-loop, zero-dimensional model of the left heart. An open-loop CVS model is a CVS model in which the circulatory loop is not closed.

[0056] In general, the model defines several relationships between various possible parameters of the cardiovascular system. These relationships can be expressed in terms of equations, some examples of which are described below.

[0057] In the exemplary model 230, the left atrial pressure P la is modeled as a constant throughout the cardiac cycle. P la (t)=P la (1)

[0058] Left ventricular blood inflow Q mv is the left atrial pressure P la and left ventricular pressure P lv and the mitral valve resistance R mvand is modeled as being driven by

number

[0059] The left ventricular pressure-volume relationship is modeled using an elastance function. P lv =E lv (V lv -V lv,0 ) (3) where E lv is the left ventricular elastance and V lv is the left ventricular volume, and V lv,0 is the left ventricular dead volume. At end diastole (ED), this gives: P lv (t=ED)=P la =LVEDP (4) Here, LVEDP is the left ventricular end-diastolic filling pressure, which is the left ventricular filling pressure at the end of diastole.

[0060] Left ventricular volume V lv Change in dV lv / dt is modeled as the difference between left ventricular inflow and outflow.

number

[0061] Amount of blood flowing from the left ventricle to the aorta Q av (i.e., aortic valve flow) is the left ventricular pressure P lv and aortic pressure P AO and the aortic valve resistance R p is modeled as being driven by

number

[0062] Aortic pressure P AO is modeled using a lumped three-element Windkessel model.

number

[0063] The CVS model 230 is solved numerically by solving a system of ordinary differential equations. In particular, the left atrial pressure P la Using multiple model parameters including the left ventricular pressure P lv , aortic pressure P AO , left ventricular volume V lv , and the blood inflow volume Q in the left ventricle mv and outflow Q av (i.e., the flow rates through the mitral and aortic valves, respectively) as a function of time.

[0064] This is more clearly or succinctly expressed by the introduction of a model parameter vector. μ=[P la ,p1,p2,…] (8) where p1, p2, ... are the left atrial pressures P la The left ventricular volume V lv , left ventricular pressure P lv , aortic pressure P AO , and mitral valve flow Q mv and aortic flow Q av The model solution is expressed as follows: V lv =V lv (t,μ),P lv =P lv (t,μ),P AO =P AO (t,μ),Q mv =Q mv (t,μ),Q av =Q av(t, μ) (9)

[0065] In other words, the CVS Model 230 measures the input left atrial pressure P la and other model parameters p1, p2, ... onto the output cardiovascular function curve (i.e., the time traces of pressure, volume, and flow). This can also be represented by collecting the model solutions into a solution vector F(t, μ). The model can be thought of as a mapping of the model parameter vector μ onto the solution vector F. μ→F(t,μ)=[V lv (t,μ),P lv (t,μ),P AO (t,μ),Q mv (t,μ),Q av (t, μ)] (10)

[0066] It will be appreciated that some elements in the solution vector F represent non-invasively measurable characteristics of cardiac function. Thus, at least some of the elements of the solution vector may represent sets of output values ​​related to non-invasively measurable characteristics of cardiac function. Other elements of the solution vector F represent additional sets of output values ​​(e.g., representing characteristics or parameters of cardiac function that historically could only be measured using invasive techniques).

[0067] Thus, the set of output data values ​​is effectively divided into an output data set (representing only non-invasively measurable characteristics of cardiac function) and a further output data set (which may represent invasively measurable characteristics of cardiac function).

[0068] One of the parameters represented by the additional output value set (if present) serves as the target cardiac parameter for purposes of the disclosed embodiments. For example, left ventricular pressure P lv serves as the target cardiac parameter. Other target cardiac parameters are envisioned for different types or instances of the CVS model.

[0069] In other examples, the target cardiac parameter is one of the model parameters. In the exemplary CVS model previously described, the left ventricular pressure P lv Also, as shown in equations (2) and (6), the blood inflow Q mv and blood outflow Q av is used in the calculation of , and therefore also serves as a model parameter.

[0070] Thus, in the exemplary CVS model, the plurality of model parameters includes a target cardiac parameter, and the output data includes a set of further output values ​​of the target cardiac parameter.

[0071] The invention will be further described with reference to this exemplary CVS model 230. However, the invention is not limited to the use of this model, and any suitable model of the cardiovascular system may be used, including closed-loop models, models with more realistic valve behavior, models in which the left atrial pressure is not constant (e.g., models that include an atrial elastance function), and models with an extended arterial system.

[0072] 1, processing system 110 acquires cardiac data 140 for a subject. The cardiac data includes a set of measurements for each set of output values. Thus, each measurement set is associated with a respective non-invasively measurable characteristic.

[0073] Cardiac data 140 is obtained from memory unit 120 and / or from a cardiac monitoring system (not shown).

[0074] In some examples, cardiac data 140 includes measurement sets for all characteristics or parameters for which output value sets are provided by CVS model 130. Of course, the CVS model may provide additional output value sets that do not have corresponding measurement sets in cardiac data 140. In other words, the multiple non-invasively measurable characteristics represented by the measurement sets of cardiac data comprise only a portion of the solution of the CVS model. In some examples, each measurement set includes multiple values ​​(of a particular non-invasively measurable characteristic of cardiac function) measured at different points in the cardiac cycle.

[0075] For example, the plurality of non-invasively measurable characteristics (represented by respective sets of output values) include left ventricular volume and at least one peripheral pressure parameter. For example, in the case of the exemplary model 230 described above, measurements of these characteristics have been found to be sufficient to determine an estimate of a particular target cardiac parameter, such as left ventricular filling pressure.

[0076] In other words, cardiac data 140 includes a set of measurements of left ventricular volume and a set of measurements of each of one or more peripheral pressure parameters, including diastolic pressure and / or systolic pressure.

[0077] Left ventricular volume is measured, for example, by 2D echocardiography (e.g., using the method of disks), 3D echocardiography (e.g., using the Philips Dynamic Heart Model), and / or cardiac MRI. Non-invasive methods for measuring left ventricular volume will be apparent to those skilled in the art.

[0078] Non-invasive methods for measuring peripheral pressure are also apparent to those skilled in the art. For example, peripheral pressure can be measured by performing systolic and diastolic pressure measurements on a cuff (e.g., to measure brachial, finger, neck, toe, ankle, and / or leg pressure). In another example, continuous arterial pressure measurements can be measured using an ultrasound device. See, e.g., Wang et al. (2018), "Monitoring of the central blood pressure waveform via a conformal ultrasonic device," Nat Biomed Eng, 2:687-695.

[0079] One or more transfer functions are used to convert this peripheral pressure to the aortic pressure P AO and the transformed measurements form part of the measurement set.

[0080] In some examples, the plurality of non-invasively measurable characteristics (having a representative output data set in the output data) further include aortic valve flow and / or mitral valve flow. In other words, cardiac data 140 further includes one or more measurement sets including at least one aortic valve flow measurement or one or more mitral valve flow measurements.

[0081] Non-invasive methods for measuring aortic and / or mitral flow will be apparent to those skilled in the art and include cardiovascular MRI flow measurements, Doppler measurements, and valve opening area measurements.

[0082] In some examples, the plurality of non-invasively measurable characteristics (having a representative output data set in the output data) further includes timing of cardiac cycle events. For example, cardiac data 140 further includes at least one measurement set including timing of mitral valve opening, timing of mitral valve closure, timing of aortic valve opening, timing of aortic valve closure, or timing of onset of left ventricular contraction.

[0083] Non-invasive methods for measuring the timing of cardiac cycle events will be apparent to those skilled in the art. For example, the timing or occurrence of cardiac cycle events can be determined using ECG data.

[0084] After acquiring the CVS model 130 and cardiac data 140, the processing system 110 iteratively modifies a plurality of model parameters of the model until one or more predetermined criteria are met, thereby correcting for differences between the set of output values ​​of the model and the set of measurements of the cardiac data. Suitable examples of the one or more predetermined criteria are described in more detail below.

[0085] The processing system then identifies the set of one or more values ​​of the target cardiac parameter (ie, the values ​​at the end of the iterative process) as the set of estimated one or more values ​​of the target cardiac parameter.

[0086] This involves identifying one or more sets of values ​​for the additional set of output values ​​representing the target cardiac parameter (e.g., compared to other set output values ​​representing non-invasively measurable characteristics of cardiac function). This approach is employed when the output of the CVS model 130 includes an additional set of output values ​​for the target cardiac parameter.

[0087] Alternatively, it includes one or more sets of values ​​defined for the model parameters of the CVS model that are used in the processing performed by the CVS model to produce the output data. This approach is taken when the model parameters of the CVS model 130 include target cardiac parameters, and thereby include a set of values ​​for the target cardiac parameters.

[0088] In other words, processing system 110 estimates a set of one or more values ​​for the subject's target cardiac parameters by fitting the model parameters to measurements in cardiac data 140. In other words, processing system 110 identifies a set of values ​​for the model parameters for which the solution or output of model 130 sufficiently corresponds to the subject's cardiac data, i.e., according to predetermined criteria.

[0089] In this way, the acquired cardiac data effectively serves as boundary conditions or values ​​for the CVS model 130 .

[0090] In some examples, the processing system 110 iteratively modifies multiple model parameters by defining a solution vector that includes multiple sets of output values ​​for the CVS model 130 and a measurement data vector that includes corresponding sets of measurement data. The solution vector is defined in the same manner as the solution vector for the example model 230 described above, i.e., in terms of a model parameter vector μ that includes the model parameters.

[0091] The processing system 110 then defines a cost function that quantifies the difference between (a portion of) the solution vector and the measurement data vector. For example, if the solution vector and the measurement data vector are each calculated at a discrete time t kIf the cost function CF is calculated or measured on a set of

number

[0092] It will be appreciated that for purposes of equation (11), the solution vector used includes only those properties of the heart that are equivalently represented in the measurement data vector. Thus, the properties of the heart defined or represented by the measurement data vector define the properties used in the cost function.

[0093] In some embodiments, the cost function is defined using only a subset of the measurement sets in the cardiac data and corresponding output value sets in the output data, and / or using only a selection of any of the corresponding output values ​​in the measurement sets and output value sets.

[0094] For example, a quantity of interest (e.g., a particular value or set of values) is identified in the set of measurements and in the corresponding set of output values ​​produced by the CVS model. This quantity of interest corresponds to a particular or predetermined characteristic (e.g., left ventricular end-diastolic volume and / or left ventricular end-systolic volume). A cost function is defined using only these quantities of interest.

[0095] Thus, a cost function is a function that uses only a subset of the quantity of interest, i.e., all possible measurement and model-determined parameter values, to define the difference between a solution vector (i.e., a set of output values) and a measurement data vector (i.e., a measurement data set).

[0096] The processing system 110 iteratively determines the value of the cost function C and modifies the model parameters based on the value of the cost function, with the goal of finding the model parameter vector μ for which the cost function is minimized.* (i.e., values ​​of the model parameters).

[0097] If the model parameters include a target cardiac parameter such as left ventricular pressure, this approach involves adding the model parameter vector μ * This facilitates identifying one or more values ​​of a target cardiac parameter from the

[0098] If the output data produced by the CVS model includes an additional set of output values ​​for a target cardiac parameter, such as left ventricular pressure, this approach facilitates identifying one or more values ​​for the target cardiac parameter from the output data (or solution vector) produced by the CVS model.

[0099] This concept is illustrated in FIG. 3, which shows an example cost function CF as a weighted sum of the areas between the curves (the dark areas in FIG. 3) for each of several non-invasively measurable characteristics.

[0100] In the example shown in FIG. 3, multiple non-invasively measurable characteristics include left ventricular volume, V LV , aortic pressure P AO , aortic valve flow Q av , and mitral valve flow Q mv The labels in Figure 3 identify the relevant areas shown by the accompanying image. The cost function is expressed as: CF=Δ rel V LV +Δ rel P AO +Δ rel Q av +Δ rel Q mv (12)

[0101] The darkened area in each graph represents the difference between the data curve of the non-invasively measurable characteristic (i.e., the curve based on measurements of cardiac data) and the solution curve (i.e., the curve based on the current output of the CVS model 130). The process of minimizing the example cost function C corresponds to minimizing the size of the darkened area.

[0102] For some non-invasively measurable characteristics, there may not be enough measurements (in the corresponding measurement set) to construct a data curve, e.g., there may be only one or two measurements of a particular non-invasively measurable characteristic. In such instances, one or more point-by-point differences between the measurements and the current output value of the CVS model 130 are used in the cost function instead of the area between one or more curves of the multiple non-invasively measurable characteristics.

[0103] For example, FIG. 4 shows the systolic and diastolic cuff pressures P cuff The point-by-point difference between the measured values ​​and the output values ​​of the CVS model 130 (respectively, ΔP sys and ΔP dia ), which is used in place of the area between the aortic pressure curves in the example cost function of FIG. 3 when aortic pressure data curves are not available. In this case, the cost function is expressed as: CF=Δ rel V LV +Δ rel P cuff +Δ rel Q av +Δ rel Q mv (13)

[0104] Several methods for iteratively determining the value of the cost function C and modifying multiple model parameters based on the value of the cost function (i.e., to minimize the cost function) are contemplated, and further suitable methods will be apparent to those skilled in the art. In particular, simultaneous optimization methods (e.g., gradient-based minimization methods such as the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm, or non-gradient-based methods such as particle swarm optimization and Bayesian optimization) and multilevel (e.g., bilevel) optimization techniques are used. Figure 5 illustrates a method 500 for iteratively determining the value of the cost function C and modifying multiple model parameters according to one embodiment of the present invention. Method 500 uses a simultaneous approach to minimize the cost function, in which the cost function is simultaneously modified as a function of all elements of the model parameter vector μ (i.e., all of the model parameters).

[0105] Method 500 begins at step 510, where an initial value set is established for each of a plurality of model parameters of CVS model 130. The initial value set may be randomly assigned values, predefined values, or values ​​obtained by user input. For example, the initial values ​​may be defined as the mean of a possible range. For some model parameters, the initial values ​​may be estimated from cardiac data 140.

[0106] In step 520, a set of output values ​​for the CVS model 130 is determined based on the current set of values ​​for each of the plurality of model parameters. In other words, the CVS model is solved using the current set of values ​​for each of the plurality of model parameters. In the first iteration of the method, the current set of values ​​is the initial set of values ​​in step 510. In subsequent iterations, the current set of values ​​is a modified set of values, as described below.

[0107] In step 530, a value of the cost function is calculated based on the determined set of output values ​​and a corresponding set of measurements of non-invasively measurable characteristics in the cardiac data 140. In a particular example, the difference between the CVS model solution (produced using the current values ​​of the model parameters) and the subject's data is determined.

[0108] In step 540, the processing system 110 determines whether one or more predetermined criteria are met. In response to determining that the one or more predetermined criteria are not met, the method 500 proceeds to step 550.

[0109] In step 550, the set of values ​​for each of the plurality of model parameters is modified based on the calculated value of the cost function, using a gradient descent technique or any other suitable modification technique for minimizing or reducing the cost function, as would be readily apparent to one skilled in the art.

[0110] Steps 520-550 are repeated iteratively until it is determined that one or more predetermined criteria have been met. In response to determining that one or more predetermined criteria have been met, method 500 proceeds to step 560.

[0111] In step 560, the processing system 110 estimates a set of one or more values ​​for the target cardiac parameters of the subject.

[0112] Step 560 is performed by identifying, if the plurality of model parameters includes a target cardiac parameter, a revised set of values ​​for the target cardiac parameter (i.e., the current set of values ​​for the target cardiac parameter in the most recent iteration) as a set of one or more values ​​for the subject's estimated target cardiac parameter.

[0113] Alternatively, step 560 is performed by, if the output data of the CVS model includes a further set of output values ​​for the target cardiac parameters, identifying this further set of output values ​​as a set of one or more values ​​for the subject's estimated target cardiac parameters.

[0114] 6 illustrates an alternative method 600 for iteratively determining values ​​of a cost function C and modifying multiple model parameters, according to another embodiment of the present invention. Method 600 uses a bi-level optimization technique to minimize the cost function, with different groups of parameters being modified independently of each other.

[0115] Method 600 begins at step 610 where a plurality of model parameters is divided into a first set of model parameters and a second set of model parameters.

[0116] In some examples, a first set of model parameters is defined such that the first set includes only the target cardiac parameter, and a second set of model parameters is defined such that the second set includes the remaining model parameters (i.e., all of the model parameters except the target cardiac parameter). Naturally, in this scenario, the model parameters include at least the target cardiac parameter. This approach allows the target cardiac parameter to be modified independently of the other model parameters. In other words, the target cardiac parameter is interpreted as a different control input from the remaining model parameters.

[0117] In another example, the processing system 110 divides the plurality of model parameters into a first set and a second set by performing a sensitivity analysis on the CVS model 130 and dividing the plurality of model parameters into a first set and a second set based on the sensitivity analysis.

[0118] An appropriate sensitivity analysis, such as the Morris or Sobol method, is used to divide the model parameters into a first and a second set. In some cases, more than one sensitivity analysis is performed. For example, the Morris method is used for a first screening and the Sobol method is used for a more detailed analysis.

[0119] The results of the sensitivity analysis are used, for example, to distinguish between model parameters that have a stronger influence on the cost function C and model parameters that have a weaker influence on the cost function C. A first set of model parameters is defined to include the model parameters that have a stronger influence, and a second set of model parameters is defined to include the model parameters that have a weaker influence.

[0120] For example, the first set of model parameters may be defined to include model parameters having a Sobol index above a predetermined threshold, or may be defined to include N model parameters having the highest Sobol indexes, where N is a predetermined number. The second set of model parameters may be defined to include model parameters having a Sobol index not exceeding a predetermined threshold, or may be defined to include all model parameters except the N model parameters having the highest Sobol indexes.

[0121] Figure 7 shows the aortic valve resistance r p , arterial resistance r d , arterial impedance c, mitral valve resistance R mv , left atrial pressure P la , elastance e min , elastance e max , m1, m2, τ1, τ2, and left ventricular dead volume V lv,0 7 shows example results 700 of a Sobol sensitivity analysis performed on a CVS model having as model parameters

[0122] In the example result 700, the left atrial pressure P lais the dominant model parameter, has the strongest influence on the cost function (S1) and has the greatest interaction with other model parameters (ST), and is influenced by the elastance e min and aortic valve resistance r p The remaining model parameters have very small Sobol indices.

[0123] The model parameters are based on the example result 700, e.g., left atrial pressure P la to the first set and the remaining model parameters to the second set, or the left atrial pressure P la , elastance e min , and aortic valve resistance r p The model parameters are divided into first and second sets by assigning the left atrial pressure to the first set and the remaining model parameters to the second set. In another example, the model parameters are divided into three sets: the first set includes only left atrial pressure, the second set includes elastance and aortic valve resistance, and the third set includes the remaining model parameters. One skilled in the art can easily adapt method 600 to a method having three sets of model parameters.

[0124] 6, an initial value is set or defined for each of a plurality of model parameters in step 620. The initial value may be a randomly assigned value, a predefined value, or a value obtained by user input.

[0125] In some examples, at least some of the initial values ​​are estimated from available measurements (e.g., from a measurement data set) in step 620. This approach utilizes existing statistical models (which may not be accurate enough for a particular subject, but can be a good starting point for fitting model parameters to a particular subject) to define or predict parameter values ​​from non-invasively measurable data.

[0126] In step 630, the processing system 110 iteratively modifies a set of values ​​for each of the first set of multiple model parameters until a first predefined convergence condition is met. Step 630 is described in more detail with reference to FIG. 8, which shows the substeps of step 630.

[0127] In sub-step 631, a set of output values ​​for the CVS model 130 is determined based on the current values ​​of each of the model parameters.

[0128] In sub-step 632, a cost function value is calculated based on the determined output values.

[0129] In sub-step 633, the processing system determines whether a first predefined convergence condition is met, for example, when the value of the cost function falls below a first predetermined threshold. Other examples are provided later in this disclosure.

[0130] In response to determining that the first predefined convergence condition is not met, step 630 proceeds to sub-step 634 .

[0131] In sub-step 634, the set of values ​​for each of the first set of model parameters is adjusted based on the calculated value of the cost function. The set of values ​​for the second set of model parameters is not adjusted during sub-step 634.

[0132] Sub-steps 631-634 are repeated iteratively until it is determined that a first predefined convergence condition is met. In response to determining that the first predefined convergence condition is met, method 600 proceeds to step 640.

[0133] Returning to FIG. 6, in step 640, the processing system 110 iteratively modifies a set of values ​​for each of the second set of multiple model parameters until a second predefined convergence condition is met.

[0134] Step 640 is described in more detail with reference to FIG. 9, which shows the substeps of step 640.

[0135] In sub-step 641, a set of output values ​​for the CVS model 130 is determined based on a current set of values ​​for each of a plurality of model parameters.

[0136] In sub-step 642, a cost function value is calculated based on the determined sets of output values ​​(and measurement sets).

[0137] In sub-step 643, the processing system determines whether a second predefined convergence condition is met, for example, when the value of the cost function falls below a second predetermined threshold. Other examples are provided later in this disclosure.

[0138] In response to determining that the second predefined convergence condition is not met, step 640 proceeds to sub-step 644 .

[0139] In sub-step 644, the set of values ​​for each of the second set of model parameters is adjusted based on the calculated value of the cost function. The set of values ​​for the first set of model parameters is not adjusted during sub-step 644.

[0140] Sub-steps 641-644 are repeated iteratively until it is determined that a second predefined convergence condition is met. In response to determining that the second predefined convergence condition is met, method 600 proceeds to step 650.

[0141] Steps 630 and 640 are better understood by considering step 610, which divides the model parameters into first and second sets, in terms of dividing the model parameter vector μ into two components. For example, if multiple model parameters are divided according to the results of a sensitivity analysis, the model parameter vector μ is represented as being divided into two components. μ=[μ strong ,μ weak ] (14) where μ strong is a vector containing the first set of model parameters (i.e., the model parameters with a stronger influence), and μ weak is a vector containing the second set of model parameters (i.e., the model parameters with a weaker influence).

[0142] In step 630, the objective is to

number

number

[0143] In step 640, the objective is to

number

number

[0144] In step 650, the processing system 110 determines whether one or more predetermined criteria are met.

[0145] In response to determining that the one or more predetermined criteria are not met, steps 630-650 are repeated iteratively until the one or more predetermined criteria are determined to be met. In some examples, the predetermined criteria are met when both the first and second convergence conditions are met and maintained.

[0146] In response to determining that the one or more determined criteria are met, the method 600 proceeds to step 660 .

[0147] In step 660, the processing system 110 estimates the target cardiac parameters of the subject by identifying the modified set of values ​​of the model parameters of the target cardiac parameters as the estimated target cardiac parameters of the subject.

[0148] The embodiment described with reference to Figures 6 to 9 effectively divides the model parameters into two sets and performs an iterative refinement process on each set of model parameters. This approach can be extended to include dividing the model parameters into N sets (N>1), with an iterative refinement run being performed on each set of parameters.

[0149] This generalized approach is illustrated by FIG. 10, which shows a method 1000 for iteratively determining values ​​of the cost function CF and modifying multiple model parameters.

[0150] The method 1000 includes a step 1010 of dividing the model parameters into a number of N sets, where N is any positive integer greater than 1. The method 1000 further includes a step 1020 of setting one or more initial values ​​for the model parameters.

[0151] Each set of model parameters is iteratively modified in turn until some predetermined (termination) criteria is met for each instance of iterative modification, as illustrated by steps 1030-1050 in Figure 10.

[0152] In step 1030, the Zth set of parameters is modified (where Z is initially 1).

[0153] In step 1040, the method determines whether some predetermined criteria have been met, e.g., whether the cost function of the Zth set of parameters meets some predetermined criteria. The cost function is different for each set of parameters. Of course, the same cost function may be shared between two or more sets of parameters.

[0154] If the result of step 1040 is negative, the method returns to step 1030. If the result of step 1040 is positive, the method determines (at step 1050) whether all sets have been processed. If all sets have been processed, the method moves to step 1060, where the value of the target cardiac parameter is estimated. Otherwise, step 1030 is performed for the next set of parameters (e.g., by adding 1 to the value of Z in step 1055 and then returning to step 1030).

[0155] In some examples, rather than moving to step 1060 when all sets have been processed, the method may return to step 1030 for the first set of model parameters processed. Thus, Z is effectively reset to 1, since modifications to subsequent sets may have affected the accuracy of the model. This iteration of modifying sets of model parameters is performed until some predetermined criteria are met, examples of which are described below.

[0156] Suitable techniques for modifying the parameters have been previously described.

[0157] The division of the model parameters into N sets is performed, for example, based on a sensitivity analysis of the model parameters.

[0158] One particularly advantageous approach for setting the model parameters of the exemplary model described with reference to FIG. 2 is described below.

[0159] The proposed approach combines assumptions based on an understanding of physiology and cardiac function, the results of a sensitivity analysis of an exemplary model of the CV system, and the relationships and interactions between model parameters, thereby generating model parameters that more accurately or more accurately represent the true processes occurring within a subject.

[0160] In this approach, the first set of model parameters is the unknown parameter Z c , R d , and C, which are the lumped impedance, resistance, and compliance of the arterial system, respectively (see equation (7)). These three parameters form part of the Windkessel model and are therefore called Windkessel (WK) parameters.

[0161] In this approach, the second set of model parameters is the filling pressure P la and left ventricular dead volume V lv,0 This second set has been identified through sensitivity analysis as being particularly effective in modifying the solution vectors produced by the example model.

[0162] In this approach, the third set of model parameters includes all remaining parameters of the example model, effectively fine-tuning the example model based on the coarser refinements performed using the first and second sets.

[0163] As previously explained, the cost function used in iteratively modifying each set of model parameters may differ depending on the set of model parameters being processed.

[0164] As an example, for the first set of model parameters, the cost function may be a function of the predicted aortic pressure (e.g., in the solution vector or output data) and / or aortic flow Q using Eq. (7). av (by processing) and the true aortic pressure P AO (in cardiac data or measurement data vectors)av is available in the cardiac data or is derived from other parameters available in the measurement data vector.

[0165] For the second and third sets of model parameters, the cost function is the quantified difference (e.g., total error, average error, or mean squared error) between all model parameters shared by the solution vector and the measurement data vector. Other suitable examples will be apparent.

[0166] To process the first set of model parameters, aortic flow Q av If is not available in the measurement data vector, for example, when operating under the assumption that there is no regurgitation through the mitral valve during systole, Q av (t)=dV lv As (t) / dt, the aortic flow rate Q av can be estimated as the time change in LV volume.

[0167] In some examples, the measurement data vector may be a vector of upper arm cuff measurements P cuff One or more transfer functions are used to convert this measurement to aortic pressure P AO Suitable transfer functions will be readily apparent to those skilled in the art.

[0168] The various iterative refinement processes performed in such an iterative process may be performed using any known refinement process, such as common gradient-based minimization methods, non-gradient-based and global optimization methods (particle swarms, genetic algorithms, Bayesian optimization, etc.).

[0169] In one example, the iterative modification process includes generating an output value set for each of a plurality of different value sets for the set of model parameters being modified, determining a cost function for each output value set, and selecting the value of the set of model parameters associated with the lowest cost function. The plurality of different values ​​may, for example, include a plurality of value sets, each value set representing a value for each of a plurality of parameters to be modified. In some examples, there are maximum and minimum values ​​for each parameter defined based on a known range for the parameter. Each value set represents a sample within these maximum and minimum boundaries. The plurality of value sets need not include all possible values ​​for each parameter; rather, a sample selection of possible values ​​for each parameter is used.

[0170] In any of the above-described embodiments, it is possible to define one or more boundaries or constraints for one or more model parameters by processing determined values ​​of other model parameters. In particular, the value of a model parameter that is part of a subsequent set of model parameters compared to a previously processed model parameter is constrained (with upper and / or lower bounds) based on the determined value of the model parameter determined in the previous processing. In this context, a subsequent set is a set that the model parameter comprises.

[0171] A particular set of model parameters determines the filling pressure P la and left ventricular dead volume V lv,0 and the later set of model parameters (i.e., the set of model parameters that are modified after a particular set) is the left ventricular elastance E lv In this scenario, the left ventricular elastance, E, can be calculated using the following equation: lv Constraint E min It is possible to define

number

[0172] Calculate the left ventricular elastance E using the following equation: lv (used in equation (3)) can be defined.

number

number

[0173] This example shows how it is possible to set constraints on a parameter based on one or more previously determined values ​​of one or more other model parameters.

[0174] As previously mentioned, the processing system iteratively modifies the model parameters until one or more predefined criteria or convergence conditions are met, thereby correcting the difference between the output values ​​and the measurements. Termination criteria for optimization processes are well known, and appropriate criteria for use as the one or more predefined criteria will be apparent to one skilled in the art.

[0175] For example, the one or more predetermined criteria or convergence conditions may include at least one of: a determination that the difference in the value of each of the plurality of model parameters between the current iteration and the immediately preceding iteration is less than a predetermined threshold; a determination that the difference in the value of the cost function between the current iteration and the immediately preceding iteration is less than a predetermined threshold; a determination that the number of iterations has exceeded a predetermined threshold; and / or a determination that the difference between each of the plurality of output values ​​of the CVS model 130 and the corresponding measurement in the cardiac data 140 is less than the uncertainty of the measurement. Suitable values ​​for the predetermined threshold will be apparent to those skilled in the art. In some examples, the one or more predetermined criteria may include at least one criterion related to the history of the model parameter values ​​to terminate the process when the solution is oscillating between two values.

[0176] If the one or more predetermined criteria or convergence conditions include several predetermined criteria, the processing system iteratively modifies the model parameters until all of the predetermined criteria are met or until at least one of the predetermined criteria is met.

[0177] In some examples, the output data of the CVS model 130 is used to generate a cardiac function curve for the subject. The processing system 110 determines multiple output value sets (and, optionally, further output value sets, if any) for the CVS model based on the revised value sets for each of the model parameters (i.e., the value sets at the end of the iterative process after one or more predetermined criteria have been met), and generates one or more cardiac function curves based on the determined output value sets and the further output value sets (if any).

[0178] The one or more cardiac function curves include one or more of a time trace of pressure (e.g., left ventricular pressure and / or aortic pressure) over one or more cardiac cycles, a time trace of volume (e.g., left ventricular volume) over one or more cardiac cycles, a time trace of flow (e.g., flow through the mitral valve and / or aortic valve) over one or more cardiac cycles, and / or a pressure-volume loop.

[0179] FIG. 11 illustrates a computer-implemented method 1100 for estimating a target cardiac parameter of a subject, according to one embodiment of the present invention.

[0180] Method 1100 begins at step 1110, where a model of the cardiovascular system is obtained. The model of the cardiovascular system uses a plurality of model parameters to generate output data including a plurality of sets of output values, each set of output values ​​associated with a different non-invasively measurable characteristic of cardiac function. Each set of output values ​​includes at least one value representing a value of the associated non-invasively measurable characteristic of cardiac function. The plurality of model parameters includes a target cardiac parameter, and / or the output data includes a further set of output values ​​for the target cardiac parameter.

[0181] Cardiac data is acquired for the subject at step 1120. The cardiac data includes a plurality of measurement sets, each measurement set associated with a different set of output values ​​and including one or more measurements of non-invasively measurable characteristics associated with the associated set of output values.

[0182] In step 1130, the method iteratively modifies a set of values ​​for a plurality of model parameters of the CVS until one or more predetermined criteria are met, thereby correcting for differences between a set of output values ​​for the output data and a corresponding set of measurements in the cardiac data.

[0183] Step 1140 is performed after the iterative modification to establish or define a set of one or more values ​​of the target cardiac parameter. Step 1140 includes, if the plurality of model parameters includes the target cardiac parameter, defining the set of one or more values ​​of the modified target cardiac parameter as a set of one or more estimated values ​​of the target cardiac parameter. Alternatively, if the output data includes a further set of output values, step 1140 includes defining the further set of output values ​​as a set of one or more estimated values ​​of the target cardiac parameter.

[0184] Method 1100 may further include step 1150 of providing a user-perceptible output in an output interface in response to the set of one or more estimated values ​​of the target cardiac parameter. The user-perceptible output may be a visual representation of the estimated set (e.g., in the form of a curve or display of any values).

[0185] It will be understood that the disclosed methods are computer-implemented methods, and therefore the concept of a computer program is also proposed, comprising code means for performing any of the described methods when said computer program is run on a processing system.

[0186] Those skilled in the art will be able to readily develop a processor to perform the methods described herein, and each step of the flowchart therefore represents a different action performed by a processor, and is performed by a respective module of the processor.

[0187] As discussed above, systems utilize processors to perform data processing. Processors may be implemented in numerous ways using software and / or hardware to perform the various functions required. Processors typically utilize one or more microprocessors that are programmed using software (e.g., microcode) to perform the required functions. Processors may be implemented as a combination of dedicated hardware to perform some functions and one or more programmed microprocessors and associated circuitry to perform other functions.

[0188] Examples of circuitry utilized in various embodiments of the present disclosure include, but are not limited to, conventional microprocessors, application specific integrated circuits (ASICs), and field programmable gate arrays (FPGAs).

[0189] In various embodiments, the processor is associated with one or more storage media, such as volatile and non-volatile computer memory, including RAM, PROM, EPROM, and EEPROM. The storage media are coded with one or more programs that, when executed by the one or more processors and / or controllers, perform the necessary functions. The various storage media may be fixed within the processor or controller, or may be portable such that one or more programs stored thereon can be loaded into the processor.

[0190] Modifications to the disclosed embodiments can be understood and effected by those skilled in the art, from a study of the drawings, the disclosure, and the appended claims, in practicing the claimed invention. In the claims, the words "comprises," "includes," and "having" do not exclude other elements or steps, and the words "a" or "an" in the singular do not exclude a plurality. A single processor or other unit fulfills the functions of several items recited in the claims. The mere fact that certain means are recited in mutually different dependent claims does not indicate that a combination of these means cannot be used to advantage. A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium, provided together with or as part of other hardware, or may further be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems. It should be noted that when the term "adapted to" is used in the claims or the description, the term "adapted to" is intended to be equivalent to the term "configured for." Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. 1. A processing system for estimating a set of one or more values ​​of target cardiac parameters of a subject, the processing system comprising: obtaining a model of the cardiovascular system using the plurality of model parameters to produce output data including a plurality of sets of output values ​​each associated with a different non-invasively measurable characteristic of cardiac function; the non-invasively measurable characteristic is a characteristic whose value is derived from measurements obtained non-invasively; each set of output values ​​includes at least one value of an associated non-invasively measurable characteristic of said cardiac function; the plurality of model parameters includes the target cardiac parameter, and / or the output data includes a further set of output values ​​for the target cardiac parameter; The processing system includes: acquiring cardiac data including a plurality of measurement sets, each measurement set associated with a different output value set and including one or more measurements of the non-invasively measurable characteristic associated with the associated output value set; iteratively modifying values ​​of the plurality of model parameters of the model, thereby correcting for differences between the plurality of sets of output values ​​of the output data and the corresponding sets of measurements of the cardiac data; Iteratively modifying values ​​of the plurality of model parameters of the model comprises: dividing the plurality of model parameters into a plurality of sets of one or more model parameters; establishing an initial value set for each of the plurality of model parameters; For each set of one or more model parameters, in turn: determining one or more sets of output values ​​of the model of the cardiovascular system based on a current set of values ​​of at least the set of model parameters; calculating a value of a cost function based on the determined set of one or more output values ​​and the set of measurements associated with the set of one or more output values; and iteratively adjusting a set of values ​​for each model parameter of the set of model parameters based on the calculated value of the cost function until one or more predetermined criteria are met; the one or more predetermined criteria are criteria for minimizing the cost function; The processing system includes: after iteratively modifying values ​​of the plurality of model parameters of the model; if the plurality of model parameters includes the target cardiac parameter, defining a set of one or more values ​​of the modified target cardiac parameter as the set of one or more estimated values ​​for the target cardiac parameter; or If the output data includes a further set of output values ​​for the target cardiac parameter, the processing system is configured to define the further set of output values ​​as the set of one or more estimated values ​​for the target cardiac parameter.

2. The processing system of claim 1 , wherein the target cardiac parameter is left ventricular pressure.

3. The processing system of claim 1 , wherein the plurality of non-invasively measurable characteristics comprises left ventricular volume and at least one peripheral pressure characteristic.

4. The processing system of claim 1 , wherein the plurality of non-invasively measurable characteristics comprises at least one of aortic valve flow, mitral valve flow, and / or timing of cardiac cycle events.

5. the processing system comprising: dividing the plurality of model parameters into a plurality of sets of one or more model parameters by dividing the plurality of model parameters into a first set of model parameters and a second set of model parameters; a set of values ​​for each of the first set of model parameters; determining the plurality of sets of output values ​​of the model of the cardiovascular system based on a current set of values ​​for each of the plurality of model parameters; calculating a value of the cost function based on the determined sets of output values; adjusting the set of values ​​for each of the first set of model parameters based on the calculated value of the cost function; iteratively repeating the steps of determining the plurality of sets of output values, calculating the values ​​of the cost function, and adjusting the set of values ​​for each of the first model parameters until a first predetermined criterion is met; and modifying it by a set of values ​​for each of the second set of model parameters; determining the plurality of sets of output values ​​of the model of the cardiovascular system based on a current set of values ​​for each of the plurality of model parameters; calculating a value of the cost function based on the determined sets of output values; adjusting the set of values ​​for each of the second set of model parameters based on the calculated value of the cost function; iteratively repeating the steps of determining the plurality of output value sets, calculating the value of the cost function, and adjusting the value set for each second model parameter until a second predetermined criterion is met; and modifying it by iteratively repeating the steps of modifying the set of values ​​for each of the first set of model parameters and modifying the set of values ​​for each of the second set of model parameters until the one or more predetermined criteria are met; The processing system of claim 1 ,

6. 6. The processing system of claim 5, wherein the plurality of model parameters includes the target cardiac parameter, and the processing system divides the plurality of model parameters into the first set of model parameters and the second set of model parameters by defining the first set of model parameters to include only the target cardiac parameter and defining the second set of model parameters to include the remaining model parameters.

7. the processing system dividing the plurality of model parameters into a plurality of sets of one or more model parameters; performing a sensitivity analysis on the model of the cardiovascular system; The processing system of claim 1 , further comprising: dividing the plurality of model parameters into a plurality of sets of the one or more model parameters based on the sensitivity analysis.

8. 2. The processing system of claim 1, wherein the one or more predetermined criteria comprise at least one of: a determination that a difference in a set of values ​​of each of the plurality of model parameters between a current iteration and a previous iteration is less than a predetermined threshold; a determination that a difference in the value of the cost function between a current iteration and a previous iteration is less than a predetermined threshold; a determination that a number of iterations has exceeded a predetermined threshold; a determination that an amount of time spent performing the iterative corrections exceeds a predetermined amount of time; and / or a determination that a difference between each of the plurality of sets of output values ​​of the model of the cardiovascular system and the corresponding measurement set is less than an uncertainty of the corresponding measurement set.

9. The processing system further comprises, after the iterative modification: determining the plurality of output value sets of the model of the cardiovascular system based on the modified value sets of each of the plurality of model parameters; The processing system according to claim 1 , wherein the processing system generates one or more cardiac function curves based on the determined plurality of sets of output values.

10. 1. A computer-implemented method for estimating a set of one or more values ​​of target cardiac parameters of a subject, the computer-implemented method comprising: obtaining a model of the cardiovascular system using a plurality of model parameters to produce output data including a plurality of sets of output values ​​each associated with a different non-invasively measurable characteristic of cardiac function; the non-invasively measurable characteristic is a characteristic whose value is derived from measurements obtained non-invasively; each said set of output values ​​includes at least one value of an associated non-invasively measurable characteristic of cardiac function; the plurality of model parameters includes the target cardiac parameter, and / or the output data includes a further set of output values ​​for the target cardiac parameter; The computer-implemented method comprises: acquiring cardiac data comprising a plurality of measurement sets, each measurement set associated with a different output value set and comprising one or more measurements of the non-invasively measurable characteristic associated with the associated output value set; iteratively modifying values ​​of the plurality of model parameters of the model, thereby correcting for differences between the plurality of sets of output values ​​of the output data and the corresponding sets of measurements of the cardiac data; Iteratively modifying values ​​of the plurality of model parameters of the model comprises: dividing the plurality of model parameters into a plurality of sets of one or more model parameters; establishing an initial value set for each of the plurality of model parameters; For each set of one or more model parameters, in turn: determining one or more sets of output values ​​of the model of the cardiovascular system based on a current set of values ​​of at least the set of model parameters; calculating a value of a cost function based on the determined set of one or more output values ​​and the set of measurements associated with the set of one or more output values; and iteratively adjusting a set of values ​​for each model parameter of the set of model parameters based on the calculated value of the cost function until one or more predetermined criteria are met; the one or more predetermined criteria are criteria for minimizing the cost function; The computer-implemented method comprises: after iteratively modifying values ​​of the plurality of model parameters of the model; if the plurality of model parameters includes the target cardiac parameter, defining a set of one or more values ​​of the modified target cardiac parameter as the set of one or more estimated values ​​for the target cardiac parameter; or If the output data includes a further set of output values ​​for the target cardiac parameter, defining the further set of output values ​​as the set of one or more estimated values ​​for the target cardiac parameter.

11. A computer program comprising code means which, when executed on a computer having a processing system, causes said processing system to perform all of the steps of the method of claim 10.

Images