Systems and methods for non-invasive assessment of cardiac pressure and efficiency

Abstract

The disclosure provides systems, methods, and non-transitory computer readable media for non-invasively assessing cardiac pressure and efficiency. An example computer implemented method includes: receiving patient-specific image data of a cardiac chamber of a patient, the patient-specific image data acquired from an imaging modality over a cardiac cycle; receiving patient-specific vitals data; deriving non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; applying the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generating, in real time, based on a ratio of the total energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient-specific cardiac efficiency value classifying a severity of a heart disease of the patient.

Description

UCLA 2024-258-2 UCLA.P0211WO SYSTEMS AND METHODS FOR NON-INVASIVE ASSESSMENT OF CARDIAC PRESSURE AND EFFICIENCY CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to U.S. provisional patent application serial number 63 / 659,513 filed June 13, 2024, the entire content of which is incorporated herein by reference and relied upon. STATEMENT REGARDING FEDERALLY SPONSORED R&D

[0002] This invention was made with government support under I01CX001901 awarded by the U.S. Department of Veterans Affairs, and HL127153 awarded by the National Institutes of Health. The government has certain rights in the invention. FIELD OF USE

[0003] The embodiments disclosed herein are generally directed towards systems, processes and methods that assess cardiac function. BACKGROUND

[0004] Patients with heart disease such as complex congenital heart disease (CHD) typically have abnormal cardiovascular anatomy, hemodynamics, and / or cardiac energetics. In particular, patients with single ventricle physiology (SVP) are dependent on one active chamber for perfusion of multiple organs including the heart, lungs, liver, kidneys, and extremities. Failure of this single ventricle can have catastrophic consequences. Markers of deterioration in function play a crucial role in long term monitoring.

[0005] Cardiac efficiency has been proposed as a means for diagnosing adverse effects of heart diseases like CHD. Typically, cardiac efficiency is measured by obtaining pressure-volume loops in the heart through invasive means by estimating pressure-volume area of the cardiac chamber (e.g., one or more ventricles of the heart). This invasive means involves an interventional maneuver where the inferior vena cava is occluded during invasive catheterization to change the ventricular loading conditions. Although cardiac efficiency obtained through estimating pressure-volume area invasively provides an estimate of the myocardial potential energy following ejection, there is ambiguity and subjectivity in estimation of myocardial potential energy following ejection, and it is impractical forUCLA 2024-258-2 UCLA.P0211WO routine clinical use. Moreover, such conventional methods of obtaining cardiac efficiency may be insensitive to atrioventricular and semilunar valve regurgitation, which is common in complex CHD. Additionally, the invasive nature of obtaining pressure-volume area for cardiac efficiencies not without risk, limiting access to optimal diagnosis and treatment of deteriorating function in CHD and other diseases.

[0006] Although non-invasive means for estimating cardiac efficiency exists, such conventional techniques are limited in patients with complex CHD. For example, ejection fraction is an example of such non-invasive technique, but ejection fraction is often insensitive to early changes associated with CHD. Furthermore, reference values for myocardial strain are not defined in the population of patients with CHD. Additional non-invasive cardiac parameters for ventricular function are therefore desirable.

[0007] There is thus a desire and need for non-invasive systems and methods for assessing cardiac pressure and efficiency. Furthermore, there is a desire and need for such non-invasive systems and methods that are sensitive to atrioventricular and semilunar valve regurgitation. Even further, there is a desire and need for such non-invasive systems and methods that are applicable for patients with complex cardiac anatomy, especially those with single ventricle palliation (SVP). Various embodiments of the present disclosure address one or more of the above-described shortcomings. SUMMARY

[0008] This specification describes various exemplary embodiments of systems, software and methods for non-invasively assessing cardiac efficiency. The disclosure, however, is not limited to these exemplary embodiments and applications or to the manner in which the exemplary embodiments and applications operate or are described herein.

[0009] In at least one aspect, a method for non-invasively assessing cardiac efficiency includes: receiving, by a processor, patient-specific image data of a cardiac chamber of a patient, the patient- specific image data acquired from an imaging modality over a cardiac cycle; receiving, by the processor, patient-specific vitals data; deriving, by the processor, non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; applying, by the processor, the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generating, in real time, based on a ratio of the totalUCLA 2024-258-2 UCLA.P0211WO energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient-specific cardiac efficiency value classifying a severity of a heart disease of the patient.

[0010] In some aspects, which may be combined with other aspects, the patient-specific vitals data comprises a heart rate, and wherein the patient-specific cardiac efficiency value is further based on the heart rate.

[0011] In some aspects, which may be combined with other aspects, the method further includes: determining, by the processor, based on the patient-specific image data, an end diastolic volume (EDV) and an end systolic volume (ESV) of the cardiac chamber, and a blood flow rate at a ventricular outlet of the cardiac chamber; wherein the non-invasive cardiac parameters applied to the trained model comprises the EDV, the ESV, and the blood flow rate.

[0012] In some aspects, which may be combined with other aspects, the patient-specific vitals data comprises non-invasive blood pressure (NIBP) measurements and a heart rate of the patient. The non- invasive cardiac parameters applied to the trained model further includes the NIBP measurements, wherein the patient-specific cardiac efficiency value is further based on the heart rate.

[0013] In some aspects, which may be combined with other aspects, applying the non-invasive cardiac parameters to the trained model includes: determining, based on trained model, a ventricular pressure at each of a plurality of events of the cardiac cycle; and determining, based on the trained model and the ventricular pressures, the total mechanical energy input and the total mechanical energy output.

[0014] In some aspects, which may be combined with other aspects, the trained model includes one or more linear regression models, wherein the one or more linear regression models are trained using a plurality of blood pressure waveforms of a plurality of respective reference patients.

[0015] In some aspects, which may be combined with other aspects, the method further includes, prior to applying the one or more parameters to the one or more linear regression models: receiving the plurality of blood pressure waveforms of the plurality of respective reference patients. For each blood pressure waveform of a respective reference patient, the method includes: identifying a ventricular pressure at each of a plurality of events of a cardiac cycle associated with the respective reference patient; determining, based on the ventricular pressures, a total energy output for blood flow out of a cardiac chamber of the respective reference patient during the cardiac cycle associated withUCLA 2024-258-2 UCLA.P0211WO the respective reference patient and a total energy input for blood flow into the cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient; and associating the total energy output of the respective reference patient and the total energy input with at least one feature vector representing non-invasive cardiac parameters of the respective reference patient. The method further includes training, by the processor, based on the at least one associated feature vector of each respective reference patient, the one or more linear regression models to predict the total mechanical energy output for blood flow out of the cardiac chamber of the patient and the total mechanical energy input for blood flow into the cardiac chamber of the patient using the non-invasive cardiac parameters of the patient.

[0016] In some aspects, which may be combined with other aspects, determining the total energy output and the total energy input of the respective reference patient includes determining, based on the ventricular pressures, one or more components of a total mechanical energy input into the cardiac chamber of the respective reference patient. Furthermore, associating the total energy output and the total energy input with the at least one feature vector representing the non-invasive cardiac parameters of the respective reference patient includes associating each component of the total mechanical energy input with a respective feature vector based on non-invasive cardiac parameters of the respective reference patient.

[0017] In some aspects, which may be combined with other aspects, each component of the total mechanical energy input includes a mechanical energy distribution for a respective phase of the cardiac cycle, wherein the respective phase is among one or more phases of the cardiac cycle corresponding to the one or more components of the total mechanical energy input.

[0018] In some aspects, which may be combined with other aspects, the one or more phases corresponding to the one or more components of the total mechanical energy input includes: an isovolumic contraction; a systolic ejection; an isovolumic relaxation; and a diastolic filling.

[0019] In some aspects, which may be combined with other aspects, the method further includes: determining, by the processor, an average ventricular pressure for each of a systole phase and a diastole phase of the cardiac cycle, wherein the one or more components of the total mechanical energy input into the cardiac chamber is further based on the average ventricular pressures.UCLA 2024-258-2 UCLA.P0211WO

[0020] In some aspects, which may be combined with other aspects, the plurality of events of the cardiac cycle includes: an end diastole or a closing of an atrioventricular (AV) valve; an opening of a semilunar valve; an end systole or a closing of the semilunar valve; and an opening of the AV valve.

[0021] In some aspects, which may be combined with other aspects, the non-invasive cardiac parameters include: diastolic, systolic, and mean blood pressures; an EDV and an ESV of the cardiac chamber; and a blood flow rate at a ventricular outlet of the cardiac chamber.

[0022] In some aspects, which may be combined with other aspects, the method further includes: adjusting, by the processor, the patient-specific cardiac efficiency value based on one or more of: a basal metabolism of the cardiac chamber, an electrical excitation of the cardiac chamber, or an efficiency of mitochondrial oxidative phosphorylation in the cardiac chamber.

[0023] In some aspects, which may be combined with other aspects, the heart disease is congenital heart disease (CHD).

[0024] In some aspects, which may be combined with other aspects, the severity of the heart disease includes an absence of the heart disease in the patient.

[0025] In some aspects, which may be combined with other aspects, the method further includes: generating a treatment recommendation based on the classified severity of the heart disease, the treatment recommendation including one or more of: a medication regimen based on a medication type, a dosage, and an administration timeline; or a surgical intervention.

[0026] In some aspects, which may be combined with other aspects, the treatment recommendation is further based on a mechanical energy distribution of the total energy input among phases of the cardiac cycle, the phases including: an isovolumic contraction, a systolic ejection, an isovolumic relaxation, and a diastolic filling.

[0027] In some aspects, which may be combined with other aspects, the cardiac chamber includes one or more ventricles.

[0028] In at least one aspect, a system is disclosed for non-invasively assessing cardiac efficiency. The system includes: a processor; and memory storing instructions that, when executed by the processor, cause the processor to perform one or more methods described herein. For example, the instructions, when executed, can cause the processor to: receive patient-specific image data of a cardiac chamber of a patient, the patient-specific image data acquired from an imaging modality overUCLA 2024-258-2 UCLA.P0211WO a cardiac cycle; receive patient-specific vitals data; derive non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; apply the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generate, in real time, based on a ratio of the total energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient- specific cardiac efficiency value classifying a severity of a heart disease of the patient.

[0029] In at least one aspect, a non-transitory computer-readable medium (CRM) is disclosed for non-invasively assessing cardiac efficiency. The non-transitory CRM has stored thereon computer- readable instructions executable to cause performance of operations including any one or more methods described herein. For example, the operations may include: receiving, by a processor, patient-specific image data of a cardiac chamber of a patient, the patient-specific image data acquired from an imaging modality over a cardiac cycle; receiving, by the processor, patient-specific vitals data; deriving, by the processor, non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; applying, by the processor, the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generating, in real time, based on a ratio of the total energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient-specific cardiac efficiency value classifying a severity of a heart disease of the patient.

[0030] Other aspects, features, and implementations will become apparent to those of ordinary skill in the art, upon reviewing the following description of specific, exemplary aspects in conjunction with the accompanying figures. While features may be discussed relative to certain aspects and figures below, various aspects may include one or more of the advantageous features discussed herein. In other words, while one or more aspects may be discussed as having certain advantageous features, one or more of such features may also be used in accordance with the various aspects. In similar fashion, while exemplary aspects may be discussed below as device, system, or method aspects, the exemplary aspects may be implemented in various devices, systems, and methods.

[0031] The foregoing has outlined, rather broadly, the features and technical advantages of examples according to the disclosure in order that the detailed description that follows may be better understood. Additional features and advantages will be described hereinafter. The conception and specific examples disclosed may be readily utilized as a basis for modifying or designing otherUCLA 2024-258-2 UCLA.P0211WO structures for carrying out the same purposes of the present disclosure. Such equivalent constructions do not depart from the scope of the appended claims. Characteristics of the concepts disclosed herein, both their organization and method of operation, together with associated advantages will be better understood from the following description when considered in connection with the accompanying figures. Each of the figures is provided for the purposes of illustration and description, and not as a definition of the limits of the claims.

[0032] While aspects and implementations are described in this application by illustration to some examples, those skilled in the art will understand that additional implementations and use cases may come about in many different arrangements and scenarios. Innovations described herein may be implemented across many differing platform types, devices, systems, shapes, sizes, and packaging arrangements. While some examples may or may not be specifically directed to use cases or applications, a wide assortment of applicability of described innovations may occur. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In addition to the features described herein, additional features and variations will be readily apparent from the following descriptions of the drawings and exemplary embodiments. It is to be understood that these drawings depict embodiments and are not intended to be limiting in scope.

[0034] FIG.1 is a graph showing the relationship between blood pressure and blood flow rate across a cardiac cycle, according to example embodiments of the present disclosure.

[0035] FIG.2 is a graph showing an example ventricular pressure-volume (PV) loop in a cardiac cycle, according to example embodiments of the present disclosure.

[0036] FIG. 3 is an illustration of an example network environment for non-invasively assessing cardiac efficiency, according to example embodiments of the present disclosure.

[0037] FIG.4 are graphs showing cardiac catheterization-based waveforms of the aortic pressure and systemic ventricular blood pressures, according to example embodiments of the present disclosure.

[0038] FIG.5 shows graphs comparing non-invasive assessments of cardiac efficiency to catheter based invasive measurements of cardiac efficiency according to example embodiments of the present disclosure.UCLA 2024-258-2 UCLA.P0211WO

[0039] FIGs. 6A-6B are block diagrams illustrating example computer-implemented methods for non-invasively assessing cardiac efficiency according to non-limiting embodiments of the present disclosure.

[0040] FIG.7 is a block diagram illustrating a computer system upon which embodiments of the present teachings may be implemented. DETAILED DESCRIPTION

[0041] This specification describes various exemplary embodiments of systems, software and methods for non-invasively assessing cardiac efficiency. The disclosure, however, is not limited to these exemplary embodiments and applications or to the manner in which the exemplary embodiments and applications operate or are described herein.

[0042] Unless otherwise defined, scientific and technical terms used in connection with the present teachings described herein shall have the meanings that are commonly understood by those of ordinary skill in the art. Further, unless otherwise required by context, singular terms shall include pluralities, and plural terms shall include the singular.

[0043] As discussed, measuring cardiac efficiency has been proposed as a way to diagnose adverse effects of heart diseases like CHD. As a pump, the heart’s function has been described by its mechanical efficiency, which in engineering, refers to the ratio between mechanical work performed by a device and the energy input within a defined time duration. Because the heart functions as an operating pump, systolic function can be coupled to myocardial oxygen consumption. However, variables such as heart rate, contractility, and wall stress can influence oxygen requirements on which the systolic function is coupled to. To calculate ventricular mechanical work, continuous measurement of force and ventricular wall curvature is needed. These are vector quantities that are experimentally challenging to measure directly. Consequently, pressure is substituted for force and by convention, ventricular mechanical work or the energy generated by the ventricle has been described as the area in the pressure-volume relationship that encompasses both external energy (net energy expended in moving blood; also termed “external work” or “stroke work”) and potential energy.

[0044] As discussed, there is a desire and need for accurate and reliable systems and methods for assessing cardiac pressure and efficiency non-invasively, for example, to assess congenital heart disease patients, especially because existing methods are either invasive (e.g., involve occlusion orUCLA 2024-258-2 UCLA.P0211WO catheterization), are insensitive to atrioventricular and semilunar valve regurgitation, and / or are not adaptive to CHD patients with complex anatomies or symptoms such as SVP patient.

[0045] The present disclosure describes a novel and nonobvious solution to the aforementioned shortcomings by describing a system and method for accurately, reliably, efficiently, and non- invasively assessing cardiac pressure and efficiency. In various embodiments, the systems and methods overcome the limitations of existing methods for estimating cardiac efficiency in heart diseases such as complex CHD by using a novel and nonobvious model for noninvasive method for determining cardiac efficiency based on thermodynamic analysis of cardiac pressure-volume loops. For example, various embodiments of the present disclosure account for the (systemic) ventricle in SVP patients having lower efficiency compared to the (systemic) left ventricle in normal subjects, thereby resulting in lesser cardiac efficiency. As single ventricle palliation could alter the total vascular resistance and ventricular pressure-volume loading, cardiac efficiency may provide insight into the physiological effect of single ventricle palliation stages (pre-single ventricle palliation, post- Glenn, and post-Fontan) on cardiac energetics. Non-invasively obtained cardiac efficiency using the techniques described herein can more accurately and reliably measure ventricular performance in CHD patients with complex anatomy. Various embodiments of the present disclosure eliminate the need for invasive occlusion of the inferior vena cava or aorta. Furthermore, the particular techniques for non-invasively assessing cardiac efficiency described herein are also sensitive to and thereby address the energy loss caused by valvular regurgitation, by utilizing the blood flow rate at ventricular outflows. The presently disclosed techniques for non-invasively assessing cardiac efficiency can be estimated using radiographic and hemodynamic metrics obtained during routine clinical exams and is expected to provide additional insight into ventricular function, which may facilitate diagnostic evaluation and medical or surgical treatment planning for complex CHD.

[0046] FIG. 1 shows graphs illustrating the relationship between blood pressure and blood flow rate across a cardiac cycle, according to example embodiments of the present disclosure. The cardiac cycle may include a diastole phase and a systole phase. During the diastole phase, the ventricles of a heart may be relaxed (e.g., not contracting), allowing blood to flow from atria of the heart into the ventricles of the heart via atrioventricular valves. During the systole phase, the ventricles may contract and eject blood into the aorta and pulmonary artery.

[0047] As shown in FIG.1, an aortic blood flow rate 102 for blood in the heart may be steady until there is an opening of the semilunar valve, which causes the flow rate to increase during systolic ejection period. During the systolic ejection, the semilunar valve begin to close, thus reducing theUCLA 2024-258-2 UCLA.P0211WO blood flow rate from its spike until the end of the systolic period (end systole). After the end systole, the aortic blood flow rate 102 remains stable again until the next time the semilunar valve opens (e.g., in the next cardiac cycle). FIG. 1 further shows the behavior of the aortic pressure, 104, the left ventricle pressure 106, and left atrial pressure 108 during this cardiac cycle. As shown in FIG.1, the aortic pressure similarly mirrors the pattern of the flow rate when the semilunar valve opens, by increasing in pressure , stabilizing (e.g., as the semilunar valve closes), and then falling after the end systole. The left ventricular pressure 206 typically increases at the end of a diastole period (e.g., before the semilunar valve opening), rises in pressure, falls in pressure, and then stabilizes after the atrioventricular valve is opened. The left atrial pressure 208 remains relatively stable during the duration of the cardiac cycle.

[0048] FIG.2 is a graph showing an example ventricular pressure-volume (PV) loop in a cardiac cycle, according to example embodiments of the present disclosure. An ideal cardiac cycle may be characterized by a relationship between ventricular pressure and volume that is shown by the loop (referred to herein as “ventricular pressure-volume loop” or “pressure-volume loop”) in FIG.2. The four circles in the pressure-volume loop represent four events in a cardiac cycle – the end diastole, which may be characterized as when an atrioventricular (AV) valve closes (shown as event 1); a semilunar valve opening (shown by event 2); an end systole, which may be characterized as when the semilunar valve closes (shown as event 3); and an AV valve opening (shown as event 4).

[0049] Furthermore, there are typically underlying processes leading to these events in the cardiac cycle. For example, after the end diastole (event 1), the ventricles of the heart may undergo isovolumic contraction 202, where energy (Êisovolumic contraction) may be expended and lost as heat. The semilunar valve opening (event 2) may cause the ventricles to undergo systolic ejection 204 where mechanical energy (Êsystolic ejection) may be used to contract the ventricles and eject blood flow. After the end systole (event 3), the ventricles may undergo isovolumic relaxation 206, where the ventricles may relax and cause energy to be lost as heat, the loss of energy here may be referred to herein as Êisovolumic relaxation. In embodiments, the aforementioned energies may be represented as ^^^^ ^^^^^^^or the energy added to the ventricular blood-volume during systole (isovolumic contraction and systolicejection) such that ^^^^ ^^^^^^^ = ^(^^^^^^^^^^^^ ^^^^^^^^^^^, ^^^^^^^^^^ ^^^^^^^^, ^^^^^^^^^^^^ ^^^^^^^^^^). Forexample, ^^^^^^^^^^^^ ^^^^^^^^^^may represent the energy expended by the ventricular myocardium during systole to develop tension and maintain pressure at end systole but is wasted during isovolumic relaxation (e.g., as heat), where the pressure falls, and no work is done. The relaxed ventricles may result in diastolic filling 208, where mechanical energy (Êdiastolic filling) may be used to add blood flowUCLA 2024-258-2 UCLA.P0211WO into the ventricles. The total mechanical energy input into the ventricular blood-volume over a cardiac cycle (^^^^) may be a function of mechanical energy added during systole and diastolic filling such

[0050] Furthermore, each event of the cardiac cycle (events 1-4) may be defined by different ventricular blood pressures. In particular, ^^^^ ^^^^^^^^,^^^^ ^^^^^^^, and ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^may represent the ventricular blood pressures at the end diastole or atrioventricular valve closure (event 1), semilunar valve opening (event 2), end systole or semilunar valve closure (event 3), and atrioventricular valve opening (event 4), respectively. Additionally, ^^^^^^^^^^ ^^^^^^^^^ ! ^^^^^^^^^^^ ^^^^^^^may represent the mean ventricular blood pressure during systolic ejection and diastolic filling phases, respectively. FIG.2 further shows, based on the volumes characterizing the boundaries of the pressure-volume loop in the cardiac cycle, an end-systolic volume (ESV), an end-diastolic pressure volume relationship (EDPVR), an end-systolic pressure volume relationship (ESPVR), the myocardial potential energy following ejection (PE), and the stroke work (SW). As will be discussed herein, various parameters of the pressure-volume loop for CHD and other patients with abnormal cardiac physiology may be obtained non-invasively to be used to determine the cardiac efficiency for such patients. EDPVR and ESPVR for such patients, as such curves may deviate from the idealized EDPVR and ESPVR shown in FIG.1. In particular, the volume axis intercept of EDPVR and ESPVR ("#) may be non-zero and / or otherwise patient-specific, such as being large or negative in patients with ischemia or infarcted ventricle. For example, image data and patient vitals data may be used to determine mechanical energy fluxes during each cardiac phases namelyenergy at ventricular outflow(s) ^^^^^to non-invasively determine the cardiac efficiency. Our method eliminates the need to estimate EDPVR, ESPVR, and "#values to estimate cardiac efficiency, which makes it easier, more accurate and repeatable to estimate cardiac efficiency both invasively and non- invasively. Example Network Environment

[0051] FIG. 3 is an illustration of an example network environment for non-invasively assessing cardiac efficiency, according to example embodiments of the present disclosure. The networkUCLA 2024-258-2 UCLA.P0211WO environment may include but is not limited to one or more medical imaging modalities 302 configured to acquire image data of at least a portion (e.g., a ventricle) of a cardiac chamber of a patient heart 304, one or more patient vital devices 316 configured for acquiring patient vitals data and a computing device and / or system 324. One or more of the components of the network environment may be communicatively coupled and exchange data via a wire connection or wirelessly. Also or alternatively, one or more of the components may be communicatively coupled via a communication network 350. In some aspects, the computing device and / or system 324 may be local. In some aspects, the computing device and / or system 324 may be located remotely. FIG.7 describes an example of the computing device and / or system 324 in additional detail.

[0052] The medical imaging modality 302 may comprise, for example, magnetic resonance imaging (MRI), multiphase, steady-state imaging with contrast enhancement (MUSIC), 4D Flow MRI, computed tomography (CT), ultrasound, or a combination thereof. The image data acquired by the medical imaging modality 302 may be of at least the portion (e.g., the ventricle) of the patient heart 304 over a cardiac cycle. The cardiac cycle may be characterized by a plurality of phases comprising isovolumic contraction, systolic ejection, isovolumic relaxation, and diastolic filling phases. The patient vitals device(s) 316 may comprise one or more devices, instruments, or sensors configured to acquire patient vitals data, such as blood pressure (e.g., peripheral non-invasive blood pressure (NIBP), heart rate, etc.), and may include, for example, a blood pressure sensor 318 (e.g., sphygmomanometer) or a heart rate sensor 320.

[0053] FIG.3 further shows an example ventricle 304 (e.g., the portion of the patient heart) from which the image data may be acquired. The dashed lines of the ventricle 304 show the control volume representing ventricular blood-volume. As discussed herein, various embodiments for non-invasive methods of assessing cardiac efficiency (e.g., ventricular efficiency) may be based on a ratio of total energy output for blood flow out of the ventricle during the cardiac cycle (also referred to as total energy output or ^^^^^) to the total energy input for blood flow into the ventricle during the cardiac cycle (also referred to as the total energy input or ^^^^). FIG.3 thus shows ^^^^^310 and two sources of ^^^^based on blood flow into the ventricle during two respective phases – systolic ejection (^^^^ ^^^^^^^) and diastolic filling (^^^^ ^^^^^^^^^ ^^^^^^^). ^^^^ ^^^^^^^306 represent mechanical energy added to ventricular blood-volume during systole (1^2 and 2^3 in Figure 2) i.e., ^^^^ ^^^^^^^=^(^^^^^^^^^^^^ ^^^^^^^^^^^, ^^^^^^^^^^ ^^^^^^^^, ^^^^^^^^^^^^ ^^^^^^^^^^) while ^^^^ ^^^^^^^^^ ^^^^^^^ 308represents mechanical energy added to ventricular blood-volume during diastolic fillingUCLA 2024-258-2 UCLA.P0211WO(^^^^ ^^^^^^^^^ ^^^^^^^ = ^(^^^^^^^^^^^ ^^^^^^^)). Total mechanical energy input to the ventricular blood-volume (^^^^) can thus be approximated to a function of mechanical energy added during systole(^^^^ ^^^^^^^) and diastolic filling (^^^^ ^^^^^^^^^ ^^^^^^^)^^^^ = ^(^^^^ ^^^^^^^ , ^^^^ ^^^^^^^^^ ^^^^^^^). Thepresently disclosed systems and methods for non-invasive assessment of cardiac efficiency also account for regurgitations in the semilunar valve 312 and atrioventricular valve 314.

[0054] The computing device and / or system 324 may receive the patient-specific image data of the cardiac chamber 304 (e.g., one or more ventricles or other part of or entirety of the cardiac chamber) of the patient heart as well as patient-specific vitals data to perform one or more methods described herein for non-invasively assessing the cardiac efficiency of the patient. The computing device / system 324 may further store or may access one or more models 326 (e.g., machine learning models, regression models, etc.) trained to predict a patient-specific cardiac efficiency value. The non- invasively obtained patient-specific vitals data and the patient-specific image data may collectively be referred to herein as non-invasive cardiac parameters 322. Also or alternatively, the non-invasive cardiac parameters 322 may refer to data derived from the patient-specific image data and the patient- specific vitals data that may be input into the trained model(s) 326. Such input data may include but is not limited to volumic information (e.g., end systolic volume (ESV), end diastolic volume (EDV)), non-invasive blood pressure measurements (e.g., diastolic, systolic, and mean blood pressures at limb, aorta, etc.), blood flow rates (e.g., at aorta, pulmonary artery, etc.), heart rate, ECG signals, and the like. Example Model Training for Assessment of Cardiac Efficiency

[0055] Various embodiments of the present disclosure describe novel and nonobvious models (e.g., machine learning models, regression models) for estimating cardiac efficiency based on thermodynamic analysis using non-invasive cardiac parameters. The non-invasive cardiac parameters may include non-invasively obtained or derived patient-specific imaging data and non-invasive blood pressure measurements. The presently disclosed models (also referred to herein as cardiac efficiency models, or "^'(^^^^model(s)) are based on a thermodynamic analysis of the ventricular pressure- volume loop (e.g., as previously discussed in relation to FIG. 2). The models may be based on or include machine learning models, such as a set of linear regression models that are specifically trained for estimating aspects of cardiac efficiency. In some embodiments, non-invasive cardiac imaging (e.g., cardiovascular magnetic resonance imaging, echocardiography, cardiac CT) of a cardiac region of a patient may be used to estimate or determine the ventricular volume and blood flow rate atUCLA 2024-258-2 UCLA.P0211WO ventricular outlet(s) of the patient. Furthermore, non-invasive blood pressure measurements may be used to estimate aortic and ventricular blood pressures based on models described below. These non- invasive estimations of the ventricular volume and blood flow rate may be used as inputs to one or more models described herein for noninvasively assessing cardiac efficiency. The one or more models may be trained to output the aortic and ventricular blood pressure based on such inputs.

[0056] In some embodiments, the model may be trained using reference training data, such as but not limited to blood pressure measurements of reference patients at key cardiac events (e.g., events 1- 4 described in relation to FIG.2). The blood pressure measurements for this reference training data may have been received non-invasively or invasively. For example, blood pressure measurements at the key cardiac events for the training data may be obtained from reference patients using cardiac catheterization.

[0057] FIG. 4 are graphs showing cardiac catheterization-based waveforms of the aortic blood pressure and systemic ventricular blood pressure, according to example embodiments of the present disclosure. In particular, graph 410 shows a catheterization-based waveform of the aortic pressure, while graph 420 shows the catheterization-based waveform of the ventricular blood pressure at key cardiac events – atrioventricular valve opening 422 (e.g., event 4 in FIG.2), end diastole 424 (event 1 in FIG. 2), semilunar valve opening 426 (event 2 in FIG.2), and end systole 428 (event 3 in FIG. 2).

[0058] One or more models may be trained to predict ventricular blood pressures at the key cardiac events (e.g., events 1-4 as shown in FIG.2), using training data comprising the actual blood pressure measurements at the key cardiac events, such as but not limited to the catheterization-based waveforms of the aortic and ventricular pressure of FIG. 4. The trained model may receive patient- specific parameters derived from patient-specific imaging data and patient-specific vitals data. The trained model may use such patient-specific parameters (e.g., ventricular volumes and blood flow rate) as inputs, and may output the estimation of aortic and ventricular blood pressures at the key cardiac events. The outputted aortic and ventricular blood pressures can then be used to determine ventricular efficiency "^'(^^^^as described herein.

[0059] In at least one embodiment, the model may be trained based on the known aortic and ventricular blood pressure waveforms of reference patients, such as those shown in FIG.4. In some aspects, derivatives of the waveforms (e.g., the first- and second-time derivatives) may be obtained to bolster the training. In some embodiments, the training may include identifying the aortic pressure atUCLA 2024-258-2 UCLA.P0211WO a dicrotic notch (^^^^^^^ ^^^^^^^^ ^^^^() 412. The dicrotic notch may be defined as the state during a descent of the aortic pressure after peak systole where there is a local minimum in the aortic pressure waveform 410. In some embodiments, the dicrotic notch 412 may be identified based on identifying when the first derivative of the aortic pressure waveform 410 is zero with respect to time (e.g., identifying a local minimum). If the first derivative does not yield a zero value in the aortic pressure waveform 410, the dicrotic notch 412 may be identified by obtaining the maximum or a local maxima of second derivative of the waveform 410. Also or alternatively, the pressure at the dicrotic notch can be estimated from the timing of certain events spanning the aortic pressure waveform 410. For example, the timing of the semilunar valve closure can be estimated using blood flow rate measured at the semilunar valve root. The blood pressure measured at the time corresponding to zero or negative blood flow at the semilunar valve root can be used to provide the blood pressure at the dicrotic notch (e.g., dicrotic notch pressure 412).

[0060] The model may be further trained based on known ventricular pressure values at key cardiac events, as provided, for example, by the catheterization based ventricular pressure waveform 420. The key cardiac events relevant for the ventricular pressure may include but are not limited to the atrioventricular valve opening 422, (e.g., event 4 in FIG. 2), end diastole 424 (event 1 in FIG. 2), semilunar valve opening 426 (event 2 in FIG.2), and end systole 428 (event 3 in FIG.2), as shown in the ventricular pressure waveform 420 in FIG.4.

[0061] The ventricular pressure during atrioventricular (AV) valve opening 422 (^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^) may be determined by identifying the AV opening 422 cardiac event in the ventricular pressure waveform 420. The AV valve opening 422 may be identified as the state during the diastole phase of the cardiac cycle where the second derivative of the ventricular pressure waveform 420 may be at its maximum and / or a local maxima. Furthermore, the AV valve opening 422 may correspond to or be based on a point or region of the ventricular pressure waveform 420 where the first derivative of the ventricular pressure waveform 420 is negative and increasing in pressure with respect to time. The ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^may correspond to the pressure at beginning of ventricular diastolic filling. In some aspects, the ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^may approximate mean atrial pressure.

[0062] The ventricular pressure during end diastole 424 (^^^^ ^^^^^^^^) may be determined by identifying the end diastole 424 cardiac event in the ventricular pressure waveform 420. The end diastole 424 may be identified as the state during the diastole phase of the cardiac cycle where thereUCLA 2024-258-2 UCLA.P0211WO is the last zero crossing in the 1stderivative of the ventricular pressure waveform 420 with respect to time. In case the zero crossing in the 1stderivative of the ventricular pressure waveform 420 is absent as observed in some cases, the end diastole 424 cardiac event can be identified by locating the zero crossing for the 2ndderivative of the ventricular pressure waveform 420. In case the zero crossing in the 2ndderivative of the ventricular pressure waveform 420 is absent as observed in some cases, the end diastole 424 cardiac event can be identified by locating peak (R-wave) in ECG signal.

[0063] The ventricular pressure during the semilunar valve opening 426 (^^^^^^^^^^ ^^^^^ ^^^^^^^) may be determined by identifying the semilunar valve opening cardiac event 426 in the ventricular pressure waveform 420. The semilunar valve opening 426 may be identified as the state during the systole phase of the cardiac cycle where the 1stderivative of the ventricular pressure waveform 420 is at its maximum or at a local maxima. Furthermore, the semilunar valve opening 426 may be located as where the 2ndderivative of the ventricular pressure waveform 420 has a 0 crossing in pressure with respect to time. The ^^^^^^^^^^ ^^^^^ ^^^^^^^may correspond to the aortic diastolic blood pressure (^^^^^^^ )*+) and may occur at the beginning of ventricular systolic ejection.

[0064] The ventricular pressure during the end systole 428 (^^^^ ^^^^^^^) may be determined by identifying the end systole cardiac event 428 in the ventricular pressure waveform 420. The end systole 428 may be identified as the state during the systole phase of the cardiac cycle where the 2ndderivative of the ventricular pressure waveform 420 is at its minimum or at a local minima. Also or alternatively, the end systole 428 can be identified or located as where the 1stderivative of the ventricular pressure waveform 420 is negative and decreasing with respect to time. The ^^^^ ^^^^^^^may correspond to aortic dicrotic notch pressure (^^^^^^^ ^^^^^^^^ ^^^^() and may be towards the end of the ventricular systolic ejection.

[0065] The systemic ventricle’s average pressure during systolic ejection (^^^^^^^^^ ^^^^^^^^) and average pressure during diastolic filling (^^^^^^^^^^^ ^^^^^^^) were obtained by taking the average of the data points along the pressure waveform during systolic ejection (e.g., curve 204 between event 2 (e.g., semilunar valve opening) and event 3 (semilunar valve closure) in FIG.2) and diastolic filling (e.g., curve 208 between event 4 (AV valve opening) and event 1 (AV valve closure) in FIG. 2), respectively.

[0066] In some embodiments, the one or more linear regression models trained for non-invasively assessing cardiac efficiency may include one or more linear regression models. For example, forUCLA 2024-258-2 UCLA.P0211WO systemic single ventricle with no severe valve disease, linear regression models can be obtained as shown below:

[0067] ^^^^^^^ ,*+ = ^(-1 × ^01*+ ,*+, -2 × ^01*+ )*+, -3 × ^01*+ 45+)By combining the above expressions, one may obtain the following additional linear regression models:

[0074] In the aforementioned equations, ^^^^^^^ ,*+and ^^^^^^^ )*+may represent aortic systolic and aortic diastolic blood pressures, respectively, and can be approximated using patient-specificvitals data (e.g. ^01*+ ,*+, ^01*+ )*+, ^ ! ^01*+ 45+ represent systolic, diastolic, and mean non-invasive blood pressure, respectively) that may be non-invasively received. Coefficients -1 ?@ -16 may be estimated using the models trained using the cardiac catheterization-based blood pressure waveforms, and non-invasive cardiac parameters (e.g., vitals data, and non-invasive blood pressure measurements, etc.) of reference patients.

[0075] In some embodiments, the presently disclosed model for ventricular mechanical efficiency ("^'(^^^^) may be estimated based on the mechanical energy (^^) during each cardiac phase, denotedis a function of pressure (^) and volume (") during the respective cardiac phase, i.e. ^^ = ^(^, ") .

[0076] A total mechanical energy input to ventricle (^^^^) during a cardiac cycle can be a function of the mechanical energy during each cardiac phase i.e.,UCLA 2024-258-2 UCLA.P0211WO

[0077] ^^^^ = ^A^^^^^^^^^^^^ ^^^^^^^^^^^, ^^^^^^^^^^ ^^^^^^^^, ^^^^^^^^^^^^ ^^^^^^^^^^, ^^^^^^^^^^^ ^^^^^^^B

[0078] The time rate of total mechanical energy input (^^^C^ ) may be obtained using the expression,

[0079] ^^C^^ =, where KL is heart rate in beats per minute.

[0080] An average power output in the blood flowing at the ventricular outlet (^^^C^^) may be given by

[0082] "^'(^^^^can thus be defined as the ratio of average power in the blood flowing through the ventricular outlet, ^^^C^^, to the average power input into the ventricular blood-volume, ^^^C^ , over a complete cardiac cycle (Eq.6). Hence, "^'(^^^^is some function ^ of ventricular and aortic blood pressure (^), ventricular blood-volume ("), blood flow rate at the ventricular outlets (N) and the heart rate (KL):

[0084] To facilitate the non-invasive estimation of ventricular diastolic blood pressure, a set of non-linear equations representing mechanical energy distributions during each cardiac phase can be solved. The set of non-linear equations may include:

[0088] In the aforementioned equations, coefficients -17 ?@ -19 may be estimated using cardiac catheterization-based blood pressure waveforms from reference patients. Model Results and ValidationUCLA 2024-258-2 UCLA.P0211WO

[0089] FIG. 5 shows graphs comparing non-invasive assessments of cardiac efficiency with catheter based invasive measurements of cardiac efficiency according to example embodiments of the present disclosure. The presently disclosed models for non-invasive assessments of cardiac efficiency were validated using a single ventricle physiology (SVP) patient’s hemodynamic measurement.

[0090] The present disclosure found that there was a strong correlation (r=0.99, N=13 SVP patients) and small mean difference (mean difference = 0.38 %, N=13 SVP patients) between ventricular efficiency models ("^'(^^^^) estimated using conventional invasive methods and the presently disclosed non-invasive method for determining ventricular efficiency, as shown in FIG.5. The results suggest that the presently disclosed methods for non-invasive ventricular efficiency estimation using "^'(^^^^can accurately and reliably predict actual ventricular efficiency measured using invasive techniques with precision. Example Computer Implemented Method

[0091] FIGs. 6A-6B are block diagrams illustrating example computer-implemented methods for non-invasively assessing cardiac efficiency according to non-limiting embodiments of the present disclosure. In particular, FIG. 6A illustrates an example computer-implemented method for non- invasively assessing cardiac efficiency using a trained model, while FIG. 6B illustrates an example computer-implemented method for training the model used for non-invasively assessing the cardiac efficiency. One or more blocks or processes described in the blocks of FIGs.6A-6B may be performed by one or more computing devices (e.g., such as but not limited to computing device / system 100, and computing system 700 as will be described herein). For example, the one or more blocks or processes may be performed by the processor 704 based on instructions provided by any one of, or a combination of memory components 706 / 708 / 710 and user input (e.g., provided via the input device 714), as will be discussed herein.

[0092] Referring to FIG.6A, at block 602, a processor (e.g., processor 704) may receive patient- specific image data of a cardiac chamber of a patient over a cardiac cycle. The cardiac chamber may comprise, for example, one or more of a ventricle, a valve, an atrium, or a septum of the patient heart. The patient-specific image data may be acquired from an imaging modality, such as, but not limited to: magnetic resonance imaging (MRI); multiphase, steady-state imaging with contrast enhancement (MUSIC); computed tomography (CT); ultrasound; or the like. A cardiac cycle may include a plurality of phases, including isovolumic contraction, systolic ejection, isovolumic relaxation, and diastolic filling.UCLA 2024-258-2 UCLA.P0211WO

[0093] In some embodiments, based on the patient-specific image data, the computing device may determine an end diastolic volume (EDV) and an end-systolic volume (ESV) of the patient cardiac chamber (e.g., the EDV and ESV of the patient ventricle). In some embodiments the computing device may further extract or determine additional non-invasive cardiac parameters from the patient-specific image data.

[0094] At block 604, the computing device may receive, by the processor, patient-specific vitals data. The patient-specific vitals data may include or comprise a heart rate associated with the patient. For example, the heart rate may be acquired from a patient vitals device 316 (e.g., HR sensor 320) and may be transmitted to the computing device (e.g., via a communication network 350) or may be received by the computing device via user input. In some embodiments, the patient-specific vitals data may be retrieved from non-invasive, invasive sources, or a combination thereof. For example, in some aspects, blood pressure may be obtained invasively via catheterization. In some embodiments, patient-specific vital data may be received entirely non-invasively.

[0095] In some embodiments, the patient-specific vital data may further include non-invasively received patient-specific blood pressure, such as a non-invasive blood pressure (NIBP) (e.g., peripheral NIBP, aortic NIBP, etc.). For example, the NIBP may be acquired from the patient vitals device 316 (e.g., BP sensor 318) and may be transmitted to the computing device (e.g., via a communication network 350) or may be received by the computing device via user input. In some embodiments, the NIBP may be used to determine ventricular pressures at two or more phases or boundaries of phases of the cardiac cycle.

[0096] At block 606, the computing device may derive non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data. For example, the computing device may be configured to process the patient-specific image data to extract non-invasive cardiac parameters, such as but not limited volume measurements of the cardiac chamber at various events or phases of the cardiac cycle (e.g., end systole volume (ESV) and end diastole volume (EDV)). In some embodiments, the volume measurement of the cardiac chamber may be extracted using machine learning techniques, such as but not limited to convolutional neural networks, classification. In some embodiments, prior to obtaining the volume measurements, the cardiac chamber or a component of the cardiac chamber may be recognized, for example, via segmentation, skeletonization, and / or image recognition. In some embodiments, for example, where the patient-specific image data is a video or an image data stream, the point or duration of time corresponding to the systole or diastole phase of the cardiac cycle may be determined, in order to obtain the correct volume measurements.UCLA 2024-258-2 UCLA.P0211WO Furthermore, the computing device may extract non-invasive cardiac parameters from the patient- specific vitals data, such as but not limited to the heart rate and the patient-specific blood pressure measurements (e.g., NIBP).

[0097] At block 608, the computing device may apply the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber (e.g., ventricle) during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber (e.g., ventricle) during the cardiac cycle. In some embodiments, for example, where the computing device has determined the EDV and the ESV of the patient cardiac chamber (e.g., based on the patient- specific image data), the computing device may apply the EDV and ESV of the patient cardiac chamber into the trained model. In some embodiments, applying the non-invasive cardiac parameters to the trained model of the cardiac cycle may include optimizing the model, for example, by optimizing the ranges of coefficients or parameters in the model (e.g., by adjusting or finding solutions within ranges of ventricular pressures at various phases of the cardiac cycle).

[0098] In at least one embodiment, non-invasive cardiac parameters of the patient may be applied to a model comprising expressions for total energy input and total energy output. The expressions approximate the energy inputs and energy output of a pressure-volume loop (e.g., as described in relation to FIG. 2), and are based on a set of trained linear regression models that approximate ventricular pressures. For example, various non-invasive parameters (e.g., EDV and ESV) may be applied to the total energy input expressions approximating the ventricular systolic ejection and diastolic filling phases of a cardiac cycle (e.g., based on isobaric processes), while various non- invasive parameters (e.g., systolic blood pressure, diastole blood pressure) may be applied to a set of linear regression models to determine ventricular pressures, average pressures, and / or blood flow rates that are then inputted into the total energy input and total energy output expressions.

[0099] In at least one embodiment, the total energy input can be divided into components corresponding to mechanical energy fluxes across the ventricular blood-volume during each phase of the cardiac cycle. The various components of the total energy input can thus be expressed as:UCLA 2024-258-2 UCLA.P0211WO

[0104] where ^^^^^^^^^^^^ ^^^^^^^^^^^represents energy added to ventricular blood-volume during the isovolumic contraction phase of a cardiac cycle, ^^^^^^^^^^ ^^^^^^^^represents boundary work done on ventricular blood-volume during the systolic ejection phase of a cardiac cycle, ^^^^^^^^^^^^ ^^^^^^^^^^represents energy leaving ventricular blood-volume during the isovolumic relaxation phase of a cardiac cycle, and ^^^^^^^^^^^ ^^^^^^^represents boundary work done by blood entering ventricle during the diastolic filling phase of a cardiac cycle.

[0105] In the aforementioned expressions corresponding to components of the total energy input, ^^^^ ^^^^^^^^ , ^^^^^^^^^^ ^^^^^ ^^^^^^^, ^^^^ ^^^^^^^ , ^ ! ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^ may representventricular blood pressure at end diastole or atrioventricular valve closure, semilunar valve opening, end systole or semilunar valve closure, and atrioventricular valve opening, respectively; and ^^^^^^^^^^ ^^^^^^^^^ ! ^^^^^^^^^^^ ^^^^^^^may represent the mean ventricular blood pressure during systolic ejection and diastolic filling phases, respectively.

[0106] Conventionally, such ventricular pressure values are typically obtained invasively through catheterized waveforms, which leads to assessments for cardiac efficiency being expensive, risky, and intrusive. However, as previously discussed, a set of linear regression models may be trained to approximate such ventricular pressure values (e.g., ^^^^ ^^^^^^^^ , ^^^^^^^^^^ ^^^^^ ^^^^^^^,^^^^ ^^^^^^^ , ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^, ^^^^^^^^^^ ^^^^^^^^, ^ ! ^^^^^^^^^^^ ^^^^^^^) using patient-specific non-invasive cardiac parameters (e.g., non-invasive blood pressure (e.g., aortic and limb systolic, diastolic, and mean blood pressures) as inputs.

[0107] For example, a set of linear regression models can be obtained as shown below:

[0108] ^^^^^^^ ,*+ = ^(-1 × ^01*+ ,*+, -2 × ^01*+ )*+, -3 × ^01*+ 45+)UCLA 2024-258-2 UCLA.P0211WO

[0113] By substituting the above two expressions, one may obtain the following additional linear regression models:

[0116] where ^^^^^^^ ,*+and ^^^^^^^ )*+may represent the aortic systolic and aortic diastolic blood pressures, respectively, and which can be derived from the non-invasively obtained patient-specific vitals (e.g. ^01*+ ,*+, ^01*+ )*+, ^ ! ^01*+ 45+ represent systolic, diastolic, and mean non-invasiveblood pressure, respectively) data (blocks 604 and 606). The aforementioned set of linear regression models may learn or otherwise determine the coefficients -1 ?@ -16 through supervised learning using a training dataset based on cardiac catheterization-based blood pressure waveforms and cardiac parameters obtained from reference patients. Example embodiments of training the model used in block 608 are described herein, for example, in relation to FIG.7.

[0117] With the ventricular pressures determined (e.g., by inputting the non-invasive blood pressures into the set of trained linear regression models with the learned coefficients -1 ?@ -16), the model may further include the ventricular pressures in the expressions for the total energy input and the total energy output. For example, ^^^^ ^^^^^^^^ , ^^^^^^^^^^ ^^^^^ ^^^^^^^,^^^^ ^^^^^^^ , ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^, ^^^^^^^^^^ ^^^^^^^^, ^ ! ^^^^^^^^^^^ ^^^^^^^, determined usingthe set of linear regression models, may be included in expressions for determining the total energy input of blood flow into the cardiac chamber and the total energy output from blood flow out of the cardiac chamber. In some embodiments, the model for computing the total energy input and total energy output may assume that the energy flux entering the ventricular blood-volume during each cardiac phase (e.g., the total energy input) is sourced by atrial and ventricular muscle contraction and relaxation. When the ventricular blood-volume returns to its initial pressure-volume state after completing a cardiac cycle, the net change in energy of the ventricular blood-volume can be zero even though the myocardium would have consumed the mechanical energy governed bymechanical work and energy flux entering the ventricular blood-volume due to ventricular and atrial muscle contraction and relaxation over a complete cardiac cycle (^^^^) may be provided by the expression:UCLA 2024-258-2 UCLA.P0211WO

[0119] where ^^^^may be representative of the total energy input for blood flow into the cardiac chamber of the patient, whereas ^^^^^^^^^^^^ ^^^^^^^^^^^, ^^^^^^^^^^ ^^^^^^^^,^^^^^^^^^^^^ ^^^^^^^^^^, and ^^^^^^^^^^^ ^^^^^^^ may represent components of the total energy input thatcorrespond to each phases of the cardiac cycle. A time rate of the total energy input (^^^C^ ) can be determined based on the heart rate of the patient, which may be obtained non-invasively (e.g., via the patient-specific vitals data). The time rate of the total energy input (^^^C^ ) can thus be determined as: ^C^^ = D^EF×GH

[0120] ^, where HR is heart rate in beats per minute. It is contemplated that other units for the hear rate may vary the denominator in the above expression.

[0121] Applying the non-invasive cardiac parameters to the trained model may also generate the total energy output, or the mechanical energy of blood flowing out of the cardiac chamber (e.g., ventricular outlet). For example, the total energy output over time (^^^C^^) may be represented as ^^^C^^= ∑^^^^^^^^^^^ ^^^^^^ ^(^^ , N^) , where ^^ ^ ! N^ represent mean blood pressure and mean flow rate atventricular outlets over a complete cardiac cycle, respectively. The fluid power can be summed over multiple ventricular outlets if a ventricle has multiple outlets e.g. double outlet right ventricle. In some aspects, ^^^^^may result from ^^^^^^^^^^ ^^^^^^^^.

[0122] At block 610, the computing device may generate, in real time, a patient-specific cardiac efficiency value based on a ratio of the total energy output to the total energy input. In some aspects, the patient-specific cardiac efficiency value ("^'(^^^^) may be defined as the ratio of average power in the blood flowing through the ventricular outlet, ^^^C^^, to the average power input into the ventricular blood-volume, ^^^C^ , over a complete cardiac cycle. In some embodiments, the patient- specific cardiac efficiency value ("^'(^^^^) can be some function ^ of ventricular and aortic blood pressure (^), ventricular blood-volume ("), blood flow rate at the ventricular outlets (N) and the heart rate (KL):

[0124] In the aforementioned expression, the total energy input over time (^^^C^ ) and the total energy output over time (^^^C^^) can be determined by applying, as inputs, non-invasive cardiac parameters (e.g., EDV, ESV, non-invasive blood pressures, heart rates) to a trained model, as discussed in block 608. For example, non-invasively derived systolic and diastolic blood pressures may be inputted intoUCLA 2024-258-2 UCLA.P0211WO a set of linear regression models that are trained (e.g., have learned coefficients) that approximate ventricular pressures (e.g., mean ventricular pressures as well as ventricular pressures at key cardiac events). The ventricular pressures may be applied with the EDV and ESV to obtain one or more components of the total energy input corresponding to one or more phases of the cardiac cycle. Another non-invasive cardiac parameter, the heart rate may be used to convert each of the total energy input or the total energy output on a per cycle basis (e.g., over time) or as power. The total energy output over time (^^^C^^) may be represented as

[0125] ^^C^^^ = ∑^^^^^^^^^^^ ^^^^^^ ^(^^ , N^) , where ^^ ^ ! N^ represents mean blood pressure andmean flow rate at the ventricular outlets over a complete cardiac cycle, respectively. The fluid power may be summed over multiple ventricular outlets if a ventricle has multiple outlets e.g. double outlet right ventricle.

[0126] In some embodiments, the patient-specific cardiac efficiency value ("^'(^^^^) may be generated in real time or near real time. In some embodiments, the patient-specific cardiac efficiency value may be further adjusted based on one or more of: a basal metabolism of the patient cardiac chamber, an electrical excitation of the patient cardiac chamber, or efficiency of mitochondrial oxidative phosphorylation of the patient cardiac chamber.

[0127] The patient-specific cardiac efficiency value may thus diagnose and / or indicate a severity of heart disease (e.g., congenital heart disease). The severity may include, for example, a stage of the heart disease, or even the absence of any heart disease. In some embodiments, the computing device may generate, based on the classified severity of the heart disease and distribution of energiesrecommendation. The treatment recommendation may include but is not limited to, for example, a medication regimen based on a medication type, a dosage, and an administration timeline. Also or alternatively, the treatment recommendation may include a surgical intervention. The recommended surgical intervention, in some aspects, may identify a location for the surgical intervention.

[0128] FIG.6B illustrates an example computer-implemented method 600B for training the model used for non-invasively assessing the cardiac efficiency. For example, method 600B may be used to train the model which the non-invasive cardiac parameters are applied, at block 608, to obtain the total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and the total energy input for blood flow into the cardiac chamber during the cardiac cycle. One or more blocks or processes described in the blocks of FIG. 6B may be performed by one or more computing devicesUCLA 2024-258-2 UCLA.P0211WO (e.g., such as but not limited to computing device / system 324, and computing system 700 as will be described herein). Also or alternatively, one or more blocks of FIG. 6B may be performed by a computing device or server that is different from (e.g., remote from) the computing device performing method 600A of FIG.6A. Thus, in some embodiments, the training of the model may be performed remotely from the application of the model to non-invasively assess a cardiac efficiency of a patient.

[0129] At block 612, the computing device may receive blood pressure waveforms of respective reference patients. The blood pressure waveforms may be obtained from the plurality of reference patients via catheterization. The blood pressure waveforms may indicate ventricular pressure over time for at least a cardiac cycle. Graphs 410 and 420 of FIG. 4 provides an example of such blood pressure waveforms.

[0130] At block 614, the computing device may identify a flow rate and ventricular pressure at each event of a cardiac cycle for each blood pressure waveform. The events may include but are not limited to an atrioventricular valve opening (e.g., event 422 in FIG.4 and event 4 in FIG.2), end diastole or a closing of an atrioventricular (AV) valve (e.g., event 424 in FIG.4 and event 1 in FIG.2), semilunar valve opening (e.g., event 426 in FIG.4 and event 2 in FIG. 2), and end systole or a closing of the semilunar valve (e.g., event 428 and event 3 in FIG.2). As previously discussed in relation to FIG.4, such events and the ventricular pressures at such events may be identified using derivatives and / or slope analysis. The flow rate of the reference patient may be obtained invasively (e.g., via catheterization) or non-invasively (e.g., via imaging) over the cardiac cycle.

[0131] At block 616, the computing device may determine a total energy output and a total energy input for each reference patient. The total energy output and total energy input may be a function of the of ventricular and aortic blood pressure (^), ventricular blood-volume ("), blood flow rate at the ventricular outlets (N) and the heart rate (KL) of each reference patient.

[0132] In some embodiments, the total energy input may be divided into components corresponding to mechanical energy fluxes associated with different phases of the cardiac cycle. For example, the computing device may determine, based on the ventricular pressures, one or more components of a total energy input into the cardiac chamber of each respective reference patient. Thus, each component of the total energy input may include a mechanical energy distribution for a respective phase of the cardiac cycle. The respective phase may be among one or more phases of the cardiac cycle corresponding to the one or more components of the total mechanical energy input. The one or moreUCLA 2024-258-2 UCLA.P0211WO phases corresponding to the one or more components of the total mechanical energy input may include: an isovolumic contraction; an isovolumic relaxation; and a diastolic filling.

[0133] The total energy output (^^^C^^) estimated by the computing device may be represented as ^^C^^^ = ∑^^^^^^^^^^^ ^^^^^^ ^(^^ , N^) , where ^^ ^ ! N^ represents mean blood pressure and mean flowrate at the ventricular outlets over a complete cardiac cycle, respectively. Thus, for each reference patient, the mean ventricular pressure obtained at block 614 can be used with a mean flow rate measured from each reference patient to determine the total energy output for the reference patient.

[0134] At block 618, the computing device may associate, for each reference patient, the total energy output and total energy input to at least one feature vector representing non-invasive cardiac parameters. For example, the computing device may receive, for each reference patient, non-invasive cardiac parameters of the reference patient, such as but not limited to aortic and / or limb systolic and diastolic blood pressures, EDV, ESV, and heart rate of the reference patient. Such non-invasive cardiac parameters of the reference patient may be vectorized (e.g., quantified) and arranged as a feature vector.

[0135] At block 620, the computing device may train, based on the associated feature vectors, one or more linear regression models to predict the total energy output and the total energy input for the patient using the non-invasive cardiac parameters of the patient. The training may involve supervised learning based on the association or mapping of inputs (e.g., non-invasive cardiac parameters) to outputs (e.g., ventricular pressures). The feature vector may serve as input variables in a machine learning model (e.g., a set of linear regression models) to be associated with the known or otherwise labeled set of outputs – ventricular pressures, flow rates, the total energy inputs, and / or the total energy output - for supervised learning. The result of the supervised learning may be a set of weights (e.g., coefficients) that may be used to predict the ventricular pressures, flow rates, the total energy input and / or the total energy output for a patient based on the non-invasive cardiac parameters of the patient.

[0136] Thus, in at least one embodiment, the set of linear regression models for which the outputs are the ventricular pressures, total energy inputs, and / or total energy outputs, may include the following. A set of models for which the outputs are ventricular pressures, and the non-invasive cardiac parameter inputs may include systolic, diastolic, and mean blood pressure measurements of the patient:UCLA 2024-258-2 UCLA.P0211WOBy combining the above expressions, one may obtain the following additional linear regression models:

[0144] In the aforementioned expressions, with the output ventricular pressures being known from the catheterized blood pressure waveforms of the reference patients (e.g., obtained at block 612), a supervised training may yield coefficients C1-C16.

[0145] A set of models for which the outputs are the total energy inputs, and the non-invasive cardiac parameter inputs include end systole volume (ESV), end diastole volume (EDV), and non- invasive blood pressures may include:

[0146] ^^^^^^^^^^^^ ^^^^^^^^^^^ = ^A^T", ^^^^^^^^^^ ^^^^^ ^^^^^^^, ^^^^ ^^^^^^^^B

[0148] ^^^^^^^^^^^^ ^^^^^^^^^^ = ^A^U", ^^^^ ^^^^^^^ , ^^^^^^^^^^^^^^^^^ ^^^^^ ^^^^^^^B, and

[0149] ^^^^^^^^^^^ ^^^^^^^ = ^A^^^^^^^^^^^ ^^^^^^^, ^T", ^U"B

[0150] In the aforementioned set of models, ^^^^^^^^^^^^ ^^^^^^^^^^^, ^^^^^^^^^^ ^^^^^^^^,^^^^^^^^^^^^ ^^^^^^^^^^, and ^^^^^^^^^^^ ^^^^^^^ may represent different components of the total energyinput representing different phases of the cardiac cycle (e.g., isovolumic contraction, systolic ejection, isovolumic relaxation, and diastolic filling). The non-invasive cardiac parameters that may be input to these models may include the EDV, ESV, and non-invasive blood pressures. As the ventricularUCLA 2024-258-2 UCLA.P0211WO pressures are already known and / or otherwise labeled (e.g., from the catheterized blood pressure waveforms) they may also be entered into the model. The outputs for the aforementioned set of models are the components of the total energy input.

[0151] The total energy output over time (^^^C^^) may be represented as ^^^C^^= ∑^^^^^^^^^^^ ^^^^^^ ^(^^ , N^) , where ^^ ^ ! N^ represents mean blood pressure and mean flow rate atthe ventricular outlets over a complete cardiac cycle, respectively. The mean blood pressure and mean blood flow rate at the ventricular outlet(s), for each respective reference patient, may also be determined using the blood pressure waveforms of the reference patient, as previously discussed in relation to FIG.4. Example Computer Implemented System

[0152] In various embodiments, the systems and methods for non-invasively assessing cardiac efficiency can be implemented via computer software or hardware.

[0153] FIG.7 is a block diagram illustrating a computer system 700 upon which embodiments of the present teachings may be implemented. In various embodiments of the present teachings, computer system 700 can include a bus 702 or other communication mechanism for communicating information and a processor 704 coupled with bus 702 for processing information. In various embodiments, computer system 300 can also include a memory, which can be a random-access memory (RAM) 706 or other dynamic storage device, coupled to bus 702 for determining instructions to be executed by processor 704. Memory can also be used for storing temporary variables or other intermediate information during execution of instructions to be executed by processor 704. In various embodiments, computer system 700 can further include a read only memory (ROM) 708 or other static storage device coupled to bus 702 for storing static information and instructions for processor 704. A storage device 710, such as a magnetic disk or optical disk, can be provided and coupled to bus 702 for storing information and instructions.

[0154] In various embodiments, computer system 700 can be coupled via bus 702 to a display 712, such as a cathode ray tube (CRT) or liquid crystal display (LCD), for displaying information to a computer user. In some embodiments, the display 712 may enable user input (e.g., via a touchscreen). For example, the display 712 may enable a user (e.g., a medical personnel) to enter patient-specific vitals data (e.g., a peripheral non-invasive blood pressure (NIBP) reading, a heart rate, etc.) by touching the display 712. An input device 714, including alphanumeric and other keys, can also enableUCLA 2024-258-2 UCLA.P0211WO the same and can be coupled to bus 702 for communication of information and command selections to processor 704. Another type of user input device is a cursor control 716, such as a mouse, a trackball or cursor direction keys for communicating direction information and command selections to processor 704 and for controlling cursor movement on display 712. This input device 714 typically has two degrees of freedom in two axes, a first axis (i.e., x) and a second axis (i.e., y), that allows the device to specify positions in a plane. However, it should be understood that input devices 714 allowing for 3-dimensional (x, y and z) cursor movement are also contemplated herein.

[0155] Consistent with certain implementations of the present teachings, results can be provided by computer system 700 in response to processor 704 executing one or more sequences of one or more instructions contained in memory 706. Such instructions can be read into memory 706 from another computer-readable medium or computer-readable storage medium, such as storage device 710. Execution of the sequences of instructions contained in memory 706 can cause processor 704 to perform the processes described herein. Alternatively, hard-wired circuitry can be used in place of or in combination with software instructions to implement the present teachings. Thus, implementations of the present teachings are not limited to any specific combination of hardware circuitry and software.

[0156] The term “computer-readable medium” (e.g., data store, data storage, etc.) or “computer- readable storage medium” as used herein refers to any media that participates in providing instructions to processor 704 for execution. Such a medium can take many forms, including but not limited to, non-volatile media, volatile media, and transmission media. Examples of non-volatile media can include, but are not limited to, dynamic memory, such as memory 706. Examples of transmission media can include, but are not limited to, coaxial cables, copper wire, and fiber optics, including the wires that comprise bus 702.

[0157] Common forms of computer-readable media include, for example, a floppy disk, a flexible disk, hard disk, magnetic tape, or any other magnetic medium, a CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes, a RAM, PROM, and EPROM, a FLASH-EPROM, another memory chip or cartridge, or any other tangible medium from which a computer can read.

[0158] In addition to computer-readable medium, instructions or data can be provided as signals on transmission media included in a communications apparatus or system to provide sequences of one or more instructions to processor 704 of computer system 700 for execution. For example, a communication apparatus may include a transceiver having signals indicative of instructions and data.UCLA 2024-258-2 UCLA.P0211WO The instructions and data are configured to cause one or more processors to implement the functions outlined in the disclosure herein. Representative examples of data communications transmission connections can include, but are not limited to, telephone modem connections, wide area networks (WAN), local area networks (LAN), infrared data connections, NFC connections, etc.

[0159] It should be appreciated that the methodologies described herein, flow charts, diagrams and accompanying disclosure can be implemented using computer system 700 as a standalone device or on a distributed network or shared computer processing resources such as a cloud computing network.

[0160] The methodologies described herein may be implemented by various means depending upon the application. For example, these methodologies may be implemented in hardware, firmware, software, or any combination thereof. For a hardware implementation, the processing unit may be implemented within one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, micro-controllers, microprocessors, electronic devices, other electronic units designed to perform the functions described herein, or a combination thereof.

[0161] In various embodiments, the methods of the present teachings may be implemented as firmware and / or a software program and applications written in conventional programming languages such as C, C++, Python, etc. If implemented as firmware and / or software, the embodiments described herein can be implemented on a non-transitory computer-readable medium in which a program is stored for causing a computer to perform the methods described above. It should be understood that the various engines described herein can be provided on a computer system, such as computer system 700, whereby processor 704 would execute the analyses and determinations provided by these engines, subject to instructions provided by any one of, or a combination of, memory components 706 / 708 / 710 and user input.

[0162] In describing the various embodiments, the specification may have presented a method and / or process as a particular sequence of steps. However, to the extent that the method or process does not rely on the particular order of steps set forth herein, the method or process should not be limited to the particular sequence of steps described. As one of ordinary skill in the art would appreciate, other sequences of steps may be possible. Therefore, the particular order of the steps set forth in the specification should not be construed as limitations on the claims. In addition, the claims directed to the method and / or process should not be limited to the performance of their steps in theUCLA 2024-258-2 UCLA.P0211WO order written, and one skilled in the art can readily appreciate that the sequences may be varied and still remain within the spirit and scope of the various embodiments. Similarly, any of the various system embodiments may have been presented as a group of particular components. However, these systems should not be limited to the particular set of components, now their specific configuration, communication and physical orientation with respect to each other. One skilled in the art should readily appreciate that these components can have various configurations and physical orientations (e.g., wholly separate components, units and subunits of groups of components, different communication regimes between components).

[0163] Although specific embodiments and applications of the disclosure have been described in this specification, these embodiments and applications are exemplary only, and many variations are possible.

Claims

UCLA 2024-258-2 UCLA.P0211WO What is claimed is:

1. A computer-implemented method for assessing cardiac efficiency, the method comprising: receiving, by a processor, patient-specific image data of a cardiac chamber of a patient, the patient-specific image data acquired from an imaging modality over a cardiac cycle; receiving, by the processor, patient-specific vitals data; deriving, by the processor, non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; applying, by the processor, the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generating, in real time, based on a ratio of the total energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient-specific cardiac efficiency value classifying a severity of a heart disease of the patient. 2 The method of claim 1, wherein the patient-specific vitals data comprises a heart rate, and wherein the patient-specific cardiac efficiency value is further based on the heart rate. 3 The method of claim 1 or 2, further comprising: determining, by the processor, based on the patient-specific image data, an end diastolic volume (EDV) and an end systolic volume (ESV) of the cardiac chamber, and a blood flow rate at a ventricular outlet of the cardiac chamber; wherein the non-invasive cardiac parameters applied to the trained model comprises the EDV, the ESV, and the blood flow rate. 4 The method of any one of the preceding claims, wherein the patient-specific vitals data comprises non-invasive blood pressure (NIBP) measurements and a heart rate of the patient, and wherein the non-invasive cardiac parameters applied to the trained model further comprises the NIBP measurements, wherein the patient-specific cardiac efficiency value is further based on the heart rate. 5 The method of any one of the preceding claims, wherein applying the non-invasive cardiac parameters to the trained model comprises:UCLA 2024-258-2 UCLA.P0211WO determining, based on trained model, a ventricular pressure at each of a plurality of events of the cardiac cycle; and determining, based on the trained model and the ventricular pressures, the total mechanical energy input and the total mechanical energy output.

6. The method of any one of the preceding claims, wherein the trained model comprises one or more linear regression models, wherein the one or more linear regression models are trained using a plurality of blood pressure waveforms of a plurality of respective reference patients. 7 The method of claim 6, further comprising, prior to applying the one or more parameters to the one or more linear regression models: receiving the plurality of blood pressure waveforms of the plurality of respective reference patients; for each blood pressure waveform of a respective reference patient, identifying a ventricular pressure at each of a plurality of events of a cardiac cycle associated with the respective reference patient; determining, based on the ventricular pressures, a total energy output for blood flow out of a cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient and a total energy input for blood flow into the cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient; associating the total energy output of the respective reference patient and the total energy input with at least one feature vector representing non-invasive cardiac parameters of the respective reference patient; and training, by the processor, based on the at least one associated feature vector of each respective reference patient, the one or more linear regression models to predict the total mechanical energy output for blood flow out of the cardiac chamber of the patient and the total mechanical energy input for blood flow into the cardiac chamber of the patient using the non-invasive cardiac parameters of the patient.UCLA 2024-258-2 UCLA.P0211WO 8. The method of claim 7, wherein determining the total energy output and the total energy input of the respective reference patient comprises determining, based on the ventricular pressures, one or more components of a total mechanical energy input into the cardiac chamber of the respective reference patient, wherein associating the total energy output and the total energy input with the at least one feature vector representing the non-invasive cardiac parameters of the respective reference patient comprises associating each component of the total mechanical energy input with a respective feature vector based on non-invasive cardiac parameters of the respective reference patient. 9 The method of claim 8, wherein each component of the total mechanical energy input comprises a mechanical energy distribution for a respective phase of the cardiac cycle, wherein the respective phase is among one or more phases of the cardiac cycle corresponding to the one or more components of the total mechanical energy input. 10 The method of claim 9, wherein the one or more phases corresponding to the one or more components of the total mechanical energy input comprises: an isovolumic contraction; a systolic ejection; an isovolumic relaxation; and a diastolic filling. 11 The method of any one of claims 7-10, further comprising: determining, by the processor, an average ventricular pressure for each of a systole phase and a diastole phase of the cardiac cycle, wherein the one or more components of the total mechanical energy input into the cardiac chamber is further based on the average ventricular pressures. 12 The method of any one of claims 7-11, wherein the plurality of events of the cardiac cycle comprises: an end diastole or a closing of an atrioventricular (AV) valve; an opening of a semilunar valve; an end systole or a closing of the semilunar valve; and an opening of the AV valve.UCLA 2024-258-2 UCLA.P0211WO 13. The method of any one of the preceding claims, wherein the non-invasive cardiac parameters comprise: diastolic, systolic, and mean blood pressures; an EDV and an ESV of the cardiac chamber; and a blood flow rate at a ventricular outlet of the cardiac chamber.

14. The method of any one of the preceding claims, further comprising: adjusting, by the processor, the patient-specific cardiac efficiency value based on one or more of: a basal metabolism of the cardiac chamber, an electrical excitation of the cardiac chamber, or an efficiency of mitochondrial oxidative phosphorylation in the cardiac chamber.

15. The method of any one of the preceding claims, wherein the heart disease is congenital heart disease (CHD).

16. The method of any one of the preceding claims, wherein the severity of the heart disease includes an absence of the heart disease in the patient.

17. The method of any one of the preceding claims, further comprising: generating a treatment recommendation based on the classified severity of the heart disease, the treatment recommendation comprising one or more of: a medication regimen based on a medication type, a dosage, and an administration timeline; or a surgical intervention.

18. The method of claim 17, wherein the treatment recommendation is further based on a mechanical energy distribution of the total energy input among phases of the cardiac cycle, the phases comprising: an isovolumic contraction, a systolic ejection, an isovolumic relaxation, and a diastolic filling.

19. The method of any one of the preceding claims, wherein the cardiac chamber comprises one or more ventricles.UCLA 2024-258-2 UCLA.P0211WO 20. A system for non-invasively assessing cardiac efficiency, the system comprising: a processor; and memory storing instructions that, when executed by the processor, cause the processor to: receive patient-specific image data of a cardiac chamber of a patient, the patient- specific image data acquired from an imaging modality over a cardiac cycle; receive patient-specific vitals data; derive non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; apply the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generate, in real time, based on a ratio of the total energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient-specific cardiac efficiency value classifying a severity of a heart disease of the patient.

21. The system of claim 20, wherein the patient-specific vitals data comprises a heart rate, and wherein the patient-specific cardiac efficiency value is further based on the heart rate.

22. The system of claim 20 or 21, wherein the instructions, when executed, further cause the processor to: determine, based on the patient-specific image data, an end diastolic volume (EDV) and an end-systolic volume (ESV) of the cardiac chamber, and a blood flow rate at a ventricular outlet of the cardiac chamber; wherein the non-invasive cardiac parameters applied to the trained model comprises the EDV, the ESV, and the blood flow rate.

23. The system of any one of claims 20-22, wherein the patient-specific vitals data comprises non- invasive blood pressure (NIBP) measurements and heart rate of the patient, and wherein the non- invasive cardiac parameters applied to the trained model further comprises the NIBP measurements, wherein the patient-specific cardiac efficiency value is further based on the heart rate.UCLA 2024-258-2 UCLA.P0211WO 24. The system of any one of claims 20-23, wherein the instructions, when executed, cause the processor to apply the non-invasive cardiac parameters to the trained model by: determining, based on trained model, a ventricular pressure at each of a plurality of events of the cardiac cycle; and determining, based on the trained model and the ventricular pressures, the total mechanical energy input and the total mechanical energy output.

25. The system of any one of claims 20-24, wherein the trained model comprises one or more linear regression models, wherein the one or more linear regression models are trained using a plurality of blood pressure waveforms of a plurality of respective reference patients.

26. The system of claim 25, wherein the instructions, when executed, cause the processor to, prior to applying the one or more parameters to the one or more linear regression models: receive the plurality of blood pressure waveforms of the plurality of respective reference patients; for each blood pressure waveform of a respective reference patient, identify, a ventricular pressure at each of a plurality of events of a cardiac cycle associated with the respective reference patient; determine, based on the ventricular pressures, a total energy output for blood flow out of a cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient and a total energy input for blood flow into the cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient; and associate the total energy output of the respective reference patient and the total energy input with at least one feature vector representing non-invasive cardiac parameters of the respective reference patient; and train, based on the at least one associated feature vector of each respective reference patient, the one or more linear regression models to predict the total mechanical energy output for blood flow out of the cardiac chamber of the patient and the total mechanical energy input for blood flow into the cardiac chamber of the patient using the non-invasive cardiac parameters of the patient.UCLA 2024-258-2 UCLA.P0211WO 27. The system of claim 26, wherein the instructions, when executed, cause the processor to determine the total energy output and the total energy input of the respective reference patient by determining, based on the ventricular pressures, one or more components of a total mechanical energy input into the cardiac chamber of the respective reference patient, wherein the instructions, when executed, cause the processor to associate the total energy output and the total energy input with the at least one feature vector representing the non-invasive cardiac parameters of the respective reference patient by associating each component of the total mechanical energy input with a respective feature vector based on non- invasive cardiac parameters of the respective reference patient.

28. The system of claim 27, wherein each component of the total mechanical energy input comprises a mechanical energy distribution for a respective phase of the cardiac cycle, wherein the respective phase is among one or more phases of the cardiac cycle corresponding to the one or more components of the total mechanical energy input.

29. The system of claim 28, wherein the one or more phases corresponding to the one or more components of the total mechanical energy input comprises: an isovolumic contraction; a systolic ejection; an isovolumic relaxation; and a diastolic filling.

30. The system of any one of claims 26-29, wherein the instructions, when executed, cause the processor to: determine an average ventricular pressure for each of a systole phase and a diastole phase of the cardiac cycle for each respective reference patient, wherein the total energy input into the cardiac chamber of the respective reference patient is further based on the average ventricular pressures.

31. The system of any one of claims 26-30, wherein the plurality of events of the cardiac cycle comprises: an end diastole or a closing of an atrioventricular (AV) valve; an opening of a semilunar valve;UCLA 2024-258-2 UCLA.P0211WO an end systole or a closing of the semilunar valve; and an opening of the AV valve.

32. The system of any one of claims 20-30, wherein the non-invasive cardiac parameters comprise: diastolic, systolic and mean blood pressures; an EDV and an ESV of the cardiac chamber, and a blood flow rate at a ventricular outlet of the cardiac chamber.

33. The system of any one of claims 20-32, wherein the instructions, when executed, cause the processor to: adjust the patient-specific cardiac efficiency value based on one or more of: a basal metabolism of the cardiac chamber, an electrical excitation of the cardiac chamber, or an efficiency of mitochondrial oxidative phosphorylation in the cardiac chamber.

34. The system of any one of claims 20-33, wherein the heart disease is congenital heart disease (CHD).

35. The system of any one of claims 20-34, wherein the severity of the heart disease includes an absence of the heart disease in the patient.

36. The system of any one of claims 20-35, wherein the instructions, when executed, cause the processor to: generate a treatment recommendation based on the classified severity of the heart disease, the treatment recommendation comprising one or more of: a medication regimen based on a medication type, a dosage, and an administration timeline; or a surgical intervention.

37. The system of claim 36, wherein the treatment recommendation is further based on a mechanical energy distribution of the total energy input among phases of the cardiac cycle, the phases comprising: an isovolumic contraction, a systolic ejection, an isovolumic relaxation, and a diastolic filling.UCLA 2024-258-2 UCLA.P0211WO 38. The system of any one of claims 20-37, wherein the cardiac chamber comprises one or more ventricles.

39. A non-transitory computer-readable medium (CRM) having stored thereon computer-readable instructions executable to cause performance of operations comprising: receiving, by a processor, patient-specific image data of a cardiac chamber of a patient, the patient-specific image data acquired from an imaging modality over a cardiac cycle; receiving, by the processor, patient-specific vitals data; deriving, by the processor, non-invasive cardiac parameters from the patient-specific image data and the patient-specific vitals data; applying, by the processor, the non-invasive cardiac parameters to a trained model to generate: a total energy output for blood flow out of the cardiac chamber during the cardiac cycle, and a total energy input for blood flow into the cardiac chamber during the cardiac cycle; and generating, in real time, based on a ratio of the total energy output to the total energy input, a patient-specific cardiac efficiency value for the cardiac chamber, the patient-specific cardiac efficiency value classifying a severity of a heart disease of the patient.

40. The non-transitory CRM of claim 39, wherein the patient-specific vitals data comprises a heart rate, and wherein the patient-specific cardiac efficiency value is further based on the heart rate.

41. The non-transitory CRM of claim 39 or 40, the operations further comprising: determining, by the processor, based on the patient-specific image data, an end diastolic volume (EDV) and an end systolic volume (ESV) of the cardiac chamber, and a blood flow rate at a ventricular outlet of the cardiac chamber; wherein the non-invasive cardiac parameters applied to the trained model comprises the EDV, the ESV, and the blood flow rate.

42. The non-transitory CRM of any one of claims 39-41, wherein the patient-specific vitals data comprises non-invasive blood pressure (NIBP) measurements of the patient and a heart rate, andUCLA 2024-258-2 UCLA.P0211WO wherein the non-invasive cardiac parameters applied to the trained model further comprises the NIBP measurements, wherein the patient-specific cardiac efficiency value is further based on the heart rate.

43. The non-transitory CRM of any one of claims 39-42, wherein applying the non-invasive cardiac parameters to the trained model comprises: determining, based on trained model, a ventricular pressure at each of a plurality of events of the cardiac cycle; and determining, based on the trained model and the ventricular pressures, the total mechanical energy input and the total mechanical energy output.

44. The non-transitory CRM of any one of claims 39-43, wherein the trained model comprises one or more linear regression models, wherein the one or more linear regression models are trained using a plurality of blood pressure waveforms of a plurality of respective reference patients.

45. The non-transitory CRM of claim 44, the operations further comprising, prior to applying the one or more parameters to the one or more linear regression models: receiving the plurality of blood pressure waveforms of the plurality of respective reference patients; for each blood pressure waveform of a respective reference patient, identifying, a ventricular pressure at each of a plurality of events of a cardiac cycle associated with the respective reference patient; determining, based on the ventricular pressures, a total energy output for blood flow out of a cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient and a total energy input for blood flow into the cardiac chamber of the respective reference patient during the cardiac cycle associated with the respective reference patient; and associating the total energy output of the respective reference patient and the total energy input with at least one feature vector representing non-invasive cardiac parameters of the respective reference patient; and training, based on the at least one associated feature vector of each respective reference patient, the one or more linear regression models to predict the total mechanical energy output for blood flow out of the cardiac chamber of the patient and the total mechanical energy input for blood flow into the cardiac chamber of the patient using the non-invasive cardiac parameters of the patient.UCLA 2024-258-2 UCLA.P0211WO 46. The non-transitory CRM of claim 45, wherein determining the total energy output and the total energy input of the respective reference patient comprises determining, based on the ventricular pressures, one or more components of a total mechanical energy input into the cardiac chamber of the respective reference patient, wherein associating the total energy output and the total energy input with the at least one feature vector representing the non-invasive cardiac parameters of the respective reference patient comprises associating each component of the total mechanical energy input with a respective feature vector based on non-invasive cardiac parameters of the respective reference patient; 47. The non-transitory CRM of claim 46, wherein each component of the total mechanical energy input comprises a mechanical energy distribution for a respective phase of the cardiac cycle, wherein the respective phase is among one or more phases of the cardiac cycle corresponding to the one or more components of the total mechanical energy input.

48. The non-transitory CRM of claim 47, wherein the one or more phases corresponding to the one or more components of the total mechanical energy input comprises: an isovolumic contraction; a systolic ejection; an isovolumic relaxation; and a diastolic filling.

49. The non-transitory CRM of any one of claims 45-48, the operations further comprising: determining, by the processor, an average ventricular pressure for each of a systole phase and a diastole phase of the cardiac cycle, wherein the one or more components of the total mechanical energy input into the cardiac chamber is further based on the average ventricular pressures.

50. The non-transitory CRM of any one of claims 45-49, wherein the plurality of events of the cardiac cycle comprises: an end diastole or a closing of an atrioventricular (AV) valve; an opening of a semilunar valve; an end systole or a closing of the semilunar valve; and an opening of the AV valve.UCLA 2024-258-2 UCLA.P0211WO 51. The non-transitory CRM of any one of claims 39-50, wherein the non-invasive cardiac parameters comprise: diastolic, systolic and mean blood pressures; an EDV and an ESV of the cardiac chamber, a blood flow rate at a ventricular outlet of the cardiac chamber.

52. The non-transitory CRM of any one of claims 39-51, the operations further comprising: adjusting, by the processor, the patient-specific cardiac efficiency value based on one or more of: a basal metabolism of the cardiac chamber, an electrical excitation of the cardiac chamber, or an efficiency of mitochondrial oxidative phosphorylation of the cardiac chamber.

53. The non-transitory CRM of any one of claims 39-52, wherein the heart disease is congenital heart disease (CHD).

54. The non-transitory CRM of any one of claims 39-53, wherein the severity of the heart disease includes an absence of the heart disease in the patient.

55. The non-transitory CRM of any one of claims 39-54, the operations further comprising: generating a treatment recommendation based on the classified severity of the heart disease, the treatment recommendation comprising one or more of: a medication regimen based on a medication type, a dosage, and an administration timeline; or a surgical intervention.

56. The non-transitory CRM of any claim 55, wherein the treatment recommendation is further based on a mechanical energy distribution of the total energy input among phases of the cardiac cycle, the phases comprising: an isovolumic contraction, a systolic ejection, an isovolumic relaxation, and a diastolic filling.

57. The non-transitory CRM of any one of claims 39-56, wherein the cardiac chamber comprises one or more ventricles.

Images