Hemodynamic parameter estimation

Abstract

An apparatus and method for estimating one or more hemodynamic parameters (e.g., cardiac output or stroke volume). Embodiments are based on the idea of incorporating information about vascular tone into hemodynamic parameter estimation to improve accuracy. More specifically, embodiments use measurements of the duration of travel of a blood pulse from the heart along a particular length of an arterial path as a proxy measure for vascular tone and incorporate it into hemodynamic parameter estimation. Embodiments also account for changes in vascular tone between different parts of the circulatory system based on incorporating vascular tone proxy measurements for multiple different arterial paths.

Description

Technical Field

[0001] This invention relates to apparatus and methods for estimating one or more hemodynamic parameters. Background Technology

[0002] To assess hemodynamic stability, various hemodynamic parameters are important for measurement and monitoring. These parameters include, for example, central hemodynamic parameters such as cardiac output (CO), stroke volume (SV), and their changes over time.

[0003] Cardiac output (CO) is defined as the amount of blood pumped from the heart per minute [liters / minute], and is therefore affected by two factors: heart rate (HR) [heart beats / minute] and stroke volume (SV), where stroke volume is the amount of blood pumped from the left ventricle of the heart to the aorta with each heartbeat [liters / heartbeat].

[0004] These central hemodynamic parameters can be measured invasively or non-invasively. Typically, this involves obtaining a measure of blood flow in one or more arteries. Blood flow Q is defined as the volume of blood flowing through a vessel per unit time. Mathematically, it can be expressed as Q = vA, where Q is the blood flow at a location along the arterial path, v is the blood velocity at that location (cm / s), and A (ml / s) is the cross-sectional area of ​​the vessel at that location (cm²). 2 ).

[0005] Various non-invasive measurement methods exist, including, for example, using an ultrasound sensing unit (e.g., using Doppler ultrasound) to measure blood flow through large arteries. Other methods for measuring blood flow are also known, such as using blood pressure measurements to indirectly determine blood flow.

[0006] like Figure 1 As illustrated in the schematic diagram, in known methods, blood flow is typically measured in one of the branches 14a-14n of the vascular tree (in an artery), and the blood flow measurement result can then be used to estimate one or more hemodynamic parameters. This can be accomplished, for example, by using a predetermined algorithm or transfer function that takes the blood flow measurement result as input and provides hemodynamic parameter estimates as output. Figure 1 The diagram schematically illustrates the blood flow from the heart 12 through the various arterial branches 14a-14n of the circulatory system. The volumetric blood flow output from the heart in each heartbeat is called stroke volume (SV), and the volumetric output per minute is called cardiac output (CO).

[0007] This method is simple and fast, but it is well-known to lack accuracy. Obtaining blood flow measurements from only a single branch of the arterial system may not accurately estimate hemodynamic parameters because flow velocities may differ between arterial branches. Furthermore, known estimation methods cannot account for differences in autoregulation and vasomotor characteristics present in different arterial branches.

[0008] Therefore, improved methods for obtaining non-invasive estimation results of hemodynamic parameters are often valuable in this field. Summary of the Invention

[0009] This invention is defined by the claims.

[0010] According to an example of one aspect of the invention, a controller is provided suitable for determining the estimation result of at least one hemodynamic parameter of an object, said controller being adapted to:

[0011] Receive input indicating arterial flow measurement results in at least one arterial path of the object, wherein the arterial path is a central arterial path or a peripheral arterial path;

[0012] A measure of the time difference ΔT between the indicated cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along the respective arterial path is obtained for each of the central and peripheral arterial pathways, as a result of blood flow from the heart to the location;

[0013] The estimation results of the hemodynamic parameters are determined based on a combination of the following: blood flow measurements from the at least one arterial pathway, and the time difference measure ΔT for each of the central arterial pathway and the peripheral arterial pathway.

[0014] Embodiments of the present invention are based on the understanding that known methods for estimating hemodynamic parameters (e.g., cardiac output or stroke volume) can be improved by taking into account vascular tension in different branches of the arterial system. Vascular tension refers to the degree of contraction a blood vessel undergoes relative to its maximum dilation state at any given time. All arteries and veins (under normal conditions) exhibit a certain degree of smooth muscle contraction, which determines the diameter of the vessel and thus its vascular tension. Vascular tension in any given local area can vary at any given time depending on hemodynamic conditions and is modulated by the influence of competing vasoconstrictors and vasodilators. The effect of changes in vascular tension is dual: it regulates both systemic arterial blood pressure and local blood flow within an organ. Therefore, vascular tension in an arterial path at any given time affects measurable hemodynamic parameters, such as cardiac output and stroke volume. Furthermore, since vascular tension varies in different arterial paths, accurate estimation of hemodynamic parameters should take this variation into account.

[0015] It is difficult to directly measure vascular tension. Embodiments of the present invention are based on the understanding that the time it takes for blood to travel a predetermined length from the heart along an arterial path depends on the vascular tension of that path. These two are interrelated. In view of this, the inventors recognized that this duration can be used as a proxy measure of vascular tension for a given arterial path.

[0016] Two standard clinical measurements provide an indication of this duration: pulse arrival time (PAT) and pulse conduction time (PTT), and certain embodiments of the invention utilize these clinical measurements to provide ΔT measurements. However, using these standard measurements is not essential for deriving ΔT measurements.

[0017] Therefore, in summary, embodiments of the present invention are based on improving the accuracy of determining central hemodynamic parameters by integrating information about vascular tension in different arterial pathways within the circulatory system. A surrogate measure of vascular tension is estimated by calculating the time taken for blood to travel along the length of one or more arterial pathways (e.g., pulse arrival time (PAT) or pulse conduction time (PTT)). This can then be integrated into an algorithm, computation, or transfer function in combination with arterial blood flow measurements for at least one arterial pathway (preferably the central arterial pathway) to determine one or more hemodynamic parameters.

[0018] This allows the calculation to be sensitive to differential autoregulation (including dynamic changes in vascular tone) across different arterial pathways. This makes the estimation of central hemodynamic parameters more accurate. Accurate estimation of central hemodynamic parameters allows for improved clinical decision-making and medical interventions, thereby improving patient outcomes.

[0019] Blood flow measurement is a measurement of the volumetric blood flow per unit time at a measurement location within an arterial path. In other words, it is a measurement of the volume of blood flowing through an arterial path (e.g., through a measurement location within the arterial path) per unit time. Mathematically, it can be expressed as Q = vA, where Q is the blood flow rate at the location along the arterial path, v is the blood flow velocity at that location (cm / s), and A (ml / s) is the cross-sectional area of ​​the blood vessel at that location (cm²). 2 ).

[0020] The controller may be adapted to provide the blood flow measurement results for the at least one arterial path and the metric or parameter derived from the metric indicating the time difference ΔT between the central arterial path and the peripheral arterial path as input to a predetermined transfer function, wherein the transfer function is adapted to generate the estimation results for the at least one hemodynamic parameter based on the input.

[0021] In other words, the controller is adapted to use a transfer function to process blood flow measurements, ΔT, or parameters derived from said ΔT for at least one arterial pathway to generate an estimate for at least one hemodynamic parameter.

[0022] A transfer function can define or embody a predetermined functional relationship between an input and at least one hemodynamic parameter. The transfer function can generate hemodynamic parameters based on this predetermined functional relationship and the input.

[0023] In some examples, the transfer function can be a classical algorithmic function and / or a machine learning model. It can include a linear input function. In some examples, it can be a multi-parameter linear regression model that models the target hemodynamic parameter as a linear sum of each input parameter, weighted by its respective weighting coefficient.

[0024] The controller is preferably also adapted to generate a data output indicating an estimation result of at least one hemodynamic parameter. This may be a data packet comprising data representing the estimation result or a series of values ​​derived therefrom over a period of time, prepared for data output (e.g., output to a data storage unit or user interface) or for data output along a network communication channel. The method may include transmitting the data output to another module or device (e.g., a user interface).

[0025] In an advantageous embodiment, ultrasound monitoring technology can be used to detect blood flow and / or the ΔT parameter. For example, Doppler ultrasound can be used to acquire blood flow measurement results.

[0026] To determine at least one hemodynamic parameter, one can combine the input of at least one blood flow measurement with the input of a ΔT measurement using a predefined function, equation, or algorithm. Machine learning engines can also be used to derive hemodynamic parameters from the input. These different methods will be collectively referred to as transfer functions in this paper.

[0027] For example, it is possible to use a pre-defined or pre-stored function (transfer function) that embodies a predetermined functional relationship between inputs, and to compute at least one hemodynamic parameter based on the inputs. A transfer function can simply be a mathematical relationship between several input parameters and a central hemodynamic parameter that serves as the output.

[0028] To derive such a transfer function, machine learning and / or statistical methods can be applied, for example, to labeled datasets. For instance, multiparameter regression has been successfully used in trials to estimate such a transfer function using clinical datasets. Given sufficient data, patient metadata (e.g., gender, BMI, and other personal patient information) can also be used to improve the transfer function.

[0029] Arterial pathway refers to a longitudinal section of a certain length, for example, along one or more arteries in the circulatory system.

[0030] The peripheral arterial pathway refers to the arterial pathway in the peripheral vascular system, that is, the part of the circulatory system that consists of arteries not located in the head, chest, or abdomen (e.g., arteries in the arms, hands, legs, and feet).

[0031] The central arterial pathway refers to the pathway in the central vascular system (e.g., arteries in the head, chest, abdomen, or neck).

[0032] Preferably, blood flow measurements can be derived from at least a central arterial path (e.g., the carotid artery). However, blood flow measurements can also be derived from different arterial paths, including peripheral arterial paths. In some examples, blood flow measurements can be derived from both peripheral and central arterial paths. Preferably, the at least one arterial path from which blood flow measurements are derived is one of the same two arterial paths (central and peripheral arterial paths) from which ΔT measurements are derived.

[0033] A cardiac ejection event refers to the event corresponding to the ejection of blood from the heart, i.e., a heartbeat event, or the systolic phase of the heart. This event can be a defined reference point (e.g., the end of the pre-ejection phase) during the ejection process (e.g., during the systolic phase of the cardiac cycle) when the aorta opens and blood begins to be ejected from the left ventricle. In other examples, the event could be the beginning of the pre-ejection phase, when the heart is first electrically activated and begins to contract.

[0034] Obtaining a metric ΔT indicating the time difference may include receiving an input at the controller indicating the time difference for each of the two arterial paths. In other examples, it may include processing or computation steps. For example, the controller may be adapted to: receive signal inputs indicating that a pulse wave has been detected arriving at a predetermined location along the arterial path for each arterial path and inputs indicating that a bleeding event has been detected, and to determine the duration ΔT for each arterial path.

[0035] While two arterial pathways have been mentioned in the description above and in the examples herein, in other embodiments, the ΔT time difference can be determined for more than two arterial pathways. Blood flow measurements can be obtained for one or even two or more arterial pathways. The controller can then be adapted to determine at least one hemodynamic parameter based on a combination of the ΔT value and one or more blood flow measurements from all arterial pathways.

[0036] Determining hemodynamic parameters can be based on the use of a pre-defined transfer function or algorithm. A transfer function is defined as a function that takes at least one blood flow measurement and a ΔT measurement for two or more arterial pathways as input and calculates an estimate of at least one hemodynamic parameter as output. A transfer function can simply be a mathematical relationship between several input parameters and a central hemodynamic parameter that serves as the output. This has been discussed above and will be discussed further below.

[0037] The transfer function may include machine learning algorithms. Various implementation options in this regard will be outlined in more detail below.

[0038] The transfer function can be stored locally in memory, the controller may include memory, or the controller may be operationally coupled to memory. In some examples, the transfer function can be incorporated into the programming of the controller itself.

[0039] According to a set of examples, obtaining a time difference metric for each arterial path can include obtaining a pulse arrival time (PAT) measurement for each arterial path.

[0040] In this context, the ejection event can correspond to the electrical activation point of the heart. It corresponds to the beginning of the pre-ejection phase, that is, the point at which the heart pulse first begins.

[0041] For example, this can be detected as the occurrence time of the QRS complex in an ECG signal. It can correspond to the occurrence time of a specific reference point within the QRS complex (e.g., one of the Q peak, R peak, and S peak (e.g., the R peak, which is the largest peak)). It can also correspond to the time at which the QRS complex begins.

[0042] However, using ECG to detect the event is not essential, and other units (such as accelerometers, inductive sensing units, or radar sensing units) can be used instead.

[0043] In some methods, PAT can be measured using ECG (electrocardiogram) and PPG (photoplethysmography) sensor measurements. ECG can be used to detect the onset of the QRS complex as the beginning of electrical activation of the heart, and PPG can be used to detect the onset of the pulse wave downstream along the arterial path via optical measurement of blood volume. Both measurements are standard in clinical practice.

[0044] According to one or more embodiments, obtaining the time difference measure ΔT for each arterial path includes obtaining a pulse conduction time (PTT) measurement for each arterial path. Pulse conduction time is the time between blood ejection into the aorta and the arrival of the corresponding pulse wave at the downstream measurement location. In other words, it is the time between the end of the pre-ejection phase (PEP) and the arrival of the pulse wave along the arterial path at the predetermined location.

[0045] In some embodiments, obtaining the PTT measurement result for each arterial path may include: obtaining the PAT measurement result for each arterial path, obtaining an estimate of the pre-ejection period (PEP) duration, and determining the PTT measurement result for each arterial path by subtracting the PEP duration from the PAT measurement result for each arterial path.

[0046] Pre-ejection phase (PEP) is a term in the art that refers to the period from the electrical activation of the heart (e.g., indicated by the QRS complex in an ECG signal) to the ejection of blood from the heart into the aorta. This is the time between cardiac electrical activation and the opening of the aortic valve.

[0047] Therefore, the relationship between pulse arrival time (PAT) and PEP is as follows: PAT = PEP + PTT, where PTT is the pulse conduction time.

[0048] There are different ways to obtain PEP estimates.

[0049] According to a set of examples, estimates of the duration of PEP can be derived based on inputs from a phonocardiography (PCG) sensing unit and / or an impedance cardiography (ICG) measurement unit. For example, both aortic valve opening and ejection produce distinctive high-pitched sounds: a click when opening and a click when ejecting. One or both of these sounds can be identified in PCG reads that occur after the first heart sound. This can, for example, be used to detect the timing of the ejection event.

[0050] Furthermore, ICG and ECG can also be used together, either additionally or alternatively, to directly identify PEP. The method for achieving this is detailed in the article "Estimated preejection period (PEP) based on the detection of the R-wave and dZ / dt-min peaks in ECG and ICG" by René van Lien, Nienke MSchutte, Jan H Meijer, and Eco JC de Geus (April 18, 2013, IOP Publishing Ltd.).

[0051] Measurements obtained from these sensing modalities can be used to identify the moment of aortic valve opening. This allows determination of the time between cardiac electrical activation (e.g., the onset of the QRS complex in an ECG measurement or other reference point) and aortic valve opening (the point at which the pressure wave begins to travel along the arterial system). This period corresponds to the pre-ejection phase.

[0052] In other examples, the controller may use a predetermined estimate (i.e., a reference value) for the pre-ejaculation phase, for example, by storing the predetermined estimate in local memory or retrieving it from a remote data source (e.g., a remote server).

[0053] According to one or more embodiments, hemodynamic parameters can be determined based on the change of time difference ΔT for each arterial path over time. For this set of embodiments, the controller can be configured to: obtain multiple time difference measurements ΔT corresponding to different cardiac cycles for each arterial path in the arterial path, determine a metric indicating the change of the ΔT value over time for each arterial path, and determine the estimated hemodynamic parameters based on the change of the ΔT value.

[0054] According to one or more embodiments, determining the estimation result of the hemodynamic parameters includes determining a quotient of one or more ΔT values ​​for the peripheral and central pathways. In other words, the estimation result of the hemodynamic parameters can be based on the quotient ΔT_cen / ΔT_peri, where ΔT_cen is the time difference ΔT value for the central arterial pathway and ΔT_peri is the time difference ΔT value for the peripheral arterial pathway. In other examples, the quotient can be calculated in the reverse manner, i.e., ΔT_peri / ΔT_cen.

[0055] Changes in surrogate measures of vascular tension (ΔT value) and / or their ratios (in the case of multiple surrogate measures) indicate changes in the (relative) tension of the branches of the vascular tree being measured. These changes can indicate both local and more systemic changes in vascular tension within the arterial system. By incorporating these measurements into the calculation of hemodynamic parameters, more accurate estimates can be obtained.

[0056] The at least one hemodynamic parameter calculated by the controller may include at least one of the following: cardiac output, stroke volume, and stroke volume variability (SVV). Other examples of hemodynamic parameters that can be derived include: blood flow velocity, stroke volume variability, systolic velocity, diastolic velocity, and blood pressure.

[0057] The details of the transfer function can vary depending on the hemodynamic parameters to be derived. The transfer function reflects the mathematical relationship between the input parameters and one or more output hemodynamic parameters. Therefore, this relationship may differ for different target hemodynamic parameters, for example, due to the weights applied to different inputs in the calculation and possibly other inputs used to calculate the target hemodynamic parameters (where, for example, there are other inputs besides blood flow and ΔT measurements).

[0058] The controller can be configured to use an ultrasound sensing unit to obtain at least one metric indicating the blood flow measurement result. The ultrasound sensing unit may include one or more ultrasound transducers. It may include, for example, an ultrasound transducer unit or probe, configured to generate ultrasound pulse emissions and detect reflected ultrasound echo signals.

[0059] The ultrasound sensing unit can output ultrasound data to the controller, or it can directly output calculated blood flow measurement results to the controller. The ultrasound sensing unit can acquire Doppler ultrasound data. In some examples, the ultrasound sensing unit may also include a dedicated ultrasound processing unit for extracting one or more blood flow measurements from the acquired ultrasound data and supplying these blood flow measurements to the controller to estimate hemodynamic parameters.

[0060] As a favorable example, wearable ultrasound sensors can be used to monitor blood flow in arteries, for instance, by being placed on at least one arterial path (e.g., the carotid artery). This allows for the measurement of blood flowing through the arterial path.

[0061] It should be noted that the use of ultrasound in the context of this invention is particularly advantageous because ultrasound sensing is versatile in acquiring a range of different information and is completely non-invasive. Furthermore, according to one or more embodiments, these ultrasound-based features can be combined in various ways to provide meaningful information, such as blood flow waveforms as the product of cross-sectional area and velocity waveform.

[0062] For example, in the context of embodiments of the present invention, pulse wave Doppler ultrasound measurements can be used to obtain ultrasound-based measurements, which can be used to generate (blood flow) velocity waveforms. Useful features, such as peak contraction velocity, maximum velocity, and many other features, can be extracted from these waveforms.

[0063] Relevant information can also be exported from B-mode ultrasound data or measurement results. This can be used, for example, to export waveforms for blood vessel diameter.

[0064] However, the use of ultrasound is not essential. For example, an alternative to using ultrasound is to indirectly obtain blood flow measurement results using blood pressure measurements. This will be explained in more detail below.

[0065] According to one or more embodiments, the controller can be configured to, for at least one of the arterial pathways, use the same ultrasound sensing unit to obtain both a measure of arterial flow and a measure of the time it takes for the pulse wave to reach the predetermined location along the arterial pathway. Therefore, the same components can be used for the dual function of detecting pulse arrival and measuring blood flow. This limits the total number of components.

[0066] The controller can be configured to detect the arrival time of a pulse wave at a predetermined location along at least one arterial path. For example, a peripheral arterial path could be an arterial path extending along an object's arm, and a PPG sensor attached to the object's finger could be used to detect the arrival event of a pulse wave at the object's finger. This is merely an example, mentioned for illustrative purposes only, and is not intended to limit the invention.

[0067] According to some examples, the ejection event is the moment of electrical activation of the heart, corresponding to the beginning of the pre-ejection phase.

[0068] In some alternative examples, the ejaculation event corresponds to the end point of the pre-ejaculation phase.

[0069] The controller can be configured to use ECG sensor input to detect the occurrence of bleeding events.

[0070] For example, when the hemorrhage event corresponds to the start of the pre-ejaculation phase, detecting the occurrence of the hemorrhage event may include detecting the start time of the QRS complex in the ECG signal.

[0071] Using an ECG is just one example. Alternative methods for detecting bleeding events include using inductive sensors, radar sensors, accelerometers, or heart rate sensors (e.g., chest heart rate sensor belts).

[0072] According to one or more embodiments, an apparatus for deriving an estimate of at least one hemodynamic parameter may also be provided, comprising:

[0073] The first sensor unit is used to detect cardiac ejection events;

[0074] A second sensor unit, which is coupled to a location along the arterial path of the object, is used to detect the time it takes for a blood pulse wave to arrive at the location from the heart;

[0075] A third sensor unit is used to detect blood flow through the arterial path; and

[0076] The controller, as described in the above overview or any example or embodiment below, or according to any claim of this application, is operatively coupled to the first sensor unit, the second sensor unit, and the third sensor unit.

[0077] The first sensor unit, the second sensor unit, and the third sensor unit may correspond to different corresponding sensor devices, or one or more sensor units may correspond to the same sensor device.

[0078] In some examples, the first sensor unit may be an ECG sensor arrangement unit.

[0079] The second sensor unit may include a PPG sensor for optical coupling to the location.

[0080] The third sensor unit can be an ultrasonic sensing unit.

[0081] Therefore, at least one set of embodiments may include means for deriving an estimate of at least one hemodynamic parameter, including:

[0082] ECG sensor array unit, used to detect cardiac ejection events;

[0083] At least one PPG sensor is used to optically couple to a location along an arterial path of an object to detect the time it takes for a blood pulse wave to arrive at said location from the heart;

[0084] An ultrasound sensing unit for detecting blood flow through at least one arterial path; and

[0085] The controller, based on any example or embodiment outlined above or described below, or according to any claim of this application, is operatively coupled to an ECG sensor arrangement unit, at least one PPG sensor, and an ultrasonic sensing unit.

[0086] As discussed above, there are alternative examples for these sensor units, which will be discussed further below.

[0087] The ultrasound sensing unit may include an ultrasound transducer arrangement unit. It can be configured to detect blood flow using Doppler ultrasound measurements. It may include a dedicated ultrasound processing unit for extracting one or more blood flow measurements from the acquired ultrasound data and supplying these measurements to a controller for use in estimating hemodynamic parameters.

[0088] According to another aspect of the invention, a computer-implemented method is provided for determining an estimate of at least one hemodynamic parameter, the method comprising:

[0089] Receive input indicating arterial blood flow measurement results for at least one arterial path of the object, wherein the arterial path is a central arterial path or a peripheral arterial path;

[0090] A measure of the time difference ΔT between the indicated cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along the respective arterial path is obtained for each of the central and peripheral arterial pathways, as a result of blood flow from the heart to the location;

[0091] The estimated hemodynamic parameters are determined based on a combination of the blood flow measurement results for the at least one arterial pathway and the time difference ΔT measure for each arterial pathway.

[0092] The determination may, for example, include providing the blood flow measurement results for the at least one arterial path and the metric or parameter derived from the metric indicating the time difference ΔT between the central arterial path and the peripheral arterial path, as input to a predetermined transfer function, wherein the transfer function is adapted to generate the estimate for the at least one hemodynamic parameter based on the input and a predetermined functional relationship between the input and the at least one hemodynamic parameter.

[0093] The method preferably further includes generating data output indicating the estimation result of the at least one hemodynamic parameter.

[0094] The present invention also provides a computer program product comprising a computer-readable medium having computer-readable code embodied therein, the computer-readable code being configured to cause, when run by a suitable computer or processor, the computer or processor to perform any of the examples or embodiments outlined above or described below, or the method described according to any claim of this application.

[0095] These and other aspects of the invention will become apparent and elucidated with reference to one or more embodiments described below. Attached Figure Description

[0096] To better understand the invention and to more clearly illustrate how it can be practiced, reference will now be made to the accompanying drawings by way of example only, in which:

[0097] Figure 1 The diagram schematically illustrates blood flow through different branches of the arterial system;

[0098] Figure 2 The overall concept of the invention is illustrated schematically;

[0099] Figure 3 The processing workflow of an example controller according to one or more embodiments is schematically illustrated;

[0100] Figure 4 The processing workflow according to another example embodiment is illustrated schematically;

[0101] Figure 5 The processing workflow according to another example embodiment is illustrated schematically;

[0102] Figure 6 The processing workflow according to another example embodiment is illustrated schematically;

[0103] Figure 7 An example apparatus according to an example embodiment is schematically depicted; and

[0104] Figure 8 An example method according to one or more embodiments is outlined in the form of a block diagram. Detailed Implementation

[0105] The invention will be described with reference to the accompanying drawings.

[0106] It should be understood that while the detailed description and specific examples indicate exemplary embodiments of the apparatus, system, and method, they are for illustrative purposes only and are not intended to limit the scope of the invention. These and other features, aspects, and advantages of the apparatus, system, and method of the present invention will become better understood from the following description, claims, and drawings. It should be understood that the drawings are merely schematic and not drawn to scale. It should also be understood that the same reference numerals are used in all drawings to indicate the same or similar parts.

[0107] This invention provides an apparatus and method for noninvasively estimating one or more hemodynamic parameters (e.g., cardiac output or stroke volume). Embodiments are based on the concept of incorporating information about vascular tone into hemodynamic parameter estimation to improve accuracy. More specifically, embodiments use measurements of the duration ΔT of a blood pulse traveling a specific length of an arterial path from the heart as a surrogate measure of vascular tone and incorporate it into the hemodynamic parameter estimation. Embodiments also consider variations in vascular tone between different parts of the circulatory system by incorporating surrogate measurements of vascular tone for multiple different arterial paths.

[0108] Known methods for estimating hemodynamic parameters are typically based on obtaining measurements of arterial flow in a single artery (e.g., the neck) and using a transfer function to derive estimates of central hemodynamic parameters (e.g., cardiac output or stroke volume), which incorporates an algorithm or computational equation that can compute the estimates of hemodynamic parameters based on the flow measurement input.

[0109] However, such estimation methods typically assume that arterial flow at a single location represents the total arterial flow throughout the entire vascular tree. This assumption is inaccurate. Consequently, currently known methods for estimating hemodynamic parameters fail to account for the differentiated autoregulation in different parts of the circulatory system. For example, variations in vascular tone are caused by both systemic factors (e.g., blood pressure regulation) and local factors (e.g., local regulation of blood flow in specific organs) that affect the entire circulatory system. Therefore, more accurate hemodynamic parameter estimation should preferably consider the possible variations in factors such as vascular tone in different parts of the circulatory system. For example, differences in these factors can be expected between the central and peripheral parts of the circulatory system.

[0110] To illustrate this, consider a scenario where blood flow measurements at the carotid artery (central arterial system) are taken and used to estimate total cardiac output (CO). The body regulates blood flow through the carotid artery to supply the brain with adequate oxygen and nutrients. However, simultaneously, due to some event (such as arm movement or exertion), blood flow to peripheral areas (e.g., the arm) may increase or decrease. This will cause a corresponding increase or decrease in total CO. However, this is undetectable in any changes in blood flow through the carotid artery, because blood flow through the carotid artery is regulated according to the brain's needs. Therefore, it is impossible to obtain an accurate estimate of total CO solely from central (carotid) arterial flow measurements.

[0111] However, by also considering peripheral measurements (e.g., peripheral PAT or PTT), changes in blood output to the peripheral region can be detected and combined with central arterial flow measurements to improve the estimation of total CO. Embodiments of the present invention are based on this principle.

[0112] According to an example of one aspect of the invention, a controller is provided suitable for deriving estimates of at least one hemodynamic parameter of an object, said controller being adapted to:

[0113] Receive input indicating arterial flow measurement results in at least one arterial path of the object, wherein the arterial path is a central arterial path or a peripheral arterial path;

[0114] A measure of the time difference ΔT between the indicated cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along the respective arterial path is obtained for each of the central and peripheral arterial pathways, as a result of blood flow from the heart to the location;

[0115] The estimation results of the hemodynamic parameters are determined based on a combination of the following: the blood flow measurement results of the at least one arterial path, and the time difference measure ΔT for each arterial path.

[0116] The embodiments aim to correct the inaccurate assumption that the flow rate through one branch of a vascular tree represents the flow rate through all other branches. The embodiments are based on longitudinal surrogate measures of vascular tension obtained from at least two branches of the vascular tree. By incorporating these measurements into hemodynamic parameter calculations, more accurate estimates are possible.

[0117] It is difficult to directly measure vascular tension. Embodiments of the present invention are based on the understanding that the time it takes for blood to travel from the heart along a predetermined length of the arterial path (in other words, the speed at which the heart pulse travels through the branches of the vascular tree) is closely dependent on the vascular tension of the arterial path. These two are interrelated. Therefore, this duration can be used as a proxy measure of vascular tension in the arterial path.

[0118] Therefore, in summary, embodiments of the present invention are based on integrating information about vascular tension in different arterial pathways within the circulatory system into the determination of central hemodynamic parameters to improve their accuracy. This information is combined with blood flow information and fed as input to a transfer function, which produces an estimate of hemodynamic parameters (e.g., cardiac output) as output.

[0119] exist Figure 2 This general principle is illustrated schematically. Embodiments of the invention are based on providing information related to or indirectly indicative of vascular tension, combined with information related to blood flow (e.g., a hemodynamic transfer function). The transfer function is configured to generate estimates of one or more hemodynamic parameters as output based on the input information.

[0120] According to one or more embodiments, a surrogate metric of vascular tension can be estimated by calculating the time taken for blood to travel along the length of one or more arterial pathways (e.g., pulse arrival time (PAT) or pulse conduction time (PTT)). This can then be integrated into an algorithm or computation (transfer function) in conjunction with at least arterial blood flow measurements (preferably for at least the central arterial pathway, but optionally, alternatively, for the peripheral pathway) to determine one or more hemodynamic parameters.

[0121] This allows the calculation to be sensitive to differential autoregulation (including dynamic changes in vascular tone) across different arterial pathways. This makes the estimation of central hemodynamic parameters more accurate. Accurate estimation of central hemodynamic parameters allows for improved clinical decision-making and medical interventions, thereby improving patient outcomes.

[0122] Arterial pathway refers to a longitudinal section of a certain length, for example, along one or more arteries in the circulatory system.

[0123] The peripheral arterial pathway refers to the arterial pathway in the peripheral vascular system, that is, the part of the circulatory system that consists of arteries not located in the head, chest, or abdomen (e.g., arteries in the arms, hands, legs, and feet).

[0124] The central arterial pathway refers to the pathway in the central vascular system, such as the arteries in the head, chest, abdomen, or neck.

[0125] A cardiac ejection event refers to the event corresponding to the ejection of blood from the heart, i.e., a heartbeat event, or the systolic phase of the heart. This event can be a defined reference point during the ejection process (e.g., the end of the pre-ejection phase), when the aorta opens and blood begins to be ejected from the left ventricle. In other examples, the event could be the beginning of the pre-ejection phase, when the heart is first electrically activated and begins to contract.

[0126] While two arterial pathways have been mentioned in the description above and in the examples herein, in other embodiments, the ΔT time difference can be determined for more than two arterial pathways. Blood flow measurements can be obtained for one or even two or more arterial pathways. The controller can then be adapted to determine at least one hemodynamic parameter based on a combination of the ΔT value and one or more blood flow measurements from all arterial pathways.

[0127] Figure 3 The general principles of the invention are schematically summarized. Figure 3 An example controller 22 is schematically depicted. The example controller 22 receives an input set 23, uses this input to perform a calculation process, and generates at least one output corresponding to an estimation result of at least one hemodynamic parameter 24. Specifically, the controller 22 receives an input set for the central arterial path, an input set for the peripheral arterial path, and at least one blood flow measurement result input corresponding to the blood flow through at least one of the central and peripheral arterial paths. Preferably, it is a measurement result for the central arterial path. In another example, blood flow measurements for both the central and peripheral arterial paths may be received as input. Furthermore, in another embodiment, inputs for the central arterial path and two or more peripheral arterial paths may be received. Additionally, in some embodiments, inputs for two or more central arterial paths may be received.

[0128] The inputs for the central artery path include an input indicating the time it takes for the pulse wave to arrive at a predetermined location along the central artery path and an input indicating the time of cardiac ejection. Using the arrival time of the pulse wave and the time of the cardiac ejection event, the time difference ΔT_centr between these two events can be calculated.

[0129] For the peripheral artery pathway, a similar set of inputs is also received. Using the arrival time of the pulse wave and the time of cardiac ejection, the time difference ΔT_peri between these two events can be calculated for the peripheral artery pathway.

[0130] Although Figure 3The controller 22 is shown to receive individual inputs for each of the central arterial pathway and peripheral arterial pathway corresponding to a cardiac ejection event, but the controller may also receive a single input indicating the occurrence of a cardiac ejection event.

[0131] Then, controller 22 uses at least one blood flow measurement result (e.g., for the central artery path) and time differences ΔT_centr and ΔT_peri to derive hemodynamic parameters 24. For example, as a non-limiting example, the hemodynamic parameters may be the subject's cardiac output (CO), the subject's stroke volume (SV), the subject's stroke volume variability (SVV), or the subject's stroke volume index (SVI).

[0132] The estimation result of hemodynamic parameter 24 can be determined based on the application of a predefined algorithm or computation function by the controller. For example, the controller 22 can apply a predetermined transfer function 21, and the controller 22 is configured to receive at least one flow measurement result and the time difference ΔT of two or more arterial pathways as input, and generate an output indicating the estimation result of one or more hemodynamic parameters based on the input. This is illustrated in... Figure 3 The transfer function 21 is schematically illustrated. As discussed above, the ΔT value provides a proxy measure of vascular tension in each of the two arterial paths, which enables the algorithm or computation or transfer function applied by the controller 22 to derive more accurate hemodynamic parameter estimates 24.

[0133] For example, a pre-determined or pre-stored function (transfer function) 21 can be used, which embodies a predetermined functional relationship between the inputs, and at least one hemodynamic parameter can be calculated based on the inputs. The transfer function can simply be a mathematical relationship between several input parameters and a central hemodynamic parameter that serves as the output. It can include a machine learning algorithm or engine that has been trained using labeled data to derive specific hemodynamic parameters based on input blood flow and ΔT measurements.

[0134] To derive the transfer function, machine learning and / or statistical methods can be applied, for example, to labeled datasets. For instance, multiparameter regression has been successfully used in trials to estimate such transfer functions based on clinical datasets. Given sufficient data, patient metadata (e.g., gender, BMI, and other personal patient information) can also be used to improve the transfer function. Other methods for providing the transfer function may include, for example, using support vector machines or Naive Bayes models.

[0135] For example, according to one or more embodiments, the transfer function may include any type of machine learning algorithm, such as a logistic regression model, decision tree algorithm, artificial neural network, support vector machine, or Naive Bayes model, or may provide any other type of machine learning algorithm, which is then trained using a training dataset that includes previously acquired data for one or more patients.

[0136] Methods for training machine learning algorithms are well-known. Typically, such methods involve obtaining a training dataset that includes training input data entries and corresponding training output data entries. In this case, the training input data entries will be the acquired inputs to the transfer function, i.e., at least the measured ΔT, PAT, PTT, or ΔPAT values ​​for each of the two arterial pathways, and at least one blood flow measurement. In some embodiments, additional inputs may also be included. The training output data entries will correspond to the hemodynamic parameters being sought, such as cardiac output, stroke volume, and / or stroke volume variability. To build the training data, the hemodynamic parameters are measured manually (e.g., using invasive methods) when the training data entries are acquired, thus ensuring their accuracy.

[0137] An initial machine learning algorithm is applied to each input data entry to generate a predicted output data entry. The error between the predicted output data entry and its corresponding training output data entry is used to modify the machine learning algorithm. This process can be repeated until the error converges and the predicted output data entry is sufficiently similar to the training output data entry (e.g., ±1%). This is often referred to as a supervised learning technique.

[0138] According to one or more embodiments, the resulting training algorithm can then provide the required transfer function.

[0139] There are different ways to obtain the input set for two or more arterial pathways. Specifically, there are different ways to obtain arterial flow measurements for at least one arterial pathway of an object and to obtain a measure of the time difference ΔT between a cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along each of the two corresponding arterial pathways. For example, different sensor modalities can be used to derive these measurement inputs.

[0140] The various possible means or methods for obtaining these measurement results will now be briefly outlined. A set of more detailed embodiments will then be described.

[0141] According to a set of examples, obtaining a time difference metric for each arterial path can include obtaining a pulse arrival time (PAT) measurement for each arterial path.

[0142] In this context, the ejection event can correspond to the electrical activation point of the heart. It corresponds to the beginning of the pre-ejection phase, that is, the point at which the heart pulse first begins.

[0143] Ejaculation events can be detected using an ECG sensor. For example, this can be detected as the time when the QRS complex begins to appear in the ECG signal. However, using an ECG to detect this event is not essential, and other methods can be used. Alternative methods include using inductive sensors, radar sensors, accelerometers, or heart rate sensors (e.g., chest heart rate sensor belts).

[0144] In some examples, a PPG sensor can be used to detect the time it takes for a pulse wave to arrive at a predetermined location along the arterial path. Alternatively, an ultrasound sensing unit can be used to detect the arrival of the pulse wave; for example, an ultrasound transducer unit configured to acquire Doppler ultrasound data at a predetermined pulse arrival location.

[0145] According to one or more embodiments, obtaining a time difference measure ΔT for each arterial path may include obtaining a pulse conduction time (PTT) measurement for each arterial path. Pulse conduction time is the time between blood ejection into the aorta and the arrival of the corresponding pulse wave at the downstream measurement location. In other words, it is the time between the end of the pre-ejection phase (PEP) and the arrival of the pulse wave along the arterial path at the predetermined location.

[0146] In some embodiments, obtaining PTT measurement results for each arterial path may include: obtaining PAT measurement results for each arterial path, obtaining an estimate of the pre-ejection period (PEP) duration, and determining the PTT measurement results for each arterial path by subtracting the PEP duration from the PAT measurement results for each arterial path.

[0147] Pre-ejection phase (PEP) is a term in the art that refers to the period of time from the electrical activation of the heart (e.g., indicated by the QRS complex in an ECG signal) that blood is ejected from the heart into the aorta. This is the time between the electrical activation of the heart and the opening of the aortic valve.

[0148] Therefore, the relationship between pulse arrival time (PAT) and PEP is as follows: PAT = PEP + PTT, where PTT is the pulse conduction time.

[0149] There are different ways to obtain PEP estimates.

[0150] According to a set of examples, the estimated duration of PEP can be derived based on inputs from a phonocardiogram (PCG) sensing unit and / or an impedance cardiogram (ICG) measurement unit.

[0151] Measurements obtained from these sensing modalities can be used to identify the moment of aortic valve opening. This allows determination of the time between cardiac electrical activation (e.g., the onset of the QRS complex in an ECG measurement) and aortic valve opening (the point at which the pressure wave begins to travel along the arterial system). This period corresponds to the pre-ejection phase.

[0152] For example, both aortic valve opening and ejection produce distinct high-pitched sounds: a click when opening and a clatter when ejecting blood. One or both of these sounds can be identified in PCG readings following the first heart sound. This can be used to detect the timing of the ejection event.

[0153] Furthermore, ICG and ECG can also be used together, either additionally or alternatively, to directly identify PEP. The method for achieving this is detailed in the article "Estimated preejection period (PEP) based on the detection of the R-wave and dZ / dt-min peaks in ECG and ICG" by René van Lien, Nienke MSchutte, Jan H Meijer, and Eco JC de Geus (April 18, 2013, IOP Publishing Ltd.).

[0154] In other examples, the controller may use a pre-defined estimate for the PEP, such as storing the pre-defined estimate in local memory or retrieving the pre-defined estimate from a remote data source (e.g., a remote server).

[0155] The controller can be configured to use an ultrasound sensing unit to acquire a metric indicating blood flow measurement results for at least one arterial pathway. The ultrasound sensing unit may include one or more ultrasound transducers. It may include, for example, an ultrasound transducer unit or a probe.

[0156] The ultrasound sensing unit can output ultrasound data to the controller, or it can directly output calculated blood flow measurement results to the controller. The ultrasound sensing unit can acquire Doppler ultrasound data. In some examples, the ultrasound sensing unit may also include a dedicated ultrasound processing unit for extracting one or more blood flow measurements from the acquired ultrasound data and supplying these blood flow measurements to the controller to estimate hemodynamic parameters.

[0157] Using ultrasound sensing is just one example of how to detect blood flow. An alternative is to obtain blood flow measurements indirectly, for example, using blood pressure measurements.

[0158] For example, cardiac output estimates can be derived from noninvasive blood pressure measurements. First, a noninvasive continuous arterial blood pressure waveform is obtained from the finger by applying a cuff around the middle phalanx and using a well-known volume-clamping method. Using a model, the brachial artery blood pressure waveform can be derived from the measured finger arterial blood pressure waveform. Finally, by applying pulse contour analysis to the brachial artery pressure waveform, an estimate of cardiac output can be obtained. This can then be used to estimate blood flow.

[0159] The pulse profilometry method mentioned above uses, for example, the area under the contraction of a blood pressure curve and combines it with, for example, a Windkessel model, and optionally also uses certain patient data (e.g., age, sex, height, and weight) to derive an estimate of stroke volume. This provides an indication of cardiac output when combined with heart rate. Blood flow information can then be derived. This is described, for example, in "Noninvasive continuous hemodynamic monitoring" by Truijen et al. (J Clin Monit Comput, 2012, Vol. 26, pp. 267-278).

[0160] A series of specific embodiments of the invention will now be outlined to further illustrate and demonstrate the principles of the inventive concept.

[0161] Figure 4 The processing workflow for the first set of example embodiments is schematically outlined.

[0162] According to this set of embodiments, pulse arrival time (PAT) is used as a proxy measure of vascular tension. In other words, the time difference ΔT between cardiac ejection events and pulse arrival is measured by acquiring PAT measurements for each of the peripheral and central arterial pathways.

[0163] For example, a peripheral arterial pathway could be an arterial pathway that extends from the heart along the arm to the fingers. A central arterial pathway could be a pathway that extends from the heart through a channel along the neck towards the head.

[0164] In operation, the PPG sensor can be positioned at a stable location along the peripheral arterial path. For example, a finger clip PPG sensor can be used and coupled to the object's finger. This PPG sensor can be used to detect the time ("Pulse arrival_peri") of each pulse wave arriving from the heart to the finger after each ejection event. This can correspond, for example, to a peak or other characteristic reference point in the PPG sensor output. Example reference points used for detection in the raw PPG pulse waveform include, for example, the time of the minimum contraction, the moment of the maximum diastole, the time of the maximum contraction slope, or any other characteristic point.

[0165] Alternatively, pulses reaching locations distal to the peripheral arterial pathway can be measured using other sensing modalities, such as ultrasound (US) sensors, accelerometers (ACC), or cardiac impact measurements.

[0166] ECG sensing devices can be used to detect the timing of ejection events in the heart (“cardiac ejection”). For example, a set of at least two ECG electrodes can be positioned on a subject’s chest to sense the heart’s electrical activity. These electrodes can be attached to a dedicated ECG sensing processor, which is configured to use the electrodes to detect electrical signals. Each time the heart contracts (systole) to eject blood from the left ventricle to pump it through the arterial tree, a characteristic electrical signal can be detected in the ECG output. This signal pattern is called the QRS complex and consists of a characteristic pattern of three peaks: the Q peak, the R peak, and the S peak. Therefore, ejection events can be detected by detecting the QRS complex in the ECG signal (e.g., the start of the QRS complex, or the timing of a specific peak in the QRS complex, such as the R peak (which is typically the maximum peak)).

[0167] In some examples, an ultrasound sensing unit can also be used and positioned, for example, to perform ultrasound sensing on the neck of an object (e.g., at a location along the central artery path (e.g., above the carotid artery)). The ultrasound sensing unit may include one or more ultrasound transducers and a dedicated ultrasound processing unit for processing the acquired ultrasound data. The ultrasound data may be Doppler ultrasound data. The ultrasound sensing unit can be used to detect the time (“pulse arrival_cen”) at which a pulse wave arrives at a defined location along the central artery path. For example, Doppler ultrasound data provides an indication of blood flow through the arterial path, thus enabling the detection of an arrival event of a pulse wave at a given location by a sudden increase in flow or the onset of an upward slope in the flow rate.

[0168] The same ultrasound sensing unit can also be used, for example, to derive a measure of arterial flow through the central arterial pathway by acquiring and processing Doppler ultrasound data.

[0169] Alternatively, different sensing modalities can be used to obtain blood flow measurements, as discussed in more detail above. Measurement of blood flow from at least one arterial pathway is required, preferably the central arterial pathway, but can also be alternatively the peripheral arterial pathway. Optionally, it is possible to acquire blood flow measurements from both the central and peripheral arterial pathways and use them to derive hemodynamic parameter estimates.

[0170] Note that in alternative examples, additional PPG sensors can be used instead of ultrasound sensors to detect pulse arrival events at locations along the central artery path, such as additional PPG sensors placed on the nose, forehead, or concha.

[0171] Using these acquired measurements, measures of pulse arrival time (PAT) for the central and peripheral arterial pathways were derived. PAT is defined as the time difference between the timing of the QRS complex in the electrocardiogram (ECG) signal and the arrival time of the corresponding pulse at the distal location. Therefore, the PAT for the central arterial pathway ("PAT_cen") can be derived by calculating the difference between the time of the cardiac ejection event and the time it takes for the pulse to reach the distal point along the central arterial pathway ("Pulse arrival_cen"). Similarly, the PAT for the peripheral arterial pathway ("PAT_peri") can be derived by calculating the difference between the time of the cardiac ejection event and the time it takes for the pulse to reach the distal point along the peripheral arterial pathway ("Pulse arrival_peri").

[0172] In some examples, the controller then feeds the PAT_cen, PAT_peri, and blood flow measurement results for at least one arterial path to the processing algorithm or computation or transfer function 21 and uses it to derive the estimated results of the hemodynamic parameters 24 (e.g., cardiac output or stroke volume).

[0173] According to another example, optionally, the quotient or ratio between the PAT_cen measurement and the PAT_peri measurement can be calculated, and then the quotient can be further used to calculate the estimated hemodynamic parameter 24. This value can be used to replace, or supplement, the individual PAT_cen and PAT_peri values ​​when calculating the estimated hemodynamic parameter.

[0174] Changes in the PAT_cen measurement indicate changes in vascular tension in the central branches, while changes in the PAT_peri measurement indicate changes in vascular tension in the peripheral branches. Furthermore, the ratio PAT_peri / PAT_cen indicates differential changes in vascular tension. Note that the quotient can also be calculated alternatively in the inverse form of PAT_cen / PAT_peri.

[0175] Note that when acquiring the various measurement results inputs discussed above (PPG, ECG, ultrasound), it is assumed that all measurement or sensor devices are synchronized in time and have millisecond-level accuracy, making the time difference calculation sufficiently accurate.

[0176] In various examples, different combinations of measurements or parameters can be used as inputs to calculate the final estimated hemodynamic parameters. These combinations include, for example, combinations of two or more PAT measurements for different corresponding central arterial paths; combinations of two or more PAT measurements for different corresponding peripheral arterial paths; combinations of at least one PAT measurement for a central arterial path and at least one PAT measurement for a peripheral arterial path; a single quotient of PAT_cen and PAT_peri, or multiple quotients for multiple pairs of different central and peripheral arterial paths.

[0177] While the ECG sensing device discussed above is used to identify ejection events (the onset of cardiac contraction), other measurement or detection units are possible. A non-limiting list of other possible units for detecting ejection events includes: one or more accelerometers (e.g., on the sternum); one or more inductive sensors placed near the heart; radar sensors, transesophageal echocardiography probes placed near the heart; or chest heart rate sensors (e.g., chest strap heart rate sensors).

[0178] Figure 5 The processing workflow for the second set of example embodiments is schematically outlined.

[0179] This set of embodiments and Figure 4 The only difference is that pulse conduction time (PTT) measurements are used as a proxy measure of vascular tone, rather than a proxy measure of pulse arrival time (PAT). In other words, the time difference ΔT between cardiac ejection and pulse arrival is measured by acquiring PTT measurements for each of the peripheral and central arterial pathways. In all other respects, the features and workflow of this set of embodiments are the same as those described above for... Figure 4 The examples described are identical. Therefore, for the sake of brevity, the common features will not be described in detail here.

[0180] To obtain PTT measurements for each arterial path, it is necessary to correct for the PAT value during the pre-ejection phase (PEP), which is the period from when the heart generates blood pressure until it exceeds the aortic pressure and the aortic valve opens. Excluding PEP yields a more accurate surrogate measure of vascular tone because no pulse travels through the arterial path during this period.

[0181] Therefore, obtaining PTT measurements for each arterial pathway can include: obtaining PAT measurements for each arterial pathway, obtaining an estimate of the pre-ejection period (PEP) duration, and determining the PTT measurement for each arterial pathway by subtracting the PEP duration from the PAT measurement for each arterial pathway. Figure 5 This situation is outlined in the text. Figure 5 This demonstrates how to obtain the corresponding PTT value for each item in the central arterial path (PTT_cen) and the peripheral arterial path (PTT_peri) by subtracting the PEP period from the corresponding PAT value for the two arterial paths.

[0182] There are different ways to obtain estimates of the duration of PEP.

[0183] In some examples, the duration of PEP can be derived based on inputs from a phonocardiogram (PCG) sensing unit and / or an impedance cardiogram (ICG) measurement unit.

[0184] Measurements obtained from these sensing modalities can be used to identify the moment of aortic valve opening. This allows determination of the time between cardiac electrical activation (e.g., the onset of the QRS complex in an ECG measurement) and aortic valve opening (the point at which the pressure wave begins to travel along the arterial system). This period corresponds to the pre-ejection phase (PEP).

[0185] As discussed above, aortic valve opening produces distinctive high-pitched sounds: a click as it opens and a clatter as it ejects blood. One or both of these sounds can be identified in PCG readings following the first heart sound. This can be used to detect the timing of aortic valve opening, which can then be used to determine the PEP (platelet-remitting pressure).

[0186] In other examples, controller 22 may use a predetermined estimate for the pre-ejaculation phase, such as storing the predetermined estimate in local memory or retrieving it from a remote data source (e.g., a remote server).

[0187] In some examples, the controller then feeds the PTT_cen, PTT_peri, and blood flow measurements for at least one arterial path to the processing algorithm or computation or transfer function 21 and uses them to derive estimates of the hemodynamic parameters 24 (e.g., cardiac output or stroke volume).

[0188] According to another example, optionally, the quotient or ratio between the PTT_cen measurement and the PTT_peri measurement can be calculated, and then the quotient can be further used to calculate the estimated hemodynamic parameters. This value can be used to replace, or supplement, the individual PTT_cen and PTT_peri values ​​when calculating the estimated hemodynamic parameter 24.

[0189] Note again that when acquiring the various measurement results inputs discussed above (PPG, ECG, ultrasound), it is assumed that all measurement or sensor devices are synchronized in time and have millisecond-level accuracy, making the time difference calculation sufficiently accurate.

[0190] Figure 6 The processing workflow for the third set of embodiments is schematically outlined.

[0191] In this set of embodiments, the controller is configured to detect changes or variations in vascular tension surrogate measurements over time for each arterial path. This provides an indication of changes in the (relative) tension of the branches of the measured vascular tree. These changes can indicate both local and more systemic changes within the vascular tree, which can also improve the accuracy of hemodynamic parameter estimations at any given time.

[0192] For example, the controller can be configured to: acquire multiple time difference measurements ΔT corresponding to different cardiac cycles for each arterial path in the arterial pathway, determine a metric indicating the change of the ΔT value over time for each arterial path, and determine estimates of hemodynamic parameters based on the changes in the ΔT value. The controller can continuously or repeatedly reacquire the ΔT value for each arterial path to monitor changes in the ΔT value as a function of time.

[0193] For example, and as Figure 6 As shown, the controller can be configured to monitor changes or alterations in the PAT value for each of the central and peripheral arterial pathways. For example, for each new heartbeat, the controller can obtain a new PAT measurement for each arterial pathway and calculate the difference ΔPAT between the new PAT value and the previous PAT value for the corresponding arterial pathway. The controller then feeds the PAT change value (ΔPAT_cen) for the central arterial pathway and the PAT change value (ΔPAT_peri) for the peripheral arterial pathway to the transfer function 21 to combine with blood flow measurement information to calculate hemodynamic parameters.

[0194] In other words, the controller uses the increase or decrease in PAT between heartbeats (rather than the absolute value of PAT) as the parameter used to calculate hemodynamic parameters.

[0195] Optionally, the quotient or ratio between the ΔPAT_cen measurement and the ΔPAT_peri measurement can be calculated, and then the quotient can be used to calculate the estimated hemodynamic parameters. This value can be used to replace, or supplement, the individual ΔPAT_cen and ΔPAT_peri values ​​when calculating the estimated hemodynamic parameters.

[0196] Although the example outlined above uses a change in the PAT value, the same principle can also be applied to, for example, a change in the PTT value.

[0197] The unit for each measurement result input in the measurement result input used to obtain the PAT value and to obtain one or more blood flow measurement results can be related to the above-mentioned... Figure 4 and Figure 5 The units described in the example embodiments are the same, so they will not be described in detail here.

[0198] Related to the discussion above Figure 4 and Figure 5 The embodiments are compared in relation to, and are based on Figure 6 In this set of embodiments, when acquiring the various measurement result inputs discussed above (PPG, ECG, ultrasound), it is not necessary for all measurement devices to be time-synchronized. For example, when using more than one measurement device to acquire inputs (e.g., ultrasound, PPG, ECG, etc.), they may not be perfectly time-synchronized. One way to address this issue is to artificially force or enforce synchronization by aligning the common reference points of the time-series data acquired from two or more measurement devices.

[0199] This process can be achieved, for example, by first identifying predefined reference points (characteristic anchors) in the signals from each measuring device. These points can be physiological or triggered by the user, for example, due to movement or by electromagnetic means. In each case, the point corresponds to an event known to have left a signal or trace on the signal from each measuring device.

[0200] Once these common reference points are identified in each signal, the signals can be time-aligned based on these points, that is, the corresponding reference points in each signal are aligned in the time domain.

[0201] Several embodiments of using transfer functions or algorithms to determine the estimation results of hemodynamic parameters based on two or more input parameters have been discussed above. The various options for implementing transfer functions have been briefly outlined above.

[0202] One particular example approach is to use a parametric regression model to map the set of input parameters to the set of output parameters. To further explain and illustrate, the application principles of such an example transfer function will now be outlined in more detail. According to any of the embodiments described above, the same principles can be applied to combine the specific sets of inputs outlined in each embodiment to derive estimated hemodynamic parameters.

[0203] The following example transfer function uses multiparameter regression modeling with three input parameters. For instance, a transfer function with the following input parameters (independent variables) is used to estimate cardiac output (CO):

[0204] Arterial blood flow at the carotid artery (“flow”) (e.g., based on pulse wave Doppler ultrasound measurements), and

[0205] A measure indicating the following: cardiac ejection events and corresponding pulse waves along the central artery path (ΔT). centr ) and peripheral arterial pathway (ΔT) peri The time difference between the arrival of the corresponding arterial path at each item in the equation at the predetermined location.

[0206] The transfer function can be estimated based on the following formula:

[0207] CO=β0+β1Flow+β2ΔT centr +β3ΔT peri

[0208] Where β0 is the intercept of the regression line, and β1, β2, and β3 are the values ​​of the corresponding input parameters Flow and ΔT. centr and ΔT peri Associated weights.

[0209] As discussed above, to train a model, a training dataset comprising historical patient data from multiple patients can be used. Specifically for multi-parameter linear regression models, the model is first built by constructing a model or algorithm that combines each of the desired input parameters as a parameter (independent variable) with its corresponding coefficient or weight. The constructed model is then trained on the training dataset to fit the model coefficients or weights to provide an optimal fit between the input parameters and the corresponding output parameters of the training dataset. The desired input parameters form the model's independent variables, while the target hemodynamic parameter is the model's dependent variable. The model expresses the relevant hemodynamic parameter being estimated as a linear sum of a constant term (intercept) and each dependent variable multiplied by its corresponding weight or coefficient.

[0210] The training dataset will include training input data entries and corresponding training output data entries. In this case, the training input data entries correspond to example values ​​of blood flow measurements for at least one arterial path, and a time difference measure ΔT between the central and peripheral arterial paths for a given patient. The training output data entries correspond to one or more predetermined hemodynamic parameters.

[0211] An initial machine learning algorithm is applied to each input data entry to generate a predicted output data entry. The error between the predicted output data entry and its corresponding training output data entry is used to modify the machine learning algorithm. This process can be repeated until the error converges and the predicted output data entry is sufficiently similar to the training output data entry (e.g., ±1%). This is commonly referred to as a supervised learning technique.

[0212] For multiparameter regression models, the training process is the process of fitting model weights / coefficients to the training dataset. Once the training or fitting process is complete, the weights or coefficients obtained during training or fitting can be used to deploy the model to map the input parameters (independent variables) to the output hemodynamic parameters.

[0213] The performance or accuracy of a generated machine learning model can be evaluated by running the model on a test dataset after training and assessing the error between the output predictions generated by the model and the actual real values. For example, for a linear regression model, performance measures can include the goodness of fit of the linear regression (R²). 2 The parameters include: standard error (SE), t-statistic (tStat) of the estimated results, p-value, root mean square error (RMSE) obtained from the correlation scatter plot, and / or reproducibility coefficient (rpc) obtained from the Bland-Altman plot.

[0214] While the example transfer function above uses blood flow measurements for at least one arterial path and a time difference measure ΔT for both the central and peripheral arterial paths as input parameters, the inputs to the transfer function may differ in other examples. For instance, the transfer function could be adapted to accept one or more parameters derived from the processing of the ΔT measure and / or blood flow measure as input. For example, see the reference above. Figure 4-6 Various examples are described.

[0215] In some examples, as referenced above Figure 4 As explained, the input parameters for the transfer function can be the values ​​of PAT_cen and PAT_peri, or functions thereof (e.g., quotients), combined with a blood flow metric (see the explanation above for more details). In some examples, and as referenced above... Figure 5As explained, the transfer function can be adapted to accept PPT_cen and PPT_peri values ​​or functions thereof (e.g., quotients) as input parameters and combine them with a blood flow metric (see the explanation above for more details). In some examples, and as referenced above... Figure 6 As explained, the transfer function can be adapted to accept ΔPAT_cen and ΔPAT_peri values ​​or functions thereof (e.g., quotients) as input parameters and incorporate a blood flow metric (see the explanation above for more details). In some examples, the transfer function can be adapted to accept a measure of pulse arrival time at a defined location along each of the peripheral and arterial pathways, a measure of cardiac ejection time, and a measure of blood flow as input parameters.

[0216] Furthermore, it should be noted that while the above example of a transfer function is based on only three input parameters, in other examples, the transfer function can accept a greater number of input parameters, i.e., a greater number of independent variables, each with a corresponding weight. Other input parameters may include, for example, parameters derived from blood flow and / or ΔT measurements and / or patient metadata (e.g., sex and body mass index).

[0217] It should be noted that although examples of hemodynamic parameters derived therefrom being stroke volume or cardiac output have been described above, the same inventive concept can be applied to derive any desired hemodynamic parameter. Other examples of hemodynamic parameters that can be derived include: blood flow velocity, stroke volume variability, systolic velocity, diastolic velocity, and blood pressure.

[0218] According to another aspect of the invention, an apparatus is provided for deriving an estimate of at least one hemodynamic parameter.

[0219] Figure 7 An example apparatus according to one or more embodiments is schematically outlined.

[0220] The device includes an ECG sensor arrangement unit 44 for detecting cardiac ejection events.

[0221] The device also includes at least one PPG sensor 46, which is optically coupled to a location along the arterial path of the object to detect the time it takes for a pulse wave to arrive at the location from the heart.

[0222] The device also includes an ultrasound sensing unit 42 for detecting blood flow through the arterial path.

[0223] The device also includes a controller 22 according to any example or embodiment outlined above or described below, or according to any claim of this application, which is operatively coupled to the ECG sensor arrangement unit, at least one PPG sensor, and the ultrasonic sensing unit.

[0224] The ultrasound sensing unit may include an ultrasound transducer arrangement unit. It may be configured to detect blood flow using Doppler ultrasound measurements. It may include a dedicated ultrasound processing unit for extracting one or more flow measurements from the acquired ultrasound data and supplying these flow measurements to a controller for estimating hemodynamic parameters.

[0225] The set of measuring devices and sensors provided for this device represents only one example, and other example means for obtaining various input measurements for the controller are also possible and have been outlined in more detail in the description above.

[0226] According to another aspect of the invention, a computer-implemented method is provided for deriving an estimate of at least one hemodynamic parameter.

[0227] exist Figure 8 The steps of method 60 implemented by an example computer according to one or more embodiments are summarized in the form of a block diagram.

[0228] Method 60 includes:

[0229] Receive 62 input indicating arterial flow measurement results for at least one arterial path of the subject, which is a central arterial path or a peripheral arterial path;

[0230] A measure of the time difference ΔT between the indicated cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along the respective arterial path is obtained for each of the 64 central and peripheral arterial pathways, as a result of blood flow from the heart to said location; and

[0231] The estimation results of 66 hemodynamic parameters are determined based on a combination of the following: blood flow measurements from at least one arterial pathway, and the time difference measure ΔT for each arterial pathway.

[0232] Based on the explanation and description provided above for the apparatus aspect (i.e., the controller aspect) of the present invention, the implementation options and details of each step in the above steps can be understood and interpreted.

[0233] Any examples, options, or embodiment features or details described above with respect to the apparatus aspect (with regard to the controller) of the present invention may be applied or combined or modified as necessary and incorporated into the method aspect of the present invention.

[0234] According to another aspect of the invention, a computer program product including code units configured to cause the processor to perform the method according to any example or embodiment outlined above or described below, or according to any claim of this application, when run on a processor.

[0235] As discussed above, the embodiments utilize controller 22. Controllers can be implemented in various ways using software and / or hardware to perform a variety of desired functions. A processor is one example of a controller employing one or more microprocessors, which can be programmed using software (e.g., microcode) to perform the desired functions. However, controllers can be implemented with or without a processor, and can also be implemented as a combination of dedicated hardware performing some functions and processors (e.g., one or more programmed microprocessors and associated circuitry) performing other functions.

[0236] Examples of controller components that may be used in various embodiments of this disclosure include, but are not limited to, conventional microprocessors, application-specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs).

[0237] In various implementations, the processor or controller may be associated with one or more storage media, such as volatile and non-volatile computer memories, such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that, when run on one or more processors and / or controllers, perform the required functions. The various storage media may be fixed within the processor or controller, or may be removable, allowing one or more programs stored thereon to be loaded into the processor or controller.

[0238] Those skilled in the art, through studying the accompanying drawings, the disclosure, and the claims, will be able to understand and implement variations of the disclosed embodiments when practicing the claimed invention. In the claims, the word "comprising" does not exclude other elements or steps, and the words "a" or "an" do not exclude a plurality.

[0239] A single processor or other unit can implement the functions of several items as described in the claims.

[0240] Although certain measures are described in different dependent claims, this does not mean that combinations of these measures cannot be used advantageously.

[0241] Computer programs can be stored / distributed on suitable media, such as optical storage media or solid-state media supplied together with or as part of other hardware, but can also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0242] If the term “suitable” is used in the claims or description, it should be noted that the term “suitable” is intended to be equivalent to the term “configured as”.

[0243] No reference numerals in the claims should be construed as limiting the scope.

Claims

1. A controller suitable for determining the estimation result of at least one hemodynamic parameter of an object, said at least one hemodynamic parameter comprising at least one of the following: cardiac output, stroke volume, and stroke volume variability, said controller being suitable for: Receives data input indicating arterial blood flow measurement results in at least one arterial path of the object, wherein the arterial path is a peripheral arterial path or a central arterial path; A measure of the time difference ΔT between the indicated cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along the respective arterial path is obtained for each of the peripheral and central arterial pathways, as a result of blood flow from the heart to said location; The system provides blood flow measurements for the at least one arterial pathway and a metric or parameter derived from the metric indicating the time difference ΔT between the central arterial pathway and the peripheral arterial pathway, as input to a predetermined transfer function, wherein... The transfer function is adapted to generate the estimation result for the at least one hemodynamic parameter based on the input and a predetermined functional relationship between the input and the at least one hemodynamic parameter; and Generate data output indicating the estimation result of the at least one hemodynamic parameter.

2. The controller according to claim 1, wherein, Obtaining time difference metrics for each arterial pathway includes obtaining pulse arrival time (PAT) measurements for each arterial pathway.

3. The controller according to claim 1 or 2, wherein, Obtaining time difference measurements for each arterial pathway includes obtaining pulse conduction time (PTT) measurements for each arterial pathway.

4. The controller according to claim 3, wherein, The process of deriving the pulse conduction time measurement for each arterial path includes: obtaining the PAT measurement for each arterial path, deriving an estimate of the pre-ejection period (PEP) duration, and determining the pulse conduction time measurement for each arterial path by subtracting the PAT duration from the PAT measurement for each arterial path.

5. The controller according to any one of claims 1-2, wherein, The controller is configured to: obtain multiple time difference measurements ΔT corresponding to different cardiac cycles for each arterial path in the arterial path, determine a metric indicating the change of the ΔT value over time for each arterial path, and determine the estimated result of the hemodynamic parameter based on the change of the ΔT value.

6. The controller according to any one of claims 1-2, wherein, Determining the estimated results of the hemodynamic parameters includes determining a quotient of one or more ΔT values ​​for the peripheral and central pathways.

7. The controller according to any one of claims 1-2, wherein, The controller is configured to use an ultrasound sensing unit to obtain a metric indicating the blood flow measurement result for at least one arterial path in the arterial path.

8. The controller according to any one of claims 1-2, wherein, The controller is configured to, for at least one of the arterial pathways, use the same ultrasound sensing unit to obtain both a measure of arterial flow and a measure of the time it takes for the pulse wave to reach the predetermined location along the arterial pathway.

9. The controller according to any one of claims 1-2, wherein, The ejection event is the moment of electrical activation of the heart, corresponding to the beginning of the pre-ejection phase of the heart; or The ejection event corresponds to the end of the pre-ejection phase of the heart.

10. The controller according to any one of claims 1-2, wherein, The transfer function includes a machine learning algorithm.

11. An apparatus for deriving an estimate of at least one hemodynamic parameter, comprising: The first sensor unit is used to detect cardiac ejection events; A second sensor unit, which is coupled to a location along the arterial path of the object, is used to detect the time it takes for a blood pulse wave to arrive at the location from the heart; The third sensor unit is used to detect blood flow through the arterial path; as well as The controller according to any one of claims 1-10 is operatively coupled to the first sensor unit, the second sensor unit and the third sensor unit.

12. A computer-implemented method for determining an estimate of at least one hemodynamic parameter, said at least one hemodynamic parameter comprising at least one of the following: cardiac output, stroke volume, and stroke volume variability, said method comprising: Receive data input indicating arterial blood flow measurement results in at least one arterial path for the subject, wherein the arterial path is a peripheral arterial path or a central arterial path; A measure of the time difference ΔT between the indicated cardiac ejection event and the arrival of the corresponding pulse wave at a predetermined location along the respective arterial path is obtained for each of the central and peripheral arterial pathways, as a result of blood flow from the heart to said location; The system provides blood flow measurement results for the at least one arterial path and a metric or parameter derived from the metric indicating the time difference ΔT between the central arterial path and the peripheral arterial path as input to a predetermined transfer function, wherein the transfer function is adapted to generate the estimation results for the at least one hemodynamic parameter based on the input and a predetermined functional relationship between the input and the at least one hemodynamic parameter. and Generate data output indicating the estimation result of the at least one hemodynamic parameter.

13. The computer-implemented method according to claim 12, wherein, Obtaining time difference measurements for each arterial pathway includes obtaining pulse arrival time (PAT) measurements and / or pulse conduction time (PTT) measurements for each arterial pathway.

14. A computer program product comprising a computer-readable medium having computer-readable code embodied therein, the computer-readable code being configured to cause the computer or processor, when run by a suitable computer or processor, to perform the method according to any one of claims 12-13.

Images