Methods used to determine stroke volume

Abstract

A method and apparatus for determining the cardiac stroke volume (SV) of an individual are provided, comprising the steps of providing a first pulse contour stroke volume or providing a conventionally derived pulse contour stroke volume based on one or more characteristics of a measured arterial blood pressure waveform, determining at least one perfusion parameter describing perfusion of fat-free mass and fat mass through the individual's body, and / or determining at least one fluid response parameter function depending on a fluid response parameter describing the individual's cardiopulmonary interaction, and adjusting the first pulse contour stroke volume or the conventionally derived pulse contour stroke volume based on at least one of the perfusion parameters and / or the fluid response parameter function to provide a second pulse contour stroke volume.

Description

Technical Field

[0001] The present invention relates to a method for determining cardiac stroke volume. Furthermore, the present invention relates to an apparatus for determining cardiac stroke volume and a computer-readable medium for performing the method. In particular, the present invention relates to a method, an apparatus and a computer-readable medium for determining cardiac stroke volume based on invasive or non-invasive arterial pulse contour analysis, i.e. the so-called pulse contour stroke volume PCSV. Background Art

[0002] Cardiac output, which is defined as the volume of blood pumped by the heart over time, is an important parameter for diagnosing and managing patients, especially during surgical procedures in intensive care units or emergency rooms. Cardiac output is usually measured in liters per minute. Cardiac output CO depends on the volume of blood pumped by the heart with each beat or heartbeat, which is called the stroke volume, and the number of heartbeats in a given time, which is called the heart rate HR and which corresponds to the pulse rate PR.

[0003] Conventionally, cardiac output or also stroke volume can be measured using the pulmonary artery thermodilution method, in which a well-defined small amount of cold fluid at a known temperature is injected into the right atrium and the time-dependent blood temperature development is measured after passing through the right ventricle in the pulmonary artery. The time-dependent blood temperature development allows the calculation of cardiac output using the Stewart-Hamilton equation, given the known volume, specific gravity and specific heat capacity of the injected fluid and the specific gravity and specific heat capacity of the blood. The (average) stroke volume can be calculated by dividing the cardiac output by the heart rate. However, the determination of stroke volume using the highly invasive pulmonary artery thermodilution method is risky due to threatening complications, time consuming, considerable cost and impracticality in an operating room or emergency room.

[0004] In order to make cardiac output (CO) comparable between individuals, it is usually normalized to body surface area and produced in l / min / m 2 The cardiac index is expressed in units of CI.

[0005] Conventionally, there are many methods to quantify cardiac output CO. One CO method that is applied minimally and non-invasively in most cases is the so-called pulse contour analysis. Thereby, certain characteristics of the arterial blood pressure waveform are used to estimate the pulse contour stroke volume PCSV on a beat-by-beat basis. Another method is to identify the part of the arterial blood pressure waveform that corresponds to left ventricular ejection and estimate PCSV based on this. Thereby, the algorithm used to determine or estimate PCSV can take into account the pressure pulse wave area (i.e., systolic pressure area) during the ejection phase, arterial compliance, impedance and systemic vascular resistance, among other parameters. Since it is based only on blood pressure and time data, the PCSV estimate obtained has not been related to the stroke volume in milliliters of blood for the individual patient (e.g., size and condition) or has not been calibrated to the stroke volume in milliliters of blood.

[0006] As described above, no invasive procedures need to be performed in order to determine the pulse contour stroke volume (PCSV). Therefore, any non-invasively measured data related to the arterial pulse contour can be used. Even in emergency surgical procedures, the non-invasive pulse contour will usually be determined, and the pressure values ​​and pressure curves used to determine the blood pressure (particularly time-dependent arterial pressure data or non-invasive tissue pressure pulse waveform) will be used to determine the stroke volume.

[0007] However, the invention is also applicable to minimally invasive measurements of the pulse contour. Even in such cases, improvements are achieved in terms of risk, time and cost compared to the pulmonary artery thermodilution method for determining stroke volume.

[0008] WO 2009 / 014420 A1 relates to a method for determining beat-to-beat stroke volume using the waveform of arterial pressure data.

[0009] However, it has been found that the (average) stroke volume determined using only time-dependent arterial pressure data (hereinafter referred to as pulse contour stroke volume (PCSV)) shows only moderate consistency and correlation with the (average) stroke volume determined by a reference method (hereinafter referred to as SVref, such as pulmonary artery and transpulmonary thermodilution or transesophageal echocardiography). Although it is possible to initially or repeatedly calibrate the PCSV using the SVref method, this is an invasive method using thermodilution, which is always expensive and very time-consuming. Therefore, it is impractical to apply such a calibrated PCSV in an intensive care unit or emergency room during surgery. More importantly, the changes in PCSV of existing methods show poor or moderate correlation, consistency and compliance and large percentage errors compared to the changes in SVref, providing strong evidence that such PCSV methods are unable to track changes in SVref. Therefore, methods based only on the patient's time-dependent arterial pressure data (especially in the operating room or emergency room) are unreliable.

[0010] Furthermore, any PCSV calibration is not sufficient by using the patient's sex, age, height, weight and body surface area (BSA) for biological calibration or setting the pulse contour stroke volume (PCSV). Normalizing CO to BSA (especially for very obese patients) can lead to an overestimation of the cardiac index (CI) (l / min / m 2 ). In addition, any conventional uncalibrated biological setting of PCSV and pulse contour cardiac output (PCCO) is complicated by the fact that extrapolation of BSA from an average body size with a normal fat-free mass / fat mass ratio (FFM / AM ratio) to the BSA of a very large obese body size does not take into account the substantially reduced FFM / AM ratio in obese individuals with much lower AM perfusion.

[0011] Furthermore, known PCSV methods exhibit modest to poor sensitivity in detecting changes in SV compared with gold standard reference methods, resulting in poor correlation, clinically unacceptably large limits of agreement, high percentage errors, and suboptimal concordance rates. Significant improvements are necessary to determine PCSV with high accuracy and precision, regardless of invasive or noninvasive measurements. Summary of the invention

[0012] It is therefore an object of the present invention to provide an improved method for reliably determining stroke volume. In particular, it is an object of the present invention to provide a method for improving the determination of stroke volume based on any pulse contour analysis. Furthermore, it is an object to increase the dynamic sensitivity of PCSV to better track real changes in PCSV (e.g. compared to changes in SVref). It is another object of the present invention to provide a method for improving any PCSV determination that calculates PCSV for each heartbeat and increasing the dynamic sensitivity of PCSV to better track real changes in PCSV.

[0013] Said object has been solved by the subject matter of the independent claims. Advantageous embodiments are disclosed in the dependent claims.

[0014] Therefore, the main idea on which the present invention is based is to "calibrate" and / or calculate the conventionally measured PCSV more accurately by using other important parameters. There are two different starting points. One starting point is the uncalibrated PCSV_uncal, which is dimensionless and needs to be calibrated for the individual. The other starting point is the conventional PCSV_conv, which has been biologically calibrated for the individual in terms of biological characteristics and is provided in ml as described above (e.g. by using BSA), but it is still not accurate enough compared to the reference PCSV measured by risky and time-consuming invasive measurements, especially with regard to rapid changes in the PCSV.

[0015] In one aspect of the invention, any uncalibrated pulse contour stroke volume PCSV can be improved by taking into account the blood flow through the metabolically most active very well perfused fat-free mass FFM and the metabolically very inactive and lightly perfused fat mass AM. In the next first PCSV or PCSV_uncal, biological calibration of any uncalibrated PCSV can be achieved by dividing the blood into the portion flowing through the lean body mass (i.e., fat-free mass FFM) and the portion flowing through the fat mass AM. Therefore, these two blood flow portions must be calculated separately and added.

[0016] Furthermore, since cardiac output CO decreases due to increased stiffness or decreased compliance of the arterial system and decreased cardiac function and also depends on demographic gender (it is generally higher in male individuals), this effect must also be taken into account when providing a second PCSV according to the invention.

[0017] In another aspect of the present invention, a fluid response parameter FRP indicative of cardiopulmonary interaction induced by any form of breathing (e.g., spontaneous breathing or fully controlled mechanical ventilation) is used to improve the first PCSV (PCSV_uncal or conventional PCSV_conv (which is derived using medical data obtained from the individual)). Therefore, unlike the first aspect, FRP can be applied to both the dimensionless uncalibrated PCSV_uncal and the dimensional conventional PCSV_conv because it is calibrated for the biological characteristics of the individual under conventional circumstances.

[0018] Each given fluid-responsive heart has a lower stroke volume before blood is supplied than after blood is supplied. Vice versa, a given fluid-responsive heart has a higher stroke volume before blood loss than after blood loss. This behavior is used to improve PCSV (PCSV_uncal or PCSV_conv).

[0019] Therefore, the fluid response parameter FRP is taken into account in the method of the present invention for improving the PCSV calculation, thereby improving the sensitivity of the PCSV to track changes in the true stroke volume SVref measured using the reference method.

[0020] Taking the FRP into account allows for improved accuracy regardless of the breathing pattern (which is for example controlled mechanical ventilation, any type of pressure support ventilation, airway pressure release ventilation or regular spontaneous breathing or any other breathing or ventilation pattern that causes cardiopulmonary interaction, preferably with a known or measurable regular cardiopulmonary interaction phase that is long enough to reliably determine the FRP). Thus, by using the FRP, the individual health status of the individual is taken into account to improve the accuracy of the PCSV.

[0021] The fluid response parameter FRP used in the present invention can be represented by, for example, pulse pressure variation PPV, mean left ventricular ejection pressure variation MEPV, stroke volume variation SVV, systolic pressure variation SPV, systolic pressure area variation SPAV or photoplethysmographic variation index PVI or any other fluid response parameter with appropriate sensitivity and specificity or a combination of several items and, for example, their relative weighted averages.

[0022] MEPV is the change in mean left ventricular ejection pressure (MEP) caused by cardiopulmonary interaction.

[0023] SPAV is the change in systolic pressure area Asys caused by cardiopulmonary interaction, Asys being the area enclosed by the arterial pressure curve during systole. This area is defined as the time from the beginning of systole to the beginning of the dicrotic notch (which corresponds to the closure of the aortic valve). Its baseline is defined as:

[0024] A horizontal line at the level of the diastolic arterial blood pressure DAP of the previous pulse, or

[0025] The straight line between the DAP of the previous pulse and the DAP of the pulse under consideration.

[0026] According to the present invention, if the FRP is a combination of several parameters, these parameters must be adjusted to comparable ranges by applying the structural regression analysis equation before generating the weighted average.

[0027] This object is achieved by a method for determining the cardiac stroke volume SV of an individual, the method comprising the following steps: providing a first pulse contour stroke volume PCSV_uncal or providing a conventionally derived pulse contour stroke volume PCSV_conv based on one or more characteristics of a measured arterial blood pressure waveform, determining at least one perfusion parameter BioCal describing the perfusion of fat-free mass and fat mass through the body of the individual, and / or determining at least one fluid response parameter function f(FRP_norm) depending on a fluid response parameter FRP_norm describing the cardiopulmonary interaction of the individual, and adjusting the first pulse contour stroke volume PCSV_uncal or the conventionally derived pulse contour stroke volume PCSV_conv based on at least one of the perfusion parameter BioCal and / or the fluid response parameter function f(FRP_norm) to provide a second pulse contour stroke volume PCSV_imp.

[0028] Therefore, this purpose can be achieved by applying the perfusion parameter BioCal or the fluid response parameter FRP_norm. Alternatively, this purpose is achieved by applying the BioCal parameter and the FRP parameter FRP_norm.

[0029] This purpose can also be achieved by applying only the FRP parameter FRP_norm when the already biologically calibrated PCSV_conv is derived in the conventional way.

[0030] Preferably, providing the first pulse contour stroke volume PCSV_uncal is based on non-invasive or invasive measurement of the arterial blood pressure waveform. Here, the first pulse contour stroke volume PCSV_uncal is a value derived from the dimensionless arterial blood pressure waveform.

[0031] In a preferred embodiment, in order to obtain the highest accuracy and precision, the method for determining the cardiac stroke volume SV of an individual includes the following steps: providing a first pulse contour cardiac stroke volume PCSV_uncal based on one or more characteristics of the measured arterial blood pressure waveform, determining at least one perfusion parameter BioCal describing the perfusion of fat-free mass and fat mass through the body of the individual, and determining at least one fluid response parameter function f(FRP_norm) depending on the fluid response parameter FRP_norm describing the cardiopulmonary interaction of the individual, and adjusting the first pulse contour cardiac stroke volume PCSV_uncal based on the perfusion parameter BioCal and the fluid response parameter function f(FRP_norm) to provide a second pulse contour cardiac stroke volume PCSV_imp.

[0032] Preferably, the step of determining at least one fluid response parameter FRP_norm is based on at least one of the following: arterial pressure variation PPV, mean left ventricular ejection pressure variation MEPV, stroke volume variation SVV, systolic pressure variation SPV, left ventricular systolic area variation SPAV, photoplethysmography variation index PVI or any other fluid response parameter or combination thereof with appropriate sensitivity and specificity.

[0033] Preferably, the fluid response parameter FRP_norm is determined based on a weighted average of the pulse pressure variation PPV and the systolic pressure variation SPV or based on MEPV and SPAV. In the most preferred solution, in order to optimally track the real PCSV variation, the fluid response parameter FRP_norm is determined based on a weighted average of the FRPs (e.g., pulse pressure variation PPV, mean left ventricular ejection pressure variation MEPV and left ventricular systolic area variation SPAV) with the highest sensitivity and specificity to cardiopulmonary interaction.

[0034] Preferably, the fluid response parameter FRP_norm is determined by applying a fluid response normalization function f_FRP_MV for mechanical ventilation, or the fluid response parameter FRP_norm is determined by applying a fluid response normalization function f_FRP_SB for spontaneous breathing, thereby eliminating contaminating parameter changes (e.g., tidal volume, lung compliance, and chest wall compliance) to obtain a pure standardized fluid response parameter that meets the standard.

[0035] Preferably, the fluid response parameter FRP_norm is determined by considering the compliance Crs of the individual's respiratory system, wherein the compliance Crs of the respiratory system includes lung compliance Cls and chest wall compliance Ccw, because various components of the respiratory system may have different effects on the intrathoracic hypotension volume system IBCS.

[0036] Preferably, in order to normalize the tidal volume TV, the fluid response parameter FRP_norm is normalized by using the FRP normalization parameter TVnorm, wherein the FRP normalization parameter TVnorm is TV normalized to predicted body weight PBW or TV normalized to body weight W, but preferably TV normalized to fat-free mass FFM.

[0037] Preferably, the fluid response parameter FRP_norm is adjusted for mechanical ventilation MV by using predetermined semi-quantitative information about the chest wall compliance Ccw.

[0038] Preferably, the fluid response parameter FRP_norm is normalized to a mean value for spontaneous breathing SB by using the lung compliance Cls and respiratory rate RR to replace TVnorm, which is unknown in most cases in SB.

[0039] Preferably, the step of determining a perfusion parameter BioCal describing the perfusion of the fat-free mass and the fat mass through the individual's body comprises at least one of: determining a pulse contour stroke volume calibration factor PCSV_FFM_Cal related to the fat-free mass, determining a pulse contour stroke volume calibration factor PCSV_AM_Cal related to the fat mass.

[0040] Preferably, the method further comprises determining individual demographic parameters, including: determining the fat-free mass FFM, determining the fat mass AM, the individual's height h, the individual's biological sex and the individual's age y, as these parameters have the greatest impact on the absolute level of resting SV.

[0041] Preferably, the pulse contour stroke volume calibration factor PCSV_FFM_Cal related to fat-free mass is determined by using the perfusion coefficient B_FFM related to perfusion through the metabolically most active fat-free mass. Alternatively or additionally, the pulse contour stroke volume calibration factor PCSV_AM_Cal related to fat mass is determined by using the perfusion coefficient B_AM related to perfusion through the metabolically very inactive fat mass.

[0042] Preferably, in order to determine the pulse contour stroke volume calibration factor PCSV_FFM_Cal related to the fat mass and / or the pulse contour stroke volume calibration factor PCSV_AM_Cal related to the fat mass, the fat-free mass layer thickness and / or the fat mass layer thickness are taken into account, because in patients under full anesthesia and muscle relaxation, the perfusion through the FFM also depends on the FFM layer thickness and the AM layer thickness, and the perfusion through the AM also depends on the AM layer thickness.

[0043] Preferably, measuring or providing one or more characteristics of the arterial blood pressure waveform comprises identifying a portion of the arterial blood pressure waveform.

[0044] The object is also achieved by a device for determining the cardiac stroke volume SV of an individual, the device comprising: an arterial blood pressure and waveform monitor, which is used to provide a first pulse contour stroke volume PCSV_uncal or a conventional pulse contour stroke volume PCSV_conv based on medical data obtained from an individual connected to the arterial blood pressure and waveform monitor; a parameter unit, which is used to provide the perfusion parameter BioCal and / or the cardiopulmonary interaction description parameter FRP_norm of the individual; a controller, which is used to adjust the first PCSV_uncal or the conventional PCSV_conv based on the perfusion parameter BioCal and / or the cardiopulmonary interaction description parameter FRP_norm to provide a second pulse contour stroke volume PCSV_imp; and a medical device, which receives the second PCSV_imp and outputs the second PCSV_imp or related data obtained from the second PCSV_imp. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0046] Figure 1 A flow chart for a first method for improving a first pulse contour stroke volume PCSV_uncal by means of a perfusion parameter BioCal and outputting a second pulse contour stroke volume PCSV_imp is shown;

[0047] Figure 1aA flow chart for determining a first pulse contour stroke volume PCSV_uncal is shown;

[0048] Figure 1b A portion of an arterial blood pressure waveform used to determine a first pulse contour stroke volume PCSV_uncal is shown;

[0049] Figure 2 shows a flow chart for determining a perfusion parameter BioCal as part of a first method for improving a first pulse contour stroke volume;

[0050] Figure 3 A model for the perfusion of fat mass AM and fat-free mass FFM is shown depending on the layer thickness AM / H of the fat mass and the layer thickness of the fat-free mass;

[0051] Figure 4 An example of the relationship between the perfusion coefficient B_FFM of the fat-free mass and the layer thickness FFM / H of the fat-free mass for different muscle tone levels is shown;

[0052] Figure 5 An example of the relationship between the reduction function AM_B_FFM_Red and the layer thickness AM / H in the fat mass for different muscle tone levels is shown;

[0053] Figure 6 The semi-quantitative effect on B_FFM and the reduction function AM_B_FFM_Red depending on the different degrees of muscle tension of the individual is shown;

[0054] Figure 7 An example of the relationship between the perfusion coefficient B_AM of a fat mass and the layer thickness AM / H of the fat mass is illustrated;

[0055] Figure 8 A flow chart for a second method for improving the pulse contour stroke volume PCSV by means of a fluid response parameter function f(FRP_norm) is shown;

[0056] Fig. 9 A flow chart for determining the FR parameter FRP is shown;

[0057] Fig.10 shows a flow chart for determining a fluid response parameter function f(FRP_norm) applied in mechanical ventilation MV as part of a second method for improving a first pulse contour stroke volume;

[0058] Fig.11 A schematic lung and factors affecting PCSV are shown;

[0059] Fig.12The figure shows the effect of chest wall compliance Ccw on the heart-lung interaction HLI during mechanical ventilation MV;

[0060] Fig.13 The semi-quantitative direction and degree of adjustment of the FRP normalization function f_FRP_MV depending on the level of chest wall compliance are illustrated;

[0061] Fig.14 shows a flow chart for determining a fluid response parameter function f(FRP_norm) applied in spontaneous breathing SB as part of a second method for improving a first pulse contour stroke volume;

[0062] Fig.15 The relationship between the pulse pressure variation PPV and the respiratory rate RR as a representative fluid response parameter FRP and an approximation function f_FRP(RR) based on a pair of measurements from a data set of measurements taken from volunteers are illustrated;

[0063] Fig.16 is based on Fig.15 The FRP standardized function f_FRP_SB for adults is derived from the approximate function f_FRP(RR);

[0064] Fig.17 The semi-quantitative direction and degree of adjustment of the FRP normalization function f_FRP_SB depending on the level of tidal volume TV are shown;

[0065] Fig.18 Schematic diagram showing the effect of the compliance Cls of the lung system on the cardiopulmonary interaction HLI during spontaneous breathing SB;

[0066] Fig.19 The semi-quantitative direction and degree of adjustment of the FRP normalization function f_FRP_SB depending on the level of compliance Cls of the lung system are shown;

[0067] Fig. 20 An example of an FRP normalization function f_FRP_SB_RR depending on the compliance Cls of the lungs in relation to the default compliance Cls0 of the lungs and the tidal volume TV is shown;

[0068] Fig.21 shows an example of the form of the fluid response parameter function f(FRP_norm) obtained from a fitting function ln_fit_fct based on pairs of measurements from a data set of measurements taken from a patient;

[0069] Fig. 22shows a regression diagram illustrating the results of a method for determining PCSV (where PCSV=PCSV_conv) based on a Wesseling algorithm for providing PCSV_uncal by using conventional biological calibration parameters BSACal but without applying a fluid response parameter function f(FRP_norm) relative to a reference SVref;

[0070] Fig.23 shows a regression plot illustrating the results of a method for determining PCSV (where PCSV=PCSV_conv) based on an alternative algorithm for providing PCSV_uncal by using conventional biological calibration parameters BSACal but without applying the fluid response parameter function f(FRP_norm) relative to a reference SVref;

[0071] Fig.24 shows a regression diagram illustrating the results of a method for determining a change ΔPCSV of PCSV (here PCSV=PCSV_conv) with respect to a corresponding initial PCSV based on a Wesseling algorithm for providing PCSV_uncal by using conventional biological calibration parameters BSACal but without applying a fluid response parameter function f(FRP_norm) versus a change ΔSVref of SVref with respect to a corresponding initial SVref;

[0072] Fig.25 shows a regression diagram illustrating the results of a method for determining a change ΔPCSV of a PCSV (here PCSV=PCSV_conv) with respect to a corresponding initial PCSV based on an alternative algorithm for providing PCSV_uncal by using a biological calibration BSACal but without applying a fluid response parameter function f(FRP_norm) relative to a reference SVref change ΔSVref with respect to a corresponding initial SVref;

[0073] Fig.26 shows a regression diagram illustrating the results of the method according to the invention for determining a PCSV (here PCSV=PCSV_imp) based on a Wesseling algorithm for providing a first PCSV_uncal by using the perfusion parameter BioCal of the invention but without applying the fluid response parameter function f(FRP_norm) relative to a reference SVref;

[0074] Fig. 27shows a regression diagram illustrating the results of a method according to the invention for determining a PCSV (here PCSV=PCSV_imp) based on an alternative algorithm for providing a first PCSV_uncal by using the perfusion parameter BioCal of the invention but without applying the fluid response parameter function f(FRP_norm) relative to a reference SVref;

[0075] Fig.28 shows a regression diagram illustrating the results of a method according to the present invention for determining a change ΔPCSV of PCSV (here PCSV=PCSV_imp) with respect to a corresponding initial PCSV based on a Wesseling algorithm for providing PCSV_uncal by using perfusion parameters BioCal but without applying a fluid response parameter function f(FRP_norm) relative to a change ΔSVref of SVref with respect to a corresponding initial SVref;

[0076] Fig.29 shows a regression diagram illustrating the results of a method according to the invention for determining a change ΔPCSV in PCSV (here PCSV=PCSV_imp) relative to a corresponding initial PCSV based on an alternative algorithm for providing PCSV_uncal by using perfusion parameters BioCal but without applying a fluid response parameter function f(FRP_norm) relative to a change ΔSVref in SVref relative to a corresponding initial SVref;

[0077] Fig.30 shows a regression diagram illustrating the results of a method according to the invention for determining a PCSV (here PCSV=PCSV_imp) based on a Wesseling algorithm for providing a first PCSV_uncal by using a perfusion parameter BioCal and applying a fluid response parameter function f(FRP_norm) relative to a reference SVref;

[0078] Fig.31 shows a regression diagram illustrating the results of a method according to the invention for determining a PCSV (here PCSV=PCSV_imp) based on an alternative algorithm for providing a first PCSV_uncal by using the perfusion parameter BioCal of the invention and applying a fluid response parameter function f(FRP_norm) relative to a reference SVref;

[0079] Fig.32shows a regression diagram illustrating the results of the method according to the invention for determining a change ΔPCSV of PCSV (here PCSV=PCSV_imp) in relation to a corresponding initial PCSV based on a Wesseling algorithm for providing PCSV_uncal by using the perfusion parameter BioCal according to the invention and applying a fluid response parameter function f(FRP_norm) versus a change ΔSVref of SVref in relation to a corresponding initial SVref;

[0080] Fig.33 shows a regression diagram illustrating the results of a method according to the invention for determining a change ΔPCSV of PCSV (here PCSV=PCSV_imp) in relation to a corresponding initial PCSV based on an alternative algorithm for providing PCSV_uncal by using the perfusion parameter BioCal of the invention and applying a fluid response parameter function f(FRP_norm) versus a change ΔSVref of SVref in relation to a corresponding initial SVref;

[0081] Fig.34 The relative errors of the second PCSV_imp and the conventional PCSV determined according to the present invention with respect to the reference stroke volume SVref are shown;

[0082] Fig.35 An arrangement according to the invention is shown. DETAILED DESCRIPTION

[0083] Figure 1 A rough overview of the method according to the present invention according to a first embodiment is provided.In the first embodiment, a biological calibration of the perfusion parameter BioCal based on any uncalibrated pulse contour stroke volume PCSV_uncal is performed by considering the separation of blood flow into fat-free mass FFM and fat mass AM.

[0084] according to Figure 1 Flowchart shown, in step S100, an uncalibrated pulse contour stroke volume PCSV_uncal (also referred to as a first PCSV) is provided. In the next step S200, a perfusion parameter BioCal is determined. This will be described in further detail below. After the perfusion parameter BioCal is determined, the perfusion parameter BioCal is applied to the first PCSV_uncal in step S280 to provide a second PCSV_imp in step S290.

[0085] Figure 1aAn exemplary flow chart for providing a first PCSV_uncal is provided. In step S70, one or more characteristics of the arterial blood pressure waveform are determined or measured. This can be invasive or non-invasive. Therefore, the result of such determination and / or measurement is a parameter of the individual's blood pressure curve, which allows the identification of a portion of such a blood pressure curve in step S80. Therefore, in step S85, the first PCSV_uncal can be determined taking into account values ​​such as amplitude, area enclosed by the curve, etc., and the first PCSV_uncal is output in step S90. Methods for providing PCSV_uncal are known in the art. A known example for providing such a PCSV_uncal is the Wesseling algorithm (Wesseling algorithm in "The Effect of Signal Quality on Six Cardiac Output Estimator" by Chen et al. in Comput Cardiol, January 1, 2009). However, other algorithms can also be used to provide PCSV_uncal.

[0086] Figure 1b An example for providing such a PCSV_uncal is illustrated. Figure 1b A portion of a blood pressure curve measured (e.g., by invasive measurement using a high-fidelity oscillometric hydraulic blood pressure cuff or by measuring the blood pressure directly using a fluid column and a pressure transducer) is illustrated. In a simple exemplary manner, a first PCSV_uncal is determined by using the blood pressure, pulse pressure, the systolic pressure area Asys indicated by hatching, and other parameters (e.g., aortic compliance and arterial impedance). The systolic pressure area Asys is the area enclosed by the arterial pressure curve during systole (i.e., the left ventricular ejection phase), which is defined as the time from the beginning of systole (with the peak of the systolic arterial blood pressure) to the beginning of the dicrotic notch (which corresponds to the closure of the aortic valve). Its baseline is defined as:

[0087] (like Figure 1b a horizontal line of the previous pulse at the level of the diastolic arterial blood pressure DAP (as shown in FIG. 1 ), or

[0088] The straight line between the DAP of the previous pulse and the DAP of the pulse under consideration.

[0089] Mean left ventricular ejection pressure (MEP) is the average value of the arterial pressure curve during systole. Figure 1b is shown as a horizontal line.

[0090] As described above, an individual's cardiac output depends on the individual's biometric / demographic parameters (eg, weight W), including fat-free mass FFM and fat mass AM, height H, age, and gender.

[0091] Therefore, in algorithms that use a summary formula to determine SV, the primary determinant of stroke volume SV is the individual's biometric / demographic parameters.

[0092] A secondary determinant of the SV of an individual is the hemodynamic state. Therefore, pulse contour evaluation of the measured arterial blood pressure pulse waveform results in dynamic input parameters from which the uncalibrated PCSV (PCSV_uncal) can be calculated.

[0093] According to a first embodiment of the invention, BioCal provides a calibration function for scaling any given PCSV_uncal to an absolute PCSV value according to the individual's biometric / demographic input parameters.

[0094] exist Figure 2 In the embodiment of the present invention, a detailed flow chart for determining the perfusion parameter BioCal is given. As shown in the first embodiment of the present invention, the fat-free mass is separated from the fat mass. Therefore, in steps S205 and S255, the fat-free mass FFM and the fat mass AM are determined respectively.

[0095] The fat-free mass FFM and fat mass AM are explained in detail below.

[0096] According to Formula 1, FFM (kg) and AM (kg) represent the two parts of an individual's body weight W (kg):

[0097] (1)W(kg)=FFM(kg)+AM(kg)

[0098] FFM is the body's highly metabolically active mass, which consumes a lot of oxygen and therefore makes up the largest portion of SV. AM is the mass of very inactive metabolically tissue, which consumes very little oxygen and therefore makes up only a very small portion of SV.

[0099] To estimate FFM (kg) and AM (kg), the formulas in "Lean body mass: the development and validation of prediction equations in healthy adults" by Yu et al. (Yu, Heitmann, and Janmahasatian, BMC Pharmacol Toxicol, October 14, 2013) were used and averaged:

[0100] (2)FFM(kg)=(56.128kg+(1.3016+9270 / (8780+244m 2 / kg·BMI(kg / m 2 )))·W(kg)-2.2268m 2 BMI (kg / m 2 )-0.1134kg / y·age(y)) / 3, for female individuals

[0101] FFM(kg)=(66.068kg+(1.4138+9270 / (6680+216m 2 / kg·BMI(kg / m 2 )))·W(kg)-2.2268m 2 BMI (kg / m 2 )-0.1134kg / y·age(y)) / 3, for male individuals

[0102] (3)AM(kg)=W(kg)-FFM(kg)

[0103] Formula (2) is an example for prediction of FFM and can be replaced by a further improved prediction formula.

[0104] For adolescents, children, and infants, other appropriate FFM and AM estimation formulas were used and averaged.

[0105] It is important to note that such FFM and AM estimates are relative to corresponding conditions in a healthy state and generally assume that 1) the water content of FFM is constant and 2) AM contains no or very little water. Therefore, when applied to a patient's clinical condition, what can be entered into such formulas is not the actual body weight (which may not be true due to, for example, widespread edema or effusion), but only the last known pre-illness body weight.

[0106] Based on the determination of FFM and AM, the relationship between stroke volume SV (ml) and FFM and AM can be given:

[0107] (4) SV (ml) B_FFM(ml / kg)·FFM(kg)+B_AM(ml / kg)·AM(kg)+…+B_Const(ml)

[0108] Among them, B_FFM is the perfusion coefficient related to FFM perfusion, B_AM is the perfusion coefficient related to AM perfusion, and B_Const takes into account other influences on SV. The perfusion coefficients B_FFM and B_AM are determined in steps S215 and S265, respectively.

[0109] In “Relations of stroke volume and cardiac output to body composition: the strong heart study” by Collis et al. (Vol. 103, No. 6, February 13, 2001) and “Relationship between body composition and left ventricular geometry using three dimensional cardiovascular magnetic resonance” by Corden et al. (J Cardiovasc Magn Reson, May 31, 2016, Vol. 18, No. 1, Page 32), it has been found that B_FFM and B_AM can distinguish between men and women.

[0110] Applying different but constant perfusion coefficients B_FFM and B_AM takes into account the difference in blood flow per mass of the FFM FFM_perf (ml / min / 0.1kg) and blood flow per mass of the AM AM_perf (ml / min / 0.1kg), but assumes that the perfusion behavior of the FFM and AM parts is independent of the relationship of the FFM to the AM. However, B_FFM and B_AM are not constant ( Figure 4 and Figure 7 ).

[0111] The blood flow rate AM_perf (ml / min / 0.1kg) through the AM is referred to as AM(kg) for short, and the perfusion coefficient B_AM depends largely on the AM share of body weight AM%=AM(kg) / W(kg), which indicates that the perfusion coefficient B_AM must be adjusted independently. However, the dependence of the perfusion coefficient B_AM on AM% is not sufficient to describe the variability of B_AM, because it does not include information on the absolute level of AM.

[0112] Therefore, according to the present invention, a simple model is provided that uses the distribution of AM and FFM over the height of an individual to determine the influence on the perfusion coefficients B_AM and B_FFM.

[0113] In the simplified model, AM and FFM are represented as two corresponding layers, such as Figure 3 In step S210, the FFM layer thickness is estimated by FFM / H. In step S260, the AM layer thickness is estimated by AM / H.

[0114] The following table gives normal values ​​for FFM / H, AM / H and other biometric / demographic parameters.

[0115]

[0116] Figure 3 The model shown is based on these considerations and appropriate perfusion functions have been developed to describe the behavior of the perfusion coefficients B_FFM and B_AM.

[0117] exist Figure 3 In the figure, the white part is the fat mass AM, and the gray part is the fat-free mass FFM. Figure 3 As shown, AM perfusion (white) and FFM perfusion (grey) depend on the layer thickness AM / H and FFM / H, where H = body height. The density of the hatching indicates the level of perfusion - dense hatching = higher perfusion ( Figure 3 Left side), wide hatching = lower perfusion ( Figure 3 Right). It can be easily seen that the perfusion through the FFM is higher in the case of a thinner FFM layer than in the case of a thicker FFM layer. Likewise, the perfusion through the AM and FFM is higher in the case of a thinner AM layer than in the case of a thicker AM layer.

[0118] In patients with low or no muscle tone (e.g., unconscious or anesthetized patients with neuromuscular blockade), the muscles lack their structural support function. Therefore, the weight of the FFM layer affects the perfusion coefficient B_FFM due to gravity, because this weight compresses most of the FFM, i.e., muscles, abdominal viscera, etc. Therefore, as the thickness of the FFM / H layer increases, the perfusion coefficient B_FFM decreases.

[0119] It has also been recognized that with increasing layer thickness AM / H of the fat mass AM, the perfusion coefficient B_FFM decreases.

[0120] In patients without muscle tone, the AM layer thickness decreases the perfusion coefficient B_FFM due to gravity as the AM layer enlarges, because the enlarged subcutaneous fat layer additionally compresses a large part of the FFM.

[0121] Furthermore, it has been found that the perfusion coefficient B_AM decreases due to gravity as AM / H increases due to the lack of supporting structures in adipose tissue. Furthermore, the perfusion coefficient B_AM is not affected by the layer thickness of FFM / H.

[0122] Therefore, if Figure 4 The function B_FFM(FFM / H) shown decreases as FFM / H increases, and is preferably a nonlinear function, such as an inverse tangent function, a sigmoid function, or a logarithmic function, preferably represented by the following formula:

[0123] (5)B_FFM(FFM / H)(ml / kg)=a_FFM(ml / kg)·arctan[(FFM / H(kg / m)–b_FFM(kg / m))·c_FFM(m / kg)]+d_FFM(ml / kg)

[0124] where a_FFM, b_FFM, c_FFM and d_FFM are individual calibration coefficients and constants related to FFM, with B_FFM ranging from 0.2…0.7 ml / kg.

[0125] exist Figure 4 An exemplary process of B_FFM (FFM / H) is given in , which shows the relationship between B_FFM and FFM / H for different levels of muscle tension, which shows that: as FFM / H increases, B_FFM decreases nonlinearly.

[0126] according to Figure 2 The flow chart in FIG. 1 needs to determine the layer thickness of FFM and AM in steps S210 and S260, respectively. This is done using Figure 3 Based on these layer thicknesses FFM / H and AM / H, the corresponding perfusion coefficients B_FFM and B_AM are determined in steps S215 and S265, respectively.

[0127] In the next step S220, the perfusion coefficient B_FFM is adjusted by applying a reduction function AM_B_FFM_Red(AM / H). The reduction function can also be replaced by an attenuation function. In the following, only the application of the reduction function is described. The reduction function AM_B_FFM_Red describes the effect of the thickness of the AM on the perfusion coefficient B_FFM. The reduction function AM_B_FFM_Red increases with the increase of AM / H and is preferably a nonlinear function, such as an inverse sine function, a sigmoid function or a logarithmic function. Figure 5 The examples shown are for different levels of muscle tone.

[0128] The reduction function AM_B_FFM_Red increases as AM / H increases, and is preferably a nonlinear function, such as an inverse sine function, a sigmoid function, or a logarithmic function as described in Equation 6:

[0129] (6)AM_B_FFM_Red(AM / H)(ml / kg)=a_att(ml / kg)·arctan[(AM / H(kg / m)–b_att(kg / m))·c_att(m / kg)]+d_att(ml / kg)

[0130] where a_att, b_att, c_att and d_att are individual calibration coefficients and constants related to AM, where AM_B_FFM_Red has a range of 0…0.2 ml / kg.

[0131] return Figure 2 In step S220 of FIG. 5 , the corrected perfusion coefficient B_FFM_corr is the difference between the preliminary B_FFM (FFM / H) and AM_B_FFM_Red (AM / H), as shown in Formula 7:

[0132] (7)B_FFM_corr(FFM / H,AM / H)(ml / kg)=B_FFM(FFM / H)(ml / kg)-AM_B_FFM_Red(AM / H)(ml / kg)

[0133] Therefore, it has been found that, depending on the level of consciousness, anesthesia and / or muscle relaxation, the perfusion coefficient B_FFM (FFM / H) and the reduction function AM_B_FFM_Red (AM / H) have to be adjusted.

[0134] Figure 6 The semi-quantitative effect on B_FFM and the reduction function AM_B_FFM_Red depending on the different degrees of muscle tension of the individual is shown. Figure 4 and Figure 5 As shown, the effects of FFM layer thickness and AM layer thickness on B_FFM and AM_B_FFM_Red are most pronounced in a state of lack of muscle tone (e.g., fully anesthetized and relaxed patients). The above effects are least pronounced in a state of normal resting muscle tone, and moderately significant in a state of reduced muscle tone (e.g., unconscious patients).

[0135] However, the function B_AM (AM / H) will also decrease as AM / H increases, and is preferably a nonlinear function, such as an inverse tangent function, a sigmoid function, or a logarithmic function. The function B_AM (AM / H) is given by Formula 8:

[0136] (8)B_AM(AM / H)(ml / kg)=a_AM(ml / kg)·arctan[(AM / H(kg / m)–b_AM(kg / m))·c_AM(m / kg)]+d_AM(ml / kg)

[0137] where a_AM, b_AM, c_AM and d_AM are individual calibration coefficients and constants related to AM, with B_AM ranging from 0…0.4 ml / kg.

[0138] exist Figure 7An exemplary course of B_AM (AM / H) is given in , which shows that the perfusion coefficient B_AM of the fat mass AM decreases with increasing layer thickness AM / H of AM.

[0139] Alternatively, the perfusion coefficient B_AM (AM / H) can be replaced by a function with similar behavior, which describes that the AM perfusion decreases with increasing AM / H. The perfusion coefficient B_FFM (FFM / H) can be replaced by a function with similar behavior, which describes that the FFM perfusion decreases with increasing FFM / H and further decreases with increasing AM / H.

[0140] Alternatively, the ratio of AM to body surface area, AM / BSA or AM / H 2 Instead of AM / H, the ratio of FFM to body surface area is used (FFM / BSA or FFM / H). 2 To replace FFM / H.

[0141] If it is not possible to distinguish between FFM and AM because premorbid weight and / or age are unknown (as is sometimes the case in cases of unidentified or unknown ED patients being transferred to the ICU), W / H or W / BSA or W / H can be used. 2 or PBW / H or PBW / BSA or PBW / H 2 This relationship can be approximated by , where PBW is the expected body weight of the individual.

[0142] Back to Figure 2 In steps S225 and S275, the individual static pulse contour stroke volume, the PCSV calibration factor related to the fat-free mass FFM and the fat mass AM are determined by using the perfusion coefficients B_FFM_corr and B_AM_corr, respectively.

[0143] (9)PCSV_FFM_Cal(ml)=B_FFM_corr(ml / kg)·FFM(kg)

[0144] (10)PCSV_AM_Cal(ml)=B_AM(ml / kg)·AM(kg)

[0145] To obtain the individual's second PCSV_imp, the first PCSV_uncal is calibrated by using the individual's static pulse contour stroke volume, PCSV calibration factors PCSV_FFM_Cal, PCSV_AM_Cal.

[0146] (11)PCSV(ml)=PCSV_uncal·(PCSV_FFM_Cal(ml)+PCSV_AM_Cal(ml))·P_Cal

[0147] Among them, PCSV_uncal and P_Cal are dimensionless.

[0148] In addition, if Figure 2 As shown in step S230 in FIG. 1 , an additional static demographic calibration factor P_Cal is applied.

[0149] The first PCSV_uncal reflects the actual hemodynamic state of the individual, which depends on dynamic influences such as cardiac preload (ie, depends on intrathoracic blood volume) and performance (ie, contractility), aortic compliance, arterial impedance and blood pressure (ie, cardiac afterload).

[0150] P_Cal is a calibration factor that depends on the biometric / demographic characteristics of the individual in addition to FFM and AM, which affect the level of SV.

[0151] The static biometric / demographic calibration factor P_Cal can be the sum or product of various biological characteristics / demographic parameters, as shown in Equation 12:

[0152] (12)P_Cal=H(m)·coeff_H(1 / m)+age(yrs)·coeff_age(1 / yrs)+const_gender

[0153] Among them, coeff_H represents the height of the individual, coeff_age represents the age of the individual (compliance of the arterial system), and const_gender is dimensionless and represents the biological sex of the individual.

[0154] If transgender, semi-quantitative input information about the individual's physique is required. This estimate can, for example, be a value in the range of 0.0 to 1.0 (eg, 0.7), where male 1.0, while women 0.0.

[0155] The static biological / demographic calibration factor P_Cal can be refined by an additional constant, dimensionless, representing additional parameters affecting the PCSV level that are not already taken into account by the input data H, age and sex:

[0156] (13)P_Cal=H(m)·coeff_H(1 / m)+age(yrs)·coeff_age(1 / yrs)+const_gender+const_BC

[0157] The PCSV calculation can be further refined by an additional calibration constant const_BC, which represents an additional parameter affecting PCSV that has not been considered in the above PCSV calibration process.

[0158] Therefore, based on PCSV_FFM_Cal and PCSV_AM_Cal and the demographic calibration factor P_Cal, a perfusion parameter BioCal is determined in step S240 , which is applied to the first PCSV_uncal.

[0159] Then, add a second PCSV_imp:

[0160] (14)PCSV_imp(ml)=PCSV_uncal·[(PCSV_FFM_Cal(ml)+PCSV_AM_Cal(ml)+const_corr(ml))·P_Cal]

[0161] BioCal(ml)=(PCSV_FFM_Cal(ml)+PCSV_AM_Cal(ml)+const_corr(ml))·P_Cal

[0162] PCSV_imp(ml)=PCSV_uncal·BioCal(ml)

[0163] The coefficients and constants of the function that produces the static input parameters of the calibration constants PCSV_FFM_Cal, PCSV_AM_Cal and P_Cal are determined in the evaluation population.

[0164] In this evaluation population, the coefficients, slopes and intercepts () of the regression and correlation analyses of PCSV calculated according to formula (14) were optimized (manually and using Microsoft Excel solver) and compared with the gold standard stroke volume SVref measured simultaneously using transpulmonary thermodilution to obtain these coefficients and constants for formula (14).

[0165] exist Figure 8 In the present invention, a flow chart describing a second embodiment of the present invention is provided. In the second embodiment of the present invention, a fluid response parameter FRP is used to better track changes in the pulse contour stroke volume. Therefore, a PCSV improvement function is developed that reflects the effect of the fluid response FR on the stroke volume. If the heart responds to changes in preload, the PCSV improvement function can improve the sensitivity of the pulse contour stroke volume algorithm.

[0166] Figure 8The general flow of providing a second PCSV_imp in step S390 by using the fluid response parameter (and in particular the normalized fluid response parameter FRP_norm and the fluid response parameter function f(FRP_norm) derived based on the fluid response parameter FRP_norm) is illustrated. The second PCSV_imp is a PCSV improved based on the method or apparatus of the present invention.

[0167] In a first step S100, similar to Figure 1-7 Step S100 in the described first embodiment of the invention provides a first pulse contour stroke volume PCSV_uncal. The process can be applied either to the first pulse contour stroke volume PCSV_uncal or to a conventional PCSV result PCSV_conv (which can be obtained using different methods). PCSV_uncal is a dimensionless (unitless) value based on data derived from blood pressure; whereas the conventional pulse contour stroke volume (PCSV_conv) is measured in ml and includes consideration of medical data obtained from the individual based on medical data obtained from the individual.

[0168] In order to provide the first pulse contour stroke volume PCSV_uncal, it is only necessary to provide an arterial blood pressure waveform that can be obtained by using invasive and / or non-invasive measurement, or the first pulse contour stroke volume PCSV_uncal can be obtained based on the reconstruction of the non-invasive tissue pressure pulse waveform and the arterial blood pressure waveform. Figure 1b The method described is used to derive the first pulse contour stroke volume PCSV_uncal. However, several other methods can also be used to provide the first pulse contour stroke volume PCSV_uncal. For the present invention, it is not important how to obtain the first pulse contour stroke volume PCSV_uncal. As mentioned above, the core of the present invention is to provide a method for improving the accuracy of the conventionally obtained first pulse contour stroke volume PCSV_uncal.

[0169] As mentioned above, cardiopulmonary interaction HLI occurs in different situations that need to be distinguished because the parameters that affect HLI are different in these situations. HLI occurs during spontaneous breathing SB and mechanical ventilation MV. MV can be divided into fully controlled mechanical ventilation CMV or assisted spontaneous breathing ASB and any other ventilation form between SB and CMV.

[0170] First, the fluid response parameters during mechanical ventilation will be discussed. In MV, an external pump (i.e., mechanical ventilator) forces air or any other ventilation gas mixture into the airways during inspiration, thereby inflating the lungs with positive pressure. Inflation of the lungs causes expansion of the thoracic cavity and caudal movement of the diaphragm. At the same time, during MV inspiration, the expanding lungs compress the intrathoracic low-pressure volume system (IBCS), which mainly consists of the vena cava, right atrium, right ventricle, pulmonary circulation, and left atrium. This compression causes a decrease in venous return, right ventricular preload, and stroke volume, which may also be detected again in the aorta and arterial system after passing through the lungs and left heart due to a decrease in stroke volume or a corresponding decrease in systolic and pulse pressures.

[0171] To understand how HLI in MV can be exploited, it helps to imagine the lungs as a test tool for relative filling of the IBCS and assessing whether cardiac preload is adequate. Relative filling means that only semi-quantitative graded information is obtained, e.g., whether cardiovascular filling is adequate, excessive, insufficient, or too little. With this, the fluid response parameter FRP is usually calculated as a percentage change in at least one breath of a ventilation cycle consisting of inspiration and expiration.

[0172] Different FRPs (e.g., pulse pressure variation PPV, mean left ventricular ejection pressure variation MEPV, stroke volume variation SVV, systolic pressure variation SPV, systolic pressure area variation SPAV or any other suitable FRP) can be used alone or in combination. In the latter case, it is necessary to adjust the different FRPs to comparable levels and ranges (e.g., by applying a regression analysis equation before generating a combined FRP (e.g., calculating a weighted average)).

[0173] An example of a function for balancing the range of values ​​of SPV to fit PPV is shown below:

[0174] (15)PPV*(%)=coeff·SPV(%)+const

[0175] Where coeff and const are derived from regression analysis and are dimensionless. PPV and PPV* are combined into FRP as weighted values:

[0176] (16)FRP(%)=(PPV(%)+w·PPV*(%)) / (1+w)

[0177] Where w is a dimensionless weighting factor.

[0178] An example of a function for equalizing the value range of the MEPV to fit the PPV is shown below:

[0179] (17)PPV1*(%)=coeff_MEPV·MEPV(%)+const_MEPV

[0180] Among them, coeff_MEPV and const_MEPV are derived from regression analysis and are dimensionless.

[0181] An example of a function for equalizing the range of SPAV values ​​to fit the PPV is shown below:

[0182] (18)PPV2*(%)=coeff_SPAV·SPAV(%)+const_SPAV

[0183] Among them, coeff_SPAV and const_SPAV are derived based on regression analysis and are dimensionless.

[0184] PPV, PPV1*, and PPV2* are combined into FRP as a weighted average:

[0185] (19)FRP(%)=(PPV(%)+w1·PPV1*(%)+w2*PPV2*(%)) / (1+w1+w2)

[0186] Among them, w1 and w2 are dimensionless weighting factors.

[0187] The absolute level of FRP depends not only on the hemodynamic filling state, but also on the effectiveness of the FR testing tool (ie, the lung) and the force and conditions of the test.

[0188] Given a constant IBCS (e.g., in MV), a smaller tidal volume will result in a lower FRP value than a larger tidal volume will result in a higher FRP value. The observed changes in FRP when the tidal volume is changed are directly caused by the concomitant changes in the intrathoracic pressure, ITP, which reduce the transmural pressure in the IBCS and thus its volume.

[0189] Thus, normalization of the FRP is accomplished, thereby eliminating changes to parameters other than the filling status of the IBCS (eg, changes in tidal volume).

[0190] The value of FRP derived in equation (16) or (19) is called the combined FRP, which is expressed in %. Fig. 9 The combined FRP value needs to be normalized to generate a normalized FRP (ie, FRP_norm). Here, MV and SB must be distinguished, which will be explained below.

[0191] With the normalized fluid response parameter FRP_norm for MV and SB, the normalized FRP_norm is used to generate or determine the fluid response parameter function f(FRP_norm) in step S300. Therefore, the fluid response parameter FRP_norm must first be determined.

[0192] In MV, the fluid response parameter FRP_norm is calculated based on Formula 20:

[0193] (20)FRP_norm(%)=FRP(%)·f_FRP_MV

[0194] Using Formula 21 in SB

[0195] (21)FRP_norm(%)=FRP(%)·f_FRP_SB

[0196] f_FRP_MV and f_FRP_SB denote normalization functions, which will be described below.

[0197] In formulas (20) and (21), FRP is Fig. 9 The combined fluid response parameters output in step S308 in Fig. 9 The single fluid response parameter output in step S304 in Fig. 9 To illustrate the determination of the FRP.

[0198] In step S302, at least one FRP is determined. The following blood pressure related data can be used to determine at least one FRP: pulse pressure variation (PPV), mean left ventricular ejection pressure variation (MEPV), stroke volume variation (SVV), systolic pressure variation (SPV), systolic pressure area variation (SPAV), photoplethysmographic variation index (PVI), or any FRP with appropriate sensitivity and specificity.

[0199] If only one of the above FRPs is determined, a decision is made in step S303 to output a single FRP (in %) in step S304. If at least two FRPs are determined, one of them is selected as the guiding FRP in step S305. Since the other of the at least two FRPs must be adjusted, an adjustment to the other FRP is performed in step S306. Thus, at least one other FRP is adjusted to a comparable level and / or range of the guiding FRP. This adjustment is performed by using a regression analysis, where coefficients and constants are used to adjust the levels of different FRPs to the guiding FRP, as shown in Formulas 15, 17, and 18.

[0200] After at least one other FRP is adjusted to be the guiding FRP in step S306, a combined FRP is generated in step S307. The dimension of the combined FRP is %. Figure 8 In step S300 of , a single FRP or a combined FRP is used to determine FRP_norm. Fig.10 In step S316 in the above process, the combined FRP or the single FRP is normalized by the FRP normalization function.

[0201] In the following, the normalization of a single FRP or a combined FRP for mechanical ventilation is described. Reference is made to the above mentioned formula 17. Therefore, in order to normalize the FRP to derive FRP_norm, a normalization function f_FRP_MV will be deployed.

[0202] First, the need for standardization is explained. When the lung is used as a test tool for the filling state of IBCS, standardization is necessary because the individual components of the respiratory system may affect IBCS differently. Therefore, the entire respiratory system and its components need to be considered, and each component may significantly change an individual from a healthy state to a diseased state.

[0203] An important factor is the compliance of the respiratory system Crs. The static compliance of the entire respiratory system Crs includes the lung compliance itself Cls and the chest wall compliance Ccw.

[0204] According to Grinnan et al. (2005[2]), the following relationship applies, as shown in formula (22):

[0205] (22)Crs(ml / cmH 2 O)=Ccw(ml / cmH 2 O)·Cls(ml / cmH 2 O) / (Ccw(ml / cmH 2 O)+Cls(ml / cmH 2 O))

[0206] The normal Crs of an adult's lung is 70-100 ml / cmH 2 O, in acute severe ARDS (adult respiratory distress syndrome) and chronic pulmonary fibrosis, lung Crs may decrease to 25ml / cmH 2 O.

[0207] Ccw describes the compliance of the chest wall, i.e., the rib cage and the diaphragm are considered as a whole. For example, in sepsis of abdominal origin, Ccw decreases due to stiffening of the diaphragm caused by increased intra-abdominal pressure; the same happens with any disease that stiffens the rib cage, such as in Bechterew's disease. Vice versa, Ccw increases due to an open abdomen during surgical procedures. Compared to a chest wall with normal Ccw, a reduced Ccw requires higher mechanical inspiratory pressures to achieve the same level of lung inflation, thus increasing ITP to a higher level. Vice versa, this also applies to a chest wall with an increased Ccw.

[0208] For illustrative purposes, refer to Fig.11 and Fig.12 , Fig.11 and Fig.12 The respiratory system including the lungs and chest wall (both of which affect the IBCS) is illustrated.

[0209] In routine clinical practice in the operating room, the mechanical ventilator is set to CMV with a target tidal volume, peak inspiratory pressure limit, positive end-expiratory pressure and ventilation rate for a given situation. It is usually impractical or even impossible to perform a standardized test procedure (e.g. with standardized tidal volume). Therefore, in addition to the level of IBCS, the resulting FRP value needs to be standardized by compensating / removing (one or more) parameters that affect FRP in order to obtain a standardized FRP_norm that is comparable between individuals.

[0210] In mechanically ventilated individuals, changes in FRP are caused by changes in intrathoracic pressure (ITP changes), which are caused by the corresponding tidal volume TV applied by the ventilator. When IBCS is kept constant, a larger TV will cause a larger FRP value, while a smaller TV will cause a smaller FRP value.

[0211] Using the same TV in a patient with a stiffer rib cage and lower Ccw will result in a greater change in ITP, and thus a higher FRP value, than in an otherwise identical patient with a normal rib cage and normal Ccw.

[0212] Therefore, FRP will be standardized based on tidal volume TV, which must also be standardized to make it comparable between individuals.

[0213] According to the definition of the NIH-NHLBI ARDS Network, in common clinical practice, TV is standardized according to predicted body weight (PBW): PBW is calculated according to formulas (23) and (24):

[0214] (23) Male, PBW (kg) = 50kg + 2.3kg / in (height (in) – 60in)

[0215] (24) Female, PBW (kg) = 45.5kg + 2.3kg / in (height (in) – 60in)

[0216] This normalization considers only sex and height, limiting the input to between 48 and 84 inches (123–213 cm), but does not consider weight and age, and is therefore rather crude. Normalizing TV to PBW based solely on equations (23) and (24) does not provide an algorithm for estimating PBW for infants and children. Furthermore, it assumes a linear relationship between height and PBW, which is highly atypical for most biological relationships.

[0217] Therefore, it is proposed to relate TV to the most metabolically active compartment of the body, which is called lean body mass or fat-free mass FFM. FFM is best correlated with oxygen uptake, cardiac output, left ventricular mass, and variables of lung function (e.g., total lung capacity TLC), where TLC is closely related to maximum lung capacity.

[0218] The predicted estimate of FFM (in kg) is described above in the first embodiment with formulas (2) and (3), and will also be applied here. The estimation of FFM using other appropriate algorithms can also be applied to children and infants.

[0219] In summary, the FRP normalization function used to normalize a single FRP or a combined FRP is calculated based on the following formula:

[0220] (25)f_FRP_MV=TVnorm0(ml / kg) / TVnorm(ml / kg)

[0221] Therefore, the FRP normalization function is normalized using the relationship of TVnorm to a default value TVnorm0 (e.g. 8 ml / kg FFM for adults), while TVnorm is calculated based on TV and FFM according to the following formula:

[0222] (26)TVnorm(ml / kg)=TV(ml) / FFM(kg)

[0223] exist Fig.10 In step S310 of , a normalized tidal volume TVnorm is provided. TV is read from a ventilator used for mechanical ventilation. If TVnorm=TVnorm0, then FRP_norm is equal to FRP. In step S312, an FRP normalization function is determined based on the normalized TVnorm.

[0224] As shown in Formula 25, the normalization of the FRP normalization function can be refined using additional coefficients coeff_MV and constants const_MV:

[0225] (27)f_FRP_MV=TVnorm0(ml / kg) / TVnorm(ml / kg)·coeff_MV+const_MV

[0226] For example, when PPV and SPV are used as combined FRP, in adults, TVnorm0 = 6...12 ml / kg FFM, coeff_MV = 0.5...2 and const_MV = -1...1.

[0227] Normalization of FRP to TVnorm already compensates for the different conditions of Cls, since the effect of MV with constant TVnorm on ITP is the same for any Cls.

[0228] As mentioned above, the respiratory system also includes the chest wall. Therefore, as required in step S314, f_FRP_MV is adjusted for different conditions of chest wall compliance Ccw. Here, input data is required to provide information about whether Ccw is normal, increased or decreased and the degree to which Ccw is increased or decreased. The preferred degree scale is:

[0229] (a) Moderate increase (↑)

[0230] (b) Normal (-)

[0231] (c) Moderate decrease (↓)

[0232] (d) Large decrease (↓↓)

[0233] (e) Very large decrease (↓↓↓)

[0234] But it is not limited to these.

[0235] For example, an increase in Ccw can be observed in cases where the abdomen has been opened (moderate enlargement) or where the chest has been opened (large to large enlargement). However, the latter situation is difficult to grade, because the situation changes greatly if the chest is partially or fully opened (maximum open situation if both lungs are freely exposed to atmospheric pressure). Therefore, due to practical problems, the graduation of such a situation is not taken into account.

[0236] Further references Fig.12 , Fig.12 The effect of Ccw on HLI in MV is illustrated, where large / low HLI is shown as large / small distance between vertical dashed lines.

[0237] Therefore, the FRP normalization function f_FRP_MV is adjusted depending on the direction and level of Ccw. An increase in Ccw requires adjusting f_FRP_MV to a higher level, and vice versa, as Fig.13 The level of adjustment can be made in quantifiable steps and will be input by a physician or medical personnel based on the health condition and physical characteristics of the individual's chest wall.

[0238] All the above standardized methods for compensating FRP for different TVnorm and Ccw can generally work in any positive pressure mechanical ventilation mode.

[0239] After considering different conditions of the compliance Ccw of the chest wall in step S314, the FRP normalization function f_FRP_MV is adjusted accordingly. In step S316, the FRP normalization function f_FRP_MV is applied to a single FRP or a combined FRP to achieve a normalized FRP_norm.

[0240] The normalized FRP_norm is used to develop a fluid response parameter function f(FRP_norm) in step S318, which is a linear function or a nonlinear function or a combination thereof, which increases the PCSV value at a small FRP_norm value and decreases the PCSV value at a large FRP_norm. Fig.21 An exemplary f(FRP_norm) is illustrated.

[0241] The possible linear form of the fluid response parameter function f(FRP_norm) is:

[0242] (28)f(FRP_norm)=a·FRP_norm(%)+d

[0243] Here, a and d are dimensionless, and a<0.

[0244] The possible basic nonlinear form of the fluid response parameter function f(FRP_norm) is:

[0245] (29)f(FRP_norm)=a·ln(FRP_norm(%))+d

[0246] Where a and d are dimensionless and a<0

[0247] It can be optimized by other adjustment factors as follows:

[0248] (30)f(FRP_norm)=a·ln(FRP_norm(%)·bc)+d

[0249] Wherein, a<0, wherein the dimensionless adjustment coefficients a, b, c and d are derived based on clinical studies.

[0250] In step S320, the final fluid response parameter function f(FRP_norm) is output for use in Figure 8 In step S350, the first PCSV_uncal is improved based on the fluid response parameter function f(FRP_norm). In step S390, the second PCSV_imp is output (eg, on a display of the medical device 330), or the second PCSV_imp is used to control an automatic medication dispenser or an infusion pump connected to or included in the medical device 330.

[0251] The basic formula 29 used to calculate the second PCSV_imp is:

[0252] (31)PCSV_imp(ml)=PCSV(ml)·f(FRP_norm)

[0253] In the following, reference will be made to Figure 14-20 To illustrate the HLI mechanism in spontaneously breathing SB and the normalization of fluid responsiveness parameters in spontaneous breathing.

[0254] Inspiration occurs during spontaneous breathing because the intercostal respiratory muscles and diaphragm contract and increase the volume of the thoracic cavity, causing negative intrathoracic pressure ITP and negative intrapulmonary pressure. This air or any other ventilation gas mixture flows into the lungs because the atmospheric pressure surrounding the individual is higher than the pressure within the airways of the lungs. At the same time, due to the negative ITP of inspiration, venous blood flow from the extrathoracic chambers to the IBCS (e.g., vena cava, right atrium) increases, so the right ventricular preload increases, and therefore the right ventricular stroke volume increases.

[0255] The increased right ventricular stroke volume travels through the lungs beat by beat, thereby increasing the left ventricular end-diastolic volume and stroke volume. The increase in left ventricular stroke volume can be detected in the aorta and arterial system by the accompanying increase in systolic pressure and pulse pressure. In normal exhalation, this HLI mechanism is reversed, the respiratory muscles and diaphragm relax, and the volume of the thoracic cavity is passively reduced in normal breathing, which is mainly due to elastic recoil when the inspiratory muscles relax. The ITP and the subsequent acquisition of positive pressure in the distal airways will cause exhalation towards a lower atmospheric pressure. During forced exhalation, the other internal intercostal muscles and internal abdominal muscles will contract and create a significantly higher positive ITP.

[0256] Both normal and forced exhalation cause an increase in ITP, compression of the IBCS, a decrease in venous return, a decrease in right ventricular preload and a decrease in stroke volume, which in turn can be detected in the aorta and arterial system by, for example, a decrease in systolic and pulse pressures.

[0257] Since all of these conditions affect the fluid response parameters, normalization is required.

[0258] Similar to the situation during mechanical ventilation, the lung itself cannot be standardized as a test tool without taking into account the different static lung compliance Cls in healthy and diseased lungs.

[0259] A stiff lung with reduced Cls (fibrosis, etc.) will require more negative ITP to achieve the same level of inflation than a lung with normal Cls. During spontaneous inspiration, the more negative the ITP, the greater the venous return, the greater the stroke volume, and the greater the final FRP value.

[0260] Unlike the situation in MV, the effect of Ccw on ITP and thus on FRP can be neglected = not applicable in SB.

[0261] Normally, neither ITP changes nor tidal volume can be measured in SB without causing stress to the patient or individual. However, it is also necessary to normalize the FRP changes caused by tidal volume and take into account static lung compliance. At the same level of physical activity and environmental conditions, alveolar ventilation remains constant over time. Any increase in RR will cause a decrease in tidal volume, while a decrease in breathing rate RR will cause a higher tidal volume. In this case, breathing rate RR or a corresponding variable is used for FRP normalization, because it is related to TV and therefore to FRP. Under normal conditions and most pathological conditions, the RR of a spontaneously breathing adult individual varies in the range of 6…40 / min. In the case of constant IBCS, a larger RR will cause a lower TV value and thus a smaller FRP value, while a lower RR will cause a larger TV value and thus a larger FRP value.

[0262] During SB, the normalization function f_FRP_SB used to normalize a single FRP or a combined FRP is any linear or nonlinear function that eliminates the RR dependence and the Cls dependence of the FRP. Any measured FRP value of an individual at a given RR and lung compliance Cls (an example of which is given above) is normalized to the RR and Cls by applying the normalization function f_FRP_SB.

[0263] Therefore, depending on whether tidal volume TV is available (which is rarely the case), different approaches can be used to determine the FRP normalization function f_FRP_SB, which is applied to a single FRP or a combined FRP to achieve a normalized FRP_norm, as shown in Equation 18 above. Fig.14In step S410, it is checked whether the tidal volume TV is available. If the tidal volume TV is not available, the FRP normalization function f_FRP_SB is determined based on the approximation function f_FRP(RR) in step S420. If TV is available, the FRP normalization function f_FRP_SB is determined based on the tidal volume in step S430.

[0264] First, determination of the FRP normalization function f_FRP_SB_RR based on RR will be described in detail.

[0265] The basic form of the FRP normalization function f_FRP_SB_RR can be described as:

[0266] (32)f_FRP_SB_RR=RR(1 / min) / RR0(1 / min)

[0267] If RR is greater than the default value RR0 (eg, 12 / min), f_FRP_SB_RR must increase the FRP value; if RR is less than RR0, f_FRP_SB_RR must decrease the FRP value.

[0268] The RR dependence of f_FRP_SB_RR can be obtained from statistical analysis of FRP measurements and simultaneous RR measurements in healthy isovolumetric individuals.

[0269] F_FRP_SB_RR can take the following forms:

[0270] (33)f_FRP_SB_RR=f_FRP(RR0) / f_FRP(RR)

[0271] Wherein, f_FRP(RR) is a respiratory rate correction function, preferably an exponential function, preferably having the following form:

[0272] (34)f_FRP(RR)=a·exp(-b(min)·RR(1 / min))

[0273] For example, for adults, a = 5 ... 20%, b = 0.02 ... 0.10 min

[0274] Fig.15 An example of a respiratory rate correction function f_FRP(RR) is provided in , which shows that as RR increases, PPV decreases.

[0275] Fig.15The results of a survey of 20 adult healthy isovolumetric (=normal volume) volunteers who breathed 6 to 30 times per minute according to the beat of a metronome (n=170 measurements in total) are shown. Non-invasive PPV was measured as FRP. The squares represent the mean (± standard deviation) of the PPV. The curve shown is an exponential approximation function f_FRP (RR) based on these data points.

[0276] in addition, Fig.15 It is shown that, compared with the default value TVnorm0, the function f_FRP(RR) increases as TVnorm increases, and decreases as TVnorm decreases.

[0277] Fig.16 The formula (33) and (34) are shown in the default lung compliance Cls0 according to Fig.15 f_FRP(RR)( Fig.16 Example of the FRP normalization function f_FRP_SB_RR for adults (solid line) derived from the example of the FRP normalization function f_FRP_SB_RR for adults (dashed line in FIG).

[0278] The FRP normalization function f_FRP_SB_RR eliminates the effect of RR on FRP by clamping FRP to the value at RR0 = 12 / min (here, FRP_norm = 7.61%). The f_FRP_SB_RR value at RR0 is 1, so the FRP value at RR = RR0 remains unchanged. By multiplying with f_FRP_SB_RR, the FRP measured at RR values ​​less than RR0 can be reduced and the FRP measured at RR values ​​greater than RR0 can be increased.

[0279] It is noted that several pathological conditions can cause an increase or decrease in TV. An increase in TV can cause an increase in respiratory-induced ITP changes. Subsequently, FRP increases while the patient's blood volume remains unchanged. This pathologically induced increase in FRP must be corrected.

[0280] To further improve the FRP correction, input data is needed to estimate whether the TV is reduced or increased and to which level.

[0281] The preferred quantitative levels of TV requiring input from a physician or medical staff are:

[0282] (a) Moderate decrease (↓)

[0283] (b) Normal (-)

[0284] (c) Moderate increase (↑)

[0285] (d) Large increase (↑↑)

[0286] (e) Very large increase (↑↑↑)

[0287] But it is not limited to these.

[0288] Based on the input values, the FRP normalization function f_FRP_SB_RR is adjusted depending on the direction and level of TV. An increase in TV requires adjusting f_FRP_SB_RR to a lower level, while a decrease in tidal volume requires adjusting f_FRP_SB_RR to a higher level. For an illustration of these levels, refer to Fig.17 .

[0289] The determination of the FRP normalization function in step S430 when the tidal volume TV is available will now be described.

[0290] If TV is available, FRP can be normalized with TVnorm normalized by TV to FFM in a similar way as in MV by applying the function f_FRP_SB_TV.

[0291] (35)FRP_norm(%)=FRP(%)·f_FRP_SB_TV

[0292] The FRP can be normalized using the relationship between TVnorm and a default value TVnorm0 of, for example, 8 ml / kg FFM for adults.

[0293] (36)f_FRP_SB_TV=TVnorm0(ml / kg) / TVnorm(ml / kg)

[0294] Preferably, the normalization algorithm can be refined using additional dimensionless coefficients coeff_MV and dimensionless constants const_MV.

[0295] (37)f_FRP_SB_TV=TVnorm0(ml / kg) / TVnorm(ml / kg)·coeff_SB_TV+const_SB_TV

[0296] In addition, the FRP normalization function f_FRP_SB (which may depend on RR (f_FRP_SB_RR) or TVnorm (f_FRP_SB_TV)) is adjusted for different conditions of static lung compliance Cls. Input data is also required to provide information about whether Cls is normal or the degree of Cls increase when Cls increases. This is performed in step S440.

[0297] The preferred level of input data entered by doctors or medical personnel is:

[0298] (a)Normal(-)

[0299] (b) Moderate decrease (↓)

[0300] (c) Large decrease (↓↓)

[0301] (d) Very large decrease (↓↓↓)

[0302] But it is not limited to these.

[0303] In addition, refer to Fig.18 and Fig.19 , Fig.18 and Fig.19 Different cases of the effect of Cls on HLI during SB are illustrated, where lower Cls is shown as darker lungs and larger / lower HLI is shown as larger / smaller distances between vertical dashed lines. Thus, the lungs shown on the left have normal Cls, while the lungs shown on the right have strongly reduced Cls.

[0304] The FRP normalization function f_FRP_SB needs to be adjusted depending on the direction and level of Cls. Thus, a reduced Cls requires adjusting f_FRP_SB to a lower level. A greatly reduced Cls requires a greatly reduced f_FRP_SB.

[0305] Further references Fig. 20 , Fig. 20 An example of the FRP normalization function f_FRP_SB_RR at Cls=Cls0 and Cls / Cls0=0.5 using equations (33) and (34) is shown. The value of f_FRP_SB_RR at RR0 is equal to the relationship Cls / Cls0.

[0306] After the FRP normalization function f_FRP_SB is adjusted, the adjusted f_FRP_SB is applied to a single FRP or a combined FRP in step S450 . Here, a normalized FRP_norm for SB is implemented.

[0307] Similar to MV, the adjustment coefficients a, b, c and d are applied to FRP_norm in step S460 to finally receive the FRP function f(FRP_norm). The FRP function f(FRP_norm) is output in step S470 to be applied to PCSV_uncal to finally obtain PCSV_imp defined in Formula 29, which is again provided here:

[0308] (29)PCSV_imp(ml)=PCSV(ml)·f(FRP_norm)

[0309] Summarizing the second embodiment, the stroke volume of the heart of an individual responsive to a fluid depends on the hemodynamic filling state. A lower filling state indicated by a larger FRP will cause a lower stroke volume and vice versa. Thus, by considering the actual FRP value, the PCSV calculation can be improved. To compensate for / remove the parameter(s) affecting FRP other than the level of IBCS to achieve this PCSV improvement, the present invention proposes using a normalized FRP_norm that is normalized using TVnorm and Ccw in the MV or a normalized FRP_norm that is normalized using RR, TVnorm, and Cls in the SB.

[0310] In the following Figures 21 to 34 are shown data of non-invasive high-resolution oscillometric measurements of a patient during a high-risk surgical procedure and simultaneously occurring invasive measurements.

[0311] Referring to Fig.21 is shown an example of a possible form of f(FRP_norm), which is derived from a logarithmic fit function ln_fit_fct that depends on SVtd_FFM_norm (ml / kgFFM) from FRP_norm, where SVtd_FFM_norm is the FFM portion of SVtd normalized to FFM.

[0312] The logarithmic fit function ln_fit_fct is based on a statistical analysis of data of 22 adult patients with a change in FRP_norm > 5%.

[0313] f(FRP_norm) is derived from ln_fit_fct by normalization with the value of ln_fit_fct at the default value of FRP_norm (here 1.2 ml / kgFFM at FRP_norm0 = 13.2%).

[0314] Furthermore, any multiplication of f(FRP_norm) with a given PCSV will increase the PCSV at FRP_norm < FRP_norm0 and decrease the PCSV at FRP_norm > FRP_norm0. Thus, it is an amplification function or attenuation function based on the normalized FRP value.

[0315] Nonetheless, the method of the present invention provides high precision for the derived value of PCSV, resulting in a more reliable diagnosis and better and potentially life-saving responses by a doctor or medical personnel during a surgical procedure.

[0316] In Figures 22 to 30In FIG. 1 , different regression graphs are illustrated, these data values ​​are based on measurements and determinations according to conventional methods and / or the method of the present invention to show the differences in their accuracy.

[0317] exist Figures 22 to 33 In FIG. 1 , different regression plots are given showing examples of comparison of biologically calibrated non-invasive PCSV calculations in 37 patients with corresponding SVref simultaneously obtained from transpulmonary thermodilution measurements. The PCSV values ​​are based on measurements and determinations according to conventional methods and / or the methods of the present invention to show the differences in their accuracy. Wherein SD denotes standard deviation, r denotes Pearson correlation coefficient, CR denotes consistency rate, PE denotes percent error, pt denotes patient, and n denotes the number of measurements.

[0318] Fig. 22 A regression plot illustrating the results of a method for determining PCSV based on a conventional algorithm for providing PCSV_uncal (Wesseling algorithm by Chen et al. in "The Effect of Signal Quality on Six Cardiac Output Estimators" (Comput Cardiol, January 1, 2009)). The Wesseling algorithm is an algorithm for providing a dimensionless PCSV_uncal and therefore requires calibration to the individual based on the individual's biological characteristics. As described above, conventional biological calibration of PCSV_uncal is performed based on BSA. In addition, FRP-based PCSV adjustments have not been applied. Fig. 22 The low accuracy and precision of the comparison of PCSV to SVref can be identified in Figure 5, as indicated by the high SD and PE values ​​and the low correlation coefficient.

[0319] Fig.23 A regression plot illustrating the results of a method for determining PCSV based on an alternative algorithm for providing PCSV_uncal is shown. Again, here, the conventional biological calibration parameter BSACal based on BSA is applied without applying the FRP-based PCSV adjustment. Fig.23 Only the algorithm used to export PCSV_uncal is different. Figure 1b As explained, in general, PCSV_uncal is dimensionless and is derived from any arterial blood pressure waveform that can be obtained invasively or non-invasively. Based on the parameters so derived, the individual's blood pressure curve is identified and several different parameters are taken into account for forming PCSV_uncal. Fig. 22In comparison, it can be monitored that the alternative PCSV_uncal algorithm produces higher accuracy and precision than the conventional algorithm. This shows that different algorithms can be used to derive the starting value PCSV_uncal required for the present invention.

[0320] Fig.24 and Fig.25 are regression plots (four-quadrant plots) that illustrate the Fig. 22 and Fig.23 Comparison of the change ΔPCSV in PCSV with the change ΔSVref of the reference thermodilution SVref for the same PCSV_uncal algorithm, ΔPCSV being related to the corresponding initial PCSV value and ΔSVref being related to the corresponding initial SVref value.

[0321] Fig.24 A regression plot of ΔPCSV and ΔSVref values ​​based on a conventional algorithm for deriving PCSV_uncal is shown. Additionally, BSACal is applied without considering FRP.

[0322] Fig.25 Regression plots of ΔPCSV and ΔSVref values ​​based on alternative algorithms for deriving PCSV_uncal are shown. Additionally, BSACal is applied without considering FRP. Fig.24 In comparison, applying the alternative algorithm to derive PCSV_uncal has improved the accuracy and precision of the comparison of ΔPCSV to ΔSVref, as indicated by lower SD values ​​and lower PE values ​​as well as higher correlation coefficient and higher CR.

[0323] Fig.26 A regression plot illustrating the results of a method for determining PCSV based on a conventional Wesseling algorithm for providing PCSV_uncal is shown. PCSV_uncal is biologically calibrated based on the perfusion parameter BioCal of the present invention. Again, no FRP-based PCSV adjustment is applied. Fig. 22 The accuracy and precision of the comparison of PCSV with SVref were significantly improved, as indicated by the lower SD and PE values ​​and the higher correlation coefficient. Figure 22 to Figure 26 The only change was the application of the perfusion parameter BioCal instead of BSACal, which improved the determination of PCSV.

[0324] Fig. 27 A regression plot illustrating the results of a method for determining PCSV based on an alternative algorithm for providing PCSV_uncal is shown. PCSV_uncal is biologically calibrated based on the perfusion parameter BioCal of the present invention. Again, no FRP-based PCSV adjustment is applied. Fig.23 and Fig.26 The accuracy and precision of the comparison of PCSV with SVref were significantly improved, as indicated by the lower SD and PE values ​​and the higher correlation coefficient. Figure 23 to Figure 27 The only change was the application of the perfusion parameter BioCal instead of BSACal, which improved the determination of PCSV.

[0325] Fig.28 and Fig.29 are regression plots (four-quadrant plots) that illustrate the Fig.26 and Fig. 27 Comparison of the change ΔPCSV in PCSV with the change ΔSVref of the reference thermodilution SVref for the same PCSV_uncal algorithm, ΔPCSV being related to the corresponding initial PCSV value and ΔSVref being related to the corresponding initial SVref value.

[0326] Fig.28 A regression plot of ΔPCSV values ​​and ΔSVref values ​​based on a conventional algorithm for deriving PCSV_uncal is shown. In addition, the perfusion parameter BioCal of the present invention was applied without considering FRP. Fig.24 The accuracy and precision of the comparison of ΔPCSV to ΔSVref were significantly improved, as indicated by the lower SD and PE values ​​as well as the higher correlation coefficient and CR. Figure 24 to Figure 28 The only change applied was the application of the perfusion parameter BioCal of the present invention instead of BSACal, thereby improving the ability to track true changes in PCSV.

[0327] Fig.29 Regression plots of ΔPCSV and ΔSVref values ​​based on alternative algorithms for deriving PCSV_uncal are shown. Additionally, BSACal is applied without considering FRP. Fig.25 The accuracy and precision of the comparison of ΔPCSV to ΔSVref were significantly improved, as indicated by the lower SD and PE values ​​as well as the higher correlation coefficient and CR. Figure 25 to Figure 29 The only change applied was the application of the perfusion parameter BioCal of the present invention instead of BSACal, thereby improving the ability to track true changes in PCSV.

[0328] Figure 30 to Figure 33 The effect of the FRP function f(FRP_norm) on the sensitivity of the pulse contour algorithm to track the true PCSV changes by taking FRP_norm into account is depicted.

[0329] Fig.30A regression plot illustrating the results of a method for determining PCSV based on a conventional Wesseling algorithm for providing PCSV_uncal is shown. PCSV_uncal is biologically calibrated based on the perfusion parameter BioCal of the present invention. In addition, an adjustment of PCSV based on f(FRP_norm) has been applied to better track the true changes. Fig.26 The accuracy and precision of the comparison of PCSV with SVref were improved as indicated by the lower SD and PE values ​​and the higher correlation coefficient. Figure 26 to Figure 30 The only change in is the application of f(FRP_norm), which further improves the determination of PCSV.

[0330] Fig.31 A regression plot illustrating the results of a method for determining PCSV based on an alternative algorithm for providing PCSV_uncal is shown. PCSV_uncal is biologically calibrated based on the perfusion parameter BioCal of the present invention. In addition, an adjustment of PCSV based on f(FRP_norm) has been applied to better track the true changes in PCSV. Fig. 27 The accuracy and precision of the comparison between PCSV and SVref are improved and the Fig.30 The accuracy and precision of the comparison of PCSV to SVref were significantly improved, as indicated by the lower SD and PE values ​​and the higher correlation coefficient. Figure 27 to Figure 31 The only change is the application of f(FRP_norm), which is the same as Fig. 22 , Fig.23 , Fig.26 , Fig. 27 , Fig.30 This further improves the determination of PCSV and produces the best accuracy and precision in comparison.

[0331] Fig.32 and Fig.33 are regression plots (four-quadrant plots) that illustrate the Fig.30 and Fig.31 Comparison of the change ΔPCSV in PCSV with the change ΔSVref of the reference thermodilution SVref for the same PCSV_uncal algorithm, ΔPCSV being related to the corresponding initial PCSV value and ΔSVref being related to the corresponding initial SVref value.

[0332] Fig.32Regression plots of ΔPCSV values ​​and ΔSVref values ​​based on the conventional algorithm (Wesseling) used to derive PCSV_uncal are shown. In addition, the perfusion parameter BioCal was applied. A PCSV adjustment based on f(FRP_norm) has been applied to better track the true changes. Fig.28 The accuracy and precision of the comparison of ΔPCSV to ΔSVref were significantly improved, as indicated by the lower SD and PE values ​​as well as the higher correlation coefficient and CR. Figure 28 to Figure 32 The only change in is the application of f(FRP_norm), which further improves the ability to track the true changes in PCSV.

[0333] Fig.33 Regression plots of ΔPCSV values ​​and ΔSVref values ​​based on an alternative algorithm for deriving PCSV_uncal are shown. In addition, the perfusion parameter BioCal of the present invention has been applied. A PCSV adjustment based on f(FRP_norm) has been applied to better track the true changes. Fig.29 and 32 The accuracy and precision of the comparison of ΔPCSV to ΔSVref were significantly improved, as indicated by the lower SD and PE values ​​as well as the higher correlation coefficient and CR. Figure 29 to Figure 33 The only change is the application of f(FRP_norm, which is the same as Fig.24 , Fig.25 , Fig.28 , Fig.29 , Fig.32 In comparison, the ability to track the true changes of PCSV is further improved and produces the best accuracy and precision.

[0334] Therefore, when Fig.31 and Fig.33 When the results shown are compared with any other comparable regression plots, it is clear that the application of the perfusion parameter BioCal of the present invention and the application of the f(FRP_norm) of the present invention describing the cardiopulmonary interaction of an individual greatly improves the accuracy of the second PCSV_imp used as the basis for diagnosis and management of the individual.

[0335] Fig.34The relative errors of a second PCSV_imp determined according to the invention and a conventional PCSV_conv are illustrated, each relative error being related to a reference stroke volume SVref. The measurement results of four different exemplary patients are illustrated. The relative error for the PCSV_imp determined using an alternative algorithm for deriving PCSV_uncal based on the invention and applying BioCal and f(FRP_norm) is calculated as (PCSV_imp-SVref) / SVref, while the relative error for the PCSV_conv determined based on an alternative algorithm for deriving PCSV_uncal but without applying BioCal and f(FRP_norm) is calculated as (PCSV_conv-SVref) / SVref. It can be easily appreciated that the dotted line for the PCSV_imp according to the invention is closer to the 0% line than the dotted line for the PCSV of the prior art. This applies to all four different patients.

[0336] Fig.35 An exemplary apparatus according to an embodiment of the present invention is illustrated. The apparatus basically comprises: a patient monitor 310 and a blood pressure and waveform processing device (i.e., an arterial blood pressure and waveform monitor) that perform the functions of arterial blood pressure and waveform measurement, a controller 320 for adjusting the derived or processed PCSV_uncal, and a medical device 330. In addition, there is a parameter unit 340 that provides a perfusion parameter BioCal and / or a FRP_norm parameter. The parameter unit 340 may be integrated in the controller 320 or may be implemented as a single unit. A patient 350 is connected to the arterial blood pressure and waveform monitor 310. It will be appreciated by those skilled in the art that Fig.35 The units shown in: controller 320 and parameter unit 340 may be implemented as functional parts of a single device (eg, patient monitor 310).

[0337] In simple form, a blood pressure cuff or invasive blood pressure measurement device (not shown) is connected to the patient 350 to provide a plurality of arterial blood pressures and corresponding waveforms of the patient based on invasively or non-invasively measured blood pressures.

[0338] like Figure 1b As shown, based on the measured blood pressure and the resulting waveform, the patient monitor 310 extracts a predetermined portion of the pulse wave. Figure 1b The blood pressure curve shown in Fig.35 The blood pressure waveform is extracted from the blood pressure waveform measured by a blood pressure cuff or an invasive blood pressure measurement device (not shown). The patient monitor 310 (the corresponding processor of the monitor) receives the blood pressure waveform and can process the blood pressure waveform by filtering, calibration or linearization to extract the blood pressure curve, such as Figure 1bAs shown. Based on this extracted blood pressure curve, an area Asys (which is called systolic pressure area Asys) is determined. During the systolic period, the blood pressure curve encloses an area Asys. This area Asys is used as a value for the first PCSV_uncal, which is improved by using the method of the invention operated by the functional unit (or device) of the invention for adjusting the first PCSV_uncal based on perfusion parameters and / or based on the fluid response parameter FRP_norm.

[0339] The second PCSV_imp is output by the controller 320 and provided to the medical device 330, which in its simplest form may be in ml and / or ml / kg FFM and / or in ml / m 2 The BSA indicates the display (eg, display) of the second PCSV_imp.

[0340] In further advanced embodiments, a second PCSV_imp may be used to control an automatic fluid dosing device and / or an infusion pump, which may be included or coupled to the medical device 330 and is not illustrated in detail here.

[0341] By controlling the infusion pump, for example, based on the second PCSV_imp, it may be possible to more reliably treat a patient who may be located in an intensive care station.

[0342] Furthermore, doctors or medical personnel use the second PCSV_imp to provide more accurate diagnosis and treatment management for patients.

[0343] The parameter BioCal for perfusion or the parameter FRP_norm for fluid response is derived based on measurements of characteristics of the patient 350 which were acquired in advance (and stored in, for example, a medical database) or derived during the process of adjusting the first PCSV_uncal, which is described in more detail above.

[0344] Thus, the patient monitor can be enabled to determine the stroke volume (SV) of the individual 350 by including a processor arranged to perform the following functions:

[0345] (1) Providing a first pulse contour stroke volume (PCSV_uncal) based on one or more characteristics of a received measured arterial blood pressure waveform, or providing a conventionally derived pulse contour stroke volume (PCSV_conv).

[0346] This can be accomplished by receiving a pressure waveform from a blood pressure cuff or invasive blood pressure measurement device and performing the above analysis.

[0347] (2) determining at least one perfusion parameter (BioCal) describing perfusion through the body's fat-free mass and fat mass of the individual, and / or

[0348] determining at least one fluid response parameter function f(FRP_norm) that depends on a fluid response parameter (FRP_norm) describing a cardiopulmonary interaction of the individual;

[0349] At this stage, the patient monitor (eg, its processor via parameter unit 340 ) retrieves appropriate patient medical data required for the analysis from the medical database; and / or the physician manually inputs appropriate patient medical data required for the analysis during the process of adjusting the first PCSV_uncal.

[0350] (3) adjusting the first pulse contour stroke volume (PCSV_uncal) based on at least one of the perfusion parameter (BioCal) and / or the fluid response parameter function f(FRP_norm), or

[0351] The conventionally derived pulse contour stroke volume (PCSV_conv) is adjusted based on the fluid response parameter function f(FRP_norm) to provide a second pulse contour stroke volume (PCSV_imp).

[0352] The second pulse contour stroke volume (PCSV_imp) can be displayed on a display of a patient monitor; PCSV_imp can further be input to other medical devices arranged to provide / control therapy to the patient.

[0353] The processor of the patient monitor can be arranged to perform the following Figure 1 , Figure 2 , Figure 8 , Fig. 9 , Fig.10 and Fig.14 Any of the steps in the above flowcharts shown.

Claims

1. A method for determining a stroke volume of an individual, comprising the steps of: providing a first pulse contour stroke volume based on one or more characteristics of a measured arterial blood pressure waveform or providing a pulse contour stroke volume derived based on medical data obtained from an individual, wherein the first pulse contour stroke volume is uncalibrated and dimensionless; and the derived pulse contour stroke volume is calibrated for biological characteristics of the individual and dimensioned, determining at least one perfusion parameter describing perfusion through the fat-free mass and fat mass of the body of the individual, and / or determining at least one fluid response parameter function f that depends on a fluid response parameter describing cardiopulmonary interaction of the individual, and adjusting the first pulse contour stroke volume to provide a second pulse contour stroke volume based on at least one of the perfusion parameter and / or the fluid response parameter function f, or The derived pulse contour stroke volume is adjusted based on the fluid response parameter function f to provide a second pulse contour stroke volume.

2. The method according to claim 1, wherein: Providing the first pulse contour stroke volume and / or the derived pulse contour stroke volume is based on a non-invasive or invasive measurement of an arterial blood pressure waveform.

3. The method according to claim 1 or 2, wherein: The step of determining a fluid response parameter is based on at least one of: pulse pressure variation, mean left ventricular ejection pressure variation, stroke volume variation, systolic pressure variation, left ventricular systolic area variation, photoplethysmographic variation index, or any other fluid response parameter or combination thereof having appropriate sensitivity and specificity.

4. The method according to claim 1 or 2, wherein: The fluid responsiveness parameter is determined by applying a fluid responsiveness normalization function for mechanical ventilation, or The fluid responsiveness parameter is determined by applying a fluid responsiveness normalization function for spontaneous breathing.

5. The method according to claim 1 or 2, wherein: The fluid responsiveness parameter is determined by taking into account compliance of the individual's respiratory system, wherein the compliance of the respiratory system includes lung compliance and chest wall compliance.

6. The method according to claim 1 or 2, wherein: The fluid responsiveness parameter is normalized using a fluid responsiveness parameter normalization parameter, wherein the fluid responsiveness parameter normalization parameter is normalized to the fat free mass, predicted weight, or weight.

7. The method according to claim 5, wherein: The fluid responsiveness parameters are adjusted for mechanical ventilation by using predetermined semi-quantitative information about the chest wall compliance.

8. The method according to claim 5, wherein: The fluid responsiveness parameters were normalized to spontaneous breathing using the lung compliance and respiratory rate.

9. The method according to claim 1 or 2, wherein: The step of determining perfusion parameters describing the perfusion through the fat-free mass and the fat mass of the individual comprises: Determine the pulse contour stroke volume correction factor related to fat free mass, and / or Determine the pulse contour stroke volume correction factor related to fat mass.

10. The method of claim 9, further comprising determining individual demographic parameters, including: Determining the fat-free mass, determining said fat mass, the height of the individual, the biological sex of the individual, and The age of the individual.

11. The method according to claim 10, wherein: The biological sex of the individual includes male, female, and transgender.

12. The method according to claim 9, wherein: The pulse contour stroke volume calibration factor associated with fat-free mass is determined by using a perfusion coefficient associated with fat-free mass perfusion and / or the pulse contour stroke volume calibration factor associated with fat mass is determined by using a perfusion coefficient associated with fat mass perfusion.

13. The method according to claim 9, wherein: In order to determine the fat-free mass-dependent pulse contour stroke volume correction factor and / or the fat-free mass-dependent pulse contour stroke volume correction factor, the fat-free mass layer thickness and the fat mass layer thickness are taken into account.

14. A patient monitor for determining a cardiac stroke volume of an individual (350), the monitor comprising a processor arranged to: providing a first pulse contour stroke volume based on one or more characteristics of the received measured arterial blood pressure waveform or providing a pulse contour stroke volume derived based on medical data obtained from the individual, wherein The first pulse contour stroke volume is uncalibrated and dimensionless; and the derived pulse contour stroke volume is calibrated for the individual with respect to biological characteristics and dimensioned, determining at least one perfusion parameter describing perfusion through the fat-free mass and fat mass of the body of the individual, and / or determining at least one fluid response parameter function f that depends on a fluid response parameter describing cardiopulmonary interaction of the individual, and adjusting the first pulse contour stroke volume to provide a second pulse contour stroke volume based on at least one of the perfusion parameter and / or the fluid response parameter function f, or The derived pulse contour stroke volume is adjusted based on the fluid response parameter function f to provide a second pulse contour stroke volume.

15. A patient monitor according to claim 14, the processor of the patient monitor being further arranged to perform the steps of the method according to any one of claims 2 to 13.

16. A patient monitor according to claim 14 or 15, further comprising a medical device (330) arranged to receive the second pulse contour stroke volume and to output the second pulse contour stroke volume or related data derived therefrom.

17. An apparatus for determining a stroke volume of an individual, comprising: An arterial blood pressure and waveform monitor (310) for providing a first pulse contour stroke volume or a pulse contour stroke volume derived based on medical data obtained from an individual (350) connected to the arterial blood pressure and waveform monitor (310), wherein the first pulse contour stroke volume is uncalibrated and dimensionless; and the derived pulse contour stroke volume is calibrated for biological characteristics of the individual and dimensioned, a parameter unit (340) for providing perfusion parameters and / or cardiopulmonary interaction description parameters of the individual (350), A controller (320) is used to adjusting the first pulse contour stroke volume based on the perfusion parameter and / or based on the cardiopulmonary interaction description parameter to provide a second pulse contour stroke volume; or The derived pulse contour stroke volume is adjusted based on the cardiopulmonary interaction description parameters to provide a second pulse contour stroke volume.

18. The apparatus according to claim 17, further comprising: A medical device (330) is arranged to receive the second pulse contour stroke volume and output the second pulse contour stroke volume or related data derived from the second pulse contour stroke volume.

Images