Reconstruction of patient-specific central arterial pressure waveform morphology using non-invasive distal blood pressure measurements

Abstract

A method, computer readable medium, and system are disclosed for reconstructing patient-specific central arterial pressure waveform morphology using non-invasive distal blood pressure measurements and non-invasive recordings of distal blood pressure waveforms. Measurements of the patient's systolic blood pressure, diastolic blood pressure, and heart rate are made at a distal location (e.g., radial artery). The invention uses patient-specific demographic and health data, such as gender, age, and / or current medications. The patient-specific central arterial pressure waveform morphology is calculated using a Windkessel-type lumped parameter multiple compartment model. The method does not assume structural rigor of the transfer relationship, but rather evolutionary laws that provide the relationship between distal and proximal blood pressure. The invention provides central arterial blood pressure values ​​and proximal blood flow that are useful for diagnosing elevated cardiac blood pressure, hypertension, or both. The method has been validated by clinical trials. The method of the invention reproduced values ​​obtained using non-invasive methods from clinical trials.

Description

[Technical field]

[0001] The present invention relates to estimating human cardiac parameters for subsequent analysis and diagnosis. In particular, the present invention relates to reconstructing central arterial pressure waveform morphology from continuous non-invasive blood pressure measurements. The reconstructed waveform morphology can be used in the diagnosis and treatment of elevated blood pressure and / or hypertension. [Background technology]

[0002] The human heart pumps approximately 100 milliliters of blood into the circulatory system during each cardiac cycle. Most of the cardiac output is stored in elastic arteries and delivered to target sites and organs during cardiac diastole. This phenomenon creates a kind of delicate dynamic equilibrium, which, when disturbed, can lead to many serious pathologies, including persistent hypertension. Elevated blood pressure is associated with hypertension. There are many definitions of hypertension, which can be found in the literature. One definition, from the Centers for Disease Control and Prevention, states that hypertension is a systolic blood pressure of 130 mmHg or higher, or a diastolic blood pressure of 80 mmHg or higher, or a current medication being used to lower hypertension (see Ostchega Y(2020). Hypertension Prevalence Among Adults Aged 18 and Over: United States, 2017-2018. NCHS data brief,(364),1-8.). According to the World Health Organization (WHO) guidelines (2021), hypertension (or elevated blood pressure) is a serious medical condition that significantly increases the risk of heart disease, brain disease, kidney disease, and other diseases. Hypertension is diagnosed using specific systolic and diastolic blood pressure values ​​or by the use of antihypertensive drugs (see World Health Organization (2021) Guidelines for the Pharmacological Treatment of Hypertension in Adults). A different definition is given in the American College of Cardiology / American Heart Association report:This report distinguishes between two stages of hypertension: stage 1 is when the systolic blood pressure is within the range of 130-139 mmHg or the diastolic blood pressure is within the range of 80-89 mmHg, and stage 2 is when the systolic blood pressure is 140 mmHg or more or the diastolic blood pressure is 90 mmHg or more (Whelton PK (2018) 2017 ACC / AHA / AAPA / ABC / ACPM / AGS / APhA / ASH / ASPC / NMA / PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: American College of Cardiology / American Heart Association Task Force on Clinical Practice Guidelines. Hypertension, 71: e13-e115). The European Society of Cardiology and the European Society of Hypertension provide a similar explanation, defining hypertension as a systolic blood pressure value of more than 140 mmHg and / or a diastolic blood pressure value of more than 90 mmHg in the office. The same classification is used for young, middle-aged and elderly people, but for children and teenagers where intervention trial data are not available, blood pressure percentiles are used (see Williams B(2018)2018 ESC / ESH Guidelines for the Management of Arterial Hypertension.European heart journal,39(33),3021-3104).

[0003] The prevalence of hypertension is related to genetic (e.g., polygenic influences) and environmental factors (e.g., diet, physical activity, sodium and potassium intake, and / or alcohol intake). Secondary hypertension is commonly caused by renal parenchymal disease, renovascular disease, primary aldosteronism, obstructive sleep apnea, and drug or alcohol use (see Whelton PK (2018) above).

[0004] Hypertension can cause left ventricular hypertrophy and coronary artery disease (CAD). Left ventricular hypertrophy is caused by pressure overload, which results in an increase in muscle mass and wall thickness without an increase in ventricular volume. The result is impaired diastolic function, slower ventricular relaxation, and delayed filling. Left ventricular hypertrophy is an independent risk factor for cardiovascular disease and can therefore cause sudden death (see Aronow WS (2017) Hypertension and left ventricular hypertrophy. Annals of Translational Medicine, 5 (15), 310). There is no threshold of blood pressure at which complications begin to develop. The impact of hypertension is determined by the severity of the condition. Increased blood pressure leads to increased morbidity at all blood pressure ranges. Chronic arterial hypertension accelerates coronary artery disease (CAD), causing myocardial ischemia and myocardial infarction, and is an important risk factor for death from coronary artery disease (CAD). Chronic pressure overload causes heart failure, which begins as diastolic dysfunction and progresses to overt systolic dysfunction with cardiac congestion. The most serious consequences of hypertension are thrombosis, thromboembolism, or stroke resulting from intracranial hemorrhage. Hypertension causes a slow progression of kidney disease, which initially manifests as microalbuminemia and becomes evident over time (see Foex P (2004) Hypertension: Pathophysiology and Treatment. Continuing Education in Anaesthesia Critical Care & Pain, 4 (3), 71).

[0005] Hypertension also appears to be associated with common non-cardiovascular diseases such as dementia, cancer, osteoporosis and oral diseases (see Kokubo Y(2015)Higher blood pressure as a risk factor for diseases other than stroke and ischemic heart disease.Hypertension,66(2),254-259). Hypertension increases the risk of atrial fibrillation (a type of chronic arrhythmia) (see Benjamin EJ (1994) Independent risk factors for atrial fibrillation in a population-based cohort. Framingham Heart Study. JAMA, 271(11), 840-844) and contributes to reduced glomerular filtration rate and progression of chronic kidney disease (see Buckalew VM (1996) Prevalence of hypertension in 1,795 subjects with chronic renal disease: the modification of diet in renal disease study baseline cohort. Modification of Diet in Renal Disease Study Group. American journal of kidney diseases: the official journal of the National Kidney Foundation, 28(6), 811-821). High blood pressure may lead to changes in blood flow, changes in the integrity of the blood-brain barrier, or brain changes in dementia (see Moretti R (2008) Vascular dementia: Lowpotension as a key point. Vascular health and risk management, 4(2), 395-402).The relationship between dementia and hypertension in the elderly is controversial, with some studies showing a link between dementia and hypertension (see Gorelick PB (2011) Vascular contributions to cognitive impairment and dementia: A statement for healthcare professionals from the American Heart Association / American Stroke Association. Stroke, 42(9), 2672-2713), while others show a link between dementia and low blood pressure (see Novak V(2010) The relationship between blood pressure and cognitive function. Nature reviews. Cardiology, 7(12), 686-698). Elevated blood pressure increases the risk of cancer incidence and mortality (see Stocks T(2012) Blood Pressure and risk of cancer incidence and mortality in the Metabolic Syndrome and Cancer Project. Hypertension, 59(4), 802-810). However, hypertension and other metabolic and carcinogenic factors may interact, so further research is needed. It has been observed that healthy lifestyle habits reduce the risk of both hypertension and cancer (see Kokubo Y (2015) above).

[0006] In 2017-2018, 45% of U.S. adults had hypertension (mean 51.0% for men and 39.7% for women). Prevalence increased with age, with over 75% of adults aged 60 years and older diagnosed with hypertension (75.2% for men and 73.9% for women). The condition occurred most frequently in non-Hispanic black men and women (see Ostchega Y (2020) above).

[0007] According to the World Health Organization (WHO), 1.28 billion adults (aged 30-79) worldwide have hypertension, two-thirds of whom live in low- and middle-income countries. 42% of adults are diagnosed and treated, but 46% do not know they have hypertension. And only about one in five people have their hypertension under control (see World Health Organization (25 August 2021) Hypertension. https: / / www.who.int / news-room / fact-sheets / detail / hypertension).

[0008] Systolic and diastolic blood pressure are the most common blood pressure measurements used in research and clinical practice. They are established as independent risk factors for cardiovascular disease and can be directly estimated (see Muntner P(2019)Measurement of Blood Pressure in Humans: A Scientific Statement from the American Heart Association. Hypertension,73(5),e35-e66).

[0009] In addition to systolic and diastolic pressure values, blood pressure waveform shape is also important in the diagnosis of cardiovascular disease. A decrease in systemic arterial compliance, which is considered the best indicator of pulsatile arterial dysfunction, can be detected by analyzing the waveform shape (see McVeigh GE (1999) Pressure pulse contour analysis identified in arterial compliance: aging and arterial compliance. Hypertension, 33 (6), 1392-1398).

[0010] Shape analysis of digital pulse waves allows a simple, non-invasive and reproducible measurement of aortic stiffness (see Millasseau SC (2002) Determination of age-related increases in large artery stiffness by digital pulse contour analysis). Indices of arterial stiffness that can be measured using waveform analysis include pulse wave velocity (velocity of pulse wave propagation along the length of the artery), amplification factor (difference between the second systolic peak and the first systolic peak divided by pulse pressure), volume compliance (ratio of pressure change to volume change during the exponential growth phase of the diastolic pressure decay), and oscillatory compliance (ratio of oscillatory pressure change to oscillatory volume change during the exponential growth phase of the diastolic pressure decay) (see Mackenzie IS (2002) Assessment of Arterial Stiffness in Clinical Practice. QJM: Monthly Journal of the Association of Physicians, 95(2), 67-74).

[0011] Similarly, cardiac output, a key determinant of oxygen delivery, can also be estimated using continuous pulse wave analysis (see Saugel B(2021)Cardiac output estimation using pulse wave analysis-physiology, algorithms, and technologies: a narrative review.British Journal of Anaesthesia,126(1),67-76).

[0012] Pulse contour analysis is a suitable method for monitoring cardiac index after surgery in pediatric patients with congenital heart disease. There is a strong correlation between cardiac index calculated using thermodilution and waveform analysis (r 2= 0.86) (see Fakler U(2007)Cardiac index monitoring by pulse contour analysis and thermodilution after pediatric cardiac surgery.The Journal of Thoracic and Cardiovascular Surgery,133(1),224-228). Arterial waveform analysis allows the calculation of derived parameters such as stroke volume, cardiac output, vascular resistance, stroke volume variation, and pulse pressure variation (see Esper SA(2014)Arterial waveform analysis.Best Practice & Research.Clinical Anaesthesiology,28(4),363-380). Other features that can be extracted from the pressure waveform are MPA and MNA (maximum positive and maximum negative amplitude, respectively). MPA and MNA measured on the index fingers of both hands using photoplethysmography are important parameters for screening the early stages of cerebral artery stenosis (Kang HG (2018) Identification of Cerebral Artery Stenosis Using Bilateral Photoplethysmography. Journal of Healthcare Engineering, 2018, 3253519).

[0013] The risk of coronary artery disease, peripheral artery disease, cardiopulmonary disease, cerebrovascular disease, and renal disease can be assessed by blood pressure, which is a strong independent risk factor. Accurate measurement is crucial for classification, evaluation, and treatment planning of blood pressure-related risk factors (Pickering TG (2005) Recommendations for blood pressure measurement in humans and experimental animals, part 1: blood pressure measurement in humans: a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research. Hypertension, 45: 142-161). The relationship between cardiovascular disease and blood pressure has been proven in many epidemiological studies. It is estimated that patients with coronary heart disease account for approximately 15% of blood pressure-related deaths due to coronary heart disease (Miura K(2001)Relationship of blood pressure to 25-year mortality due to coronary heart disease,cardiovascular diseases,and all causes in young adult men:the Chicago Heart Association Detection Project in Industry.Archives of Internal Medicine,161(12),1501-1508). Blood pressure fluctuations are the rule rather than the exception.

[0014] For about 100 years, the main method of measuring blood pressure has been the auscultation method. The Korotkoff method has been used without significant improvement, despite its limited accuracy. However, this method is being replaced by newer methods that are more suitable for automated measurement.

[0015] The gold standard for traditional clinical measurements has been the mercury sphygmomanometer. The newer mercury sphygmomanometers are protected from spillage if dropped, and their use has changed little over the past 50 years. Aneroid sphygmomanometers use a mechanical system of metal bellows and levers. The drawback of this system is that it loses stability and requires periodic calibration. The accuracy of aneroid sphygmomanometers is lower than that of mercury sphygmomanometers and varies greatly between manufacturers. Hybrid sphygmomanometers are devices that combine the characteristics of both stethoscopes and electronic devices, where the mercury column is replaced by an electronic manometer, such as that used in an oscillometer. As the accuracy of electronic devices improves, hybrid devices may replace mercury devices (see Pickering TG (2005) above).

[0016] The next generation of blood pressure measurement is the oscillometric method, first described by Marey in 1876 (see Marey EJ (1876) Physiologie experimentale: Marey (1876) Physiologie experimentale: Travaux du Laboratoire de M. Marey). Further observations of this method showed that the point of maximum oscillations in cuff measurements corresponds to the mean intra-arterial pressure recorded with gradually deflated cuffs (see Mauck G (1980) The meaning of the point of maximum oscillations in cuff pressure in the indirect measurement of blood pressure. Part II. Journal of Biomechanical Engineering, 102(1), 28-33).

[0017] The oscillations begin at pressures above the systolic pressure and progress to pressures below the diastolic pressure. As a result, the systolic and diastolic pressures are indirectly assessed using algorithms based on empirical data. The advantages of this method are that it does not require placement of a transducer on the brachial artery, it is not essential to wear a cuff, it is less susceptible to external noise, and the cuff is removable (see Pickering TG (2005) above).

[0018] The oscillometric method has limitations because the oscillations depend not only on blood pressure but also on several other factors, the most important of which is arterial stiffness. In elderly people with arterial stiffness, the mean arterial pressure value may be underestimated. Also, the algorithms for detecting diastolic and systolic blood pressure are not published by the manufacturers, so there are significant measurement differences when using devices from different manufacturers (see Amoore JN (2000) Can simulators evaluate systematic differences between oscillometric non-invasive blood-pressure monitors? Blood pressure monitoring, 5 (2), 81-89). The oscillometric device is well compatible with intra-arterial and Korotkoff measurements. In addition, the device is inexpensive compared to the devices used in intra-arterial and Korotkoff measurements, making it suitable for both portable and home monitoring applications (see Pickering TG (2005) above).

[0019] Another method is the finger cuff method, which uses the principle of the "unloaded arterial wall". Arterial pulsation is detected by a photoplethysmograph in a finger under a pressure cuff. The cuff is inflated to the same pressure as the intra-arterial pressure and further inflated until the cuff is on the verge of collapse and the transmural pressure is close to zero. The output of the photoplethysmograph is used in a servomechanism system that controls the cuff pressure (see Muntner P (2019) above). A solution of this kind was presented by Penaz. In addition to finger arteries, measurements can be made in other arteries that are easily transilluminated (arteries that are accessible from the surface and are located in soft tissue against a bone or other background). Typical examples are the forearm or the side of the head (Penaz J (1988) Automatic noninvasive blood pressure monitor. U.S. Pat. No. 4,869,261). In general, oscillometric measurements with cuffs provide only information on systolic and diastolic blood pressure values.

[0020] The last group are those using tonometry. The principle of these methods is based on compressing or clamping the artery against the bone, and the resulting pulsation is proportional to the intra-arterial pressure. There are methods that can be used to measure the pressure signal at the wrist, where the radial artery is located on the radius bone. The measured signal is sensitive to position, so the transducer must be placed directly over the center of the artery. This method is not suitable for routine clinical practice, as it requires calibration for each patient. In applanation tonometry, a single transducer is held in the hand to record the pressure waveform over the radial artery. The brachial artery is used to monitor systolic and diastolic pressure (see Pickering TG (2005) above).

[0021] At the first visit, blood pressure should be measured in each arm and the arm with the higher blood pressure value should be selected. At the next visit, blood pressure should be measured in the arm selected at the first visit (Williams B(2018)2018 ESC / ESH Guidelines for the management of arterial hypertension.European Heart Journal,39(33),3021-3104). However, no clear difference was found between the two arms, but a difference of about 10 mmHg was observed in 20% of subjects (see Lane D(2002) Inter-arm differences in blood pressure:when they are clinically significant?Journal of Hypertension,20(6),1089-1095). It is recommended to use a cuff similar to a bladder with a length of 80% of the arm circumference and a width of at least 40%. First, inflate the cuff to a pressure of at least 30 mmHg higher than the radial artery pulse vanishing point. The contraction rate should be 2-3 mmHg per second (see Pickering TG (2005) above). Measurements are generally performed in the sitting or supine position, but results differ between the sitting and supine positions. Diastolic blood pressure measured in the sitting position is about 5 mmHg higher than when measured in the supine position (see Netea RT (2003) Influence of body and arm position on blood pressure reading: an overview. Journal of Hypertension, 21 (2), 237-241). If the cuff is at the level of the right atrium, the supine position will give 8 mmHg higher systolic blood pressure than the upright position (see Terent A (1994) Epidemiological perspective of body position and arm level in blood pressure measurement. Blood pressure, 3 (3), 156-163). If the patient's back is not supported, the diastolic blood pressure may be 6 mmHg higher (according to Cushman, 1990).In addition, crossing the legs may increase systolic blood pressure by 2 to 8 mmHg (Peters GL (1999) The effect of crossing legs on blood pressure: a randomized single-blind crossover study. Blood pressure monitoring, 4 (2), 97-101). The above considerations show that the results obtained by tonometry are ambiguous. Even when the waveform shape is obtained using applanation sphygmomanometry, the accuracy of tonometry depends on the experience of the person measuring it.

[0022] Considering all the above mentioned methods, it can be seen that they have limitations, in this case noninvasive blood pressure (NIBP) measurement method is considered to be the optimal waveform measurement method for clinical purposes. Noninvasive blood pressure (NIBP) measurement devices can use a finger cuff with a photoelectric volumetric volume plethysmograph, or to avoid user bias, applanation tonometry devices can be implemented that do not hold the sensor during measurement, but place it, for example, on a bracelet (see Lakhal K (2018). Noninvasive BP Monitoring in the Critically Ill: Time to Abandon the Arterial Catheter? Chest, 153 (4), 1023-1039).

[0023] There are three main approaches to deriving central arterial pressure from radial artery measurements: spectral (frequency) domain representation and resynthesis of the signal, variations on the autoregressive moving average model, and reduced order or lumped parameter models.

[0024] The spectral representation (also called the frequency approach to determining central arterial pressure from radial arterial pressure) was derived directly from the theory of signal analysis and signal processing. The method was formulated in the early 1990s by Mustafa Karamanoglu and his colleagues at the University of New South Wales, Australia (see Karamanoglu (1993) M An analysis of the relationship between central aortic and peripheral upper limb pressure waves in man. European Heart Journal. Feb;14(2):160-7). The subjects were 14 patients. For each patient, the brachial artery and ascending aortic blood pressure waveforms were recorded with a micromanometer, while the radial artery pressure was recorded with an applanation tonometry device. The blood pressure waveform measurements were made at steady state before and after the administration of a nitroglycerin tablet. The transfer functions were obtained using Fourier analysis, each of which was defined by the following equation:

[0025]

number

[0026]

number

[0027]

number

[0028] The second group of models (called autoregressive moving average model type, ARMA type) was first proposed and popularized in the late 1990s through the work of Chen-Huan Chen and co-authors (a team from the Johns Hopkins University Medical Institution in Baltimore) (see Chen CH (1997) Estimation of central aortic pressure waveform by mathematical transformation of radial tonometry pressure. Validation of generalized transfer function. Circulation. Apr 1; ​​95 (7): 1827-36). The basic concept of these models is based on calculating the individual transfer functions (TFs) between a pair of central aortic pressure and radial aortic pressure using the ARMAX (AutoRegressive Moving Average with eXogenous input) model and generalizing each transfer function (TF) to obtain a generalized transfer function (GTF). The main criticism of the spectral method is that the inter- and intra-patient variability of the transfer functions (TFs) was not systematically evaluated. Chen's goal was to determine the magnitude of transfer function (TF) variation between central aortic and radial aortic pressures both under control conditions and under physiological manipulations that significantly altered blood pressure. For each of the 20 patients, aortic pressure was recorded with a micromanometer and radial artery pressure was recorded with an automated tonometry device. Data were recorded for each subject at steady state and then during at least one of several hemodynamic transient maneuvers. A transfer function (TF) between aortic and radial artery pressures was calculated for each subject using a linear ARMAX model. A direct transfer function (TF) corresponding to the physiological system was derived from the aortic pressure input and the radial artery tonometer signal output.

[0029]

number

[0030]

number

[0031] When the estimated waveforms were obtained using either the individual or the generalized inverse transfer functions, the pulse amplitude and shape of the estimated waveforms at steady state were similar to the measured central arterial waveforms. The accuracy of the waveform estimation was slightly higher with the individual inverse transfer functions (accuracy was compared using measurements of the minimum area of ​​regression of the estimated and measured waveform plots). When calculated using the generalized transfer function, the calculated central arterial pressure differed from the measured value by ≤0.2 ± 3.8 mmHg. The variance of the individual transfer functions (ITFs) was 0.9 mmHg. Comparison of augmentation indexes (AI) revealed important differences, with AI values ​​from reconstructed waves being lower than those calculated based on measured waveforms. Generalized transfer functions (GTFs) reduced AI values ​​by 30 ± 45%, while the use of individual transfer functions (ITFs) reduced the variance of underestimation. Reconstruction of the aortic waveform under transient load changes relied on the constancy of the transfer function for both the generalized transfer function (GTF) and the individual transfer function (ITF). On average, the transfer function (TF) was constant, but in several patients the transfer function (TF) changed significantly during transients, preventing accurate reconstruction of blood pressure. Intrapatient variability of the transfer function (TF coefficient of variation of peak amplitude >20%) was observed in 4 out of 14 patients. The authors concluded that the generalized transfer function (GTF) gave results that were almost as reliable as the individual transfer function (ITF). This suggests that the upper limb vascular branching, which leads to pressure amplification, is a stronger factor influencing the transfer function (TF) compared to factors such as age, sex, and body type. A similar method is the n-point moving average (NPMA), which acts as a digital low-pass filter used to smooth and remove high-frequency noise. However, the optimal moving average denominator in this method is empirically determined using validation data from a selected population. As a result, the accuracy of this method is only as good as that obtained with spectral methods (see Miyashita H (2012) Clinical Assessment of Central Blood Pressure. Current hypertension reviews, 8 (2), 80-90).

[0032] The third group of methods for determining the aortic pressure waveform was based on the tube load model proposed by Mukkamala (R Mukkamala (2019) Methods and apparatus for determining a central aortic pressure waveform from a peripheral artery pressure waveform. U.S. Pat. No. 10,251,566) or Gao (Gao M (2016) A simple adaptive transfer function for deriving the central blood pressure waveform from a radial blood pressure waveform. Scientific Reports, 6(1), 1-9). Below, we explain how to use the more sophisticated Mukkamala method. A mathematical transformation between the aortic pressure (AP) waveform and the peripheral arterial pressure (PAP) waveform was built on the assumption that the central aortic blood flow during diastole due to the closure of the aortic valve is negligible, together with a distributed model representing the arterial tree. The first step was to define the transfer functions between the aortic pressure (AP) and the peripheral arterial pressure (PAP) and between the peripheral arterial pressure (PAP) and the central arterial blood flow, using the distributed model. The parameters of the model were estimated by finding another transfer function that minimized the magnitude of the central arterial blood flow waveform under diastolic conditions when applied to the measured peripheral arterial pressure (PAP). These parameters were then substituted into the previous transfer function to finally convert the peripheral arterial pressure (PAP) to aortic pressure (AP). The parameters were updated by the transfer function each time a new waveform segment was available.

[0033] The distributed model representing the arterial tree consists of parallel segments of uniform tubes (pathways between the aorta and peripheral arteries) placed in series with lumped parameter end loads (distal arterial beds of peripheral arteries). The transfer function from blood pressure to blood pressure is given by:

[0034]

number

[0035]

number

[0036] The prior art search revealed papers presenting clinical evaluations of the non-invasive methods of hemodynamic monitoring mentioned above. One of the papers compared non-invasive measurement of arterial pressure with ClearSight™ (vascular unloading technology) with invasive measurement performed after induction of anesthesia during cardiac surgery, i.e., at the time when measurement of mean arterial pressure is required. The conclusion was that non-invasive measurement is a useful alternative, even taking into account limitations such as blue-finger syndrome. The use of non-invasive measurement was a feasible alternative to invasive measurement, especially in anxious patients or when the difficult situation of radial artery puncture was expected (see Frank P (2021) Noninvasive continuous arterial pressure monitoring during anesthesia induction in patients undergoing cardiac surgery. Annals of Cardiac Anaesthesia, 24: 281-7). The opposite position was presented in a review article by Kim (see Kim SH(2014)Accuracy and precision of continuous noninvasive arterial pressure monitoring compared with invasive arterial pressure: a systematic review and meta-analysis.Anesthesiology,120(5),1080-1097).

[0037] The researchers performed a comprehensive review and meta-analysis comparing continuous noninvasive arterial pressure monitoring with invasive arterial pressure monitoring. The results of continuous noninvasive arterial pressure monitoring were approved and recommended by the Association for the Advancement of Medical Instrumentation when the pooled bias estimate was 5 mmHg or less and the standard deviation was 8 mmHg or less. The findings showed that the imprecision of noninvasive arterial pressure monitoring devices was beyond the acceptable range. It should be noted that the purpose of this study was to evaluate the relative accuracy of noninvasive arterial pressure monitoring, not to evaluate the potential clinical utility of the devices used for said monitoring. Overall, more accurate and precise measurement devices are needed for clinical decision-making processes, final outcomes, or safety. Another meta-analysis confirmed the above results (see Saugell B (2020) Continuous noninvasive pulse wave analysis using finger cuff technologies for arterial blood pressure and cardiac output monitoring in perioperative and intensive care medicine: a systematic review and meta-analysis. British Journal of Anaesthesia, 125 (1), 25-37). This analysis consisted of the results of several studies, which found that arterial pressure, cardiac output, and cardiac index were compatible when measured using noninvasive finger cuff devices and invasive methods as a reference. However, the combined results presented in this meta-analysis showed that noninvasive and invasive methods were not compatible. This result was due to the substantial variability and significant heterogeneity of arterial pressure values, cardiac output, and cardiac index in the above studies.In general, heterogeneity of results occurs due to several factors related to patients (e.g., different populations), clinical environment (e.g., use of vasopressors and inotropes), and devices (e.g., different versions of software used for monitoring). In this review article, only the agreement of the results of the study and the reference method was analyzed, and the trend understood as the relative change in arterial pressure or cardiac output and indicators over time was not analyzed. There is also another article on the clinical evaluation of the CNAP (trademark) device (Continuous Noninvasive Arterial Pressure; CNSystems Medizintechnik AG). According to one such article (Ilies C (2012) Investigation of the agreement of a continuous non-invasive arterial pressure device in comparison with invasive radial artery measurement. British Journal of Anaesthesia, 108 (2), 202-210.), in normotensive conditions, CNAP was compatible with invasive blood pressure monitoring for mean arterial pressure values, but after induction of anesthesia, the results were different and the arterial pressure was low and compatibility was not obtained. In conclusion, the CNAP™ can be used as an additional blood pressure monitoring device, although it is not statistically equivalent to the invasive method under anesthesia. A similar conclusion was also presented in a paper by R. Hahn (see Hahn R (2012) Clinical validation of a continuous non-invasive haemodynamic monitor (CNAP™ 500) during general anaesthesia. British Journal of Anaesthesia, 108 (4), 581-585). That is, the CNAP™ monitoring device was found to be consistent with the invasive device for arterial pressure, but the non-invasive device did not meet the predefined requirements.

[0038] The cited prior art documents confirm the uncertainty of the agreement between invasive and non-invasive measurement techniques. Since non-invasive measurement techniques have many advantages (such as safety and ease of use), there is a need to develop mathematical methods to convert non-invasively obtained blood pressures to invasively obtained blood pressures.

[0039] As shown by the above prior art review, there are various different proposals for extrapolating blood pressure measurements from distal to proximal (central). Among the most traditional approaches are the black box methods, which are well established and most widespread in practice. These methods promise to solve the problem with a single, simple and universal formula, independent of gender, age and other health-specific factors. Opposed to these methods are complex distributed models with dozens of unknown empirical parameters (related to physiology, even if slightly different).

[0040] Current cardiovascular interaction models include high-dimensional models (e.g., 2D and 3D) and low-dimensional models (e.g., 1D, 0D, and tube loading models) (see Zhou S, et al. A review on low-dimensional physics-based models of systemic arteries: application to estimation of central aortic pressure. Biomedical Engineering Online. 2019;18(1):41,1-25). High-dimensional models are predicated on describing local phenomena in specific regions of the circulatory system. These models are most often derived from mass balance equations in the Euler formula and Navier-Stokes equations, sometimes complemented with constitutive relations of vascular tissue (see Morris PD, et al Computational fluid dynamics modelling in cardiovascular medicine, Heart 2016;102:18-28). High-dimensional models are complex and computationally demanding, so they can only be used in strictly limited circulatory regions. For broader applications, high-dimensional models need to be simplified.This is done in stages, in the first step the spatial dimension is reduced to a one-dimensional (1D) model (Raines JK, Jaffrin MY, Shapiro AH. A computer simulation of arterial dynamics in the human leg. Journal of Biomechanics. 1974;7(1):77-91, Formaggia L, Lamponi D, Tuveri M, Veneziani A. Numerical modeling of 1D arterial networks coupled with a lumped parameters description of the heart. Computer Methods in Biomechanics and Biomedical Engineering. 2006;9(5):273-288) and discontinuities in the spatial canal loading model (Swamy G, Mukkamala R, Olivier N. Estimation of the aortic pressure waveform from a peripheral artery pressure waveform via an adaptive transfer function. Annual International Conference of the IEEE Engineering in Medicine and Biology Society.2008:1385-1388).

[0041] Despite their simplicity, models using lumped parameters provide a very effective way to describe important phenomena observed in cardiac hemodynamics (Vlachopoulos Ch, O'Rourke M, Nichols WW. McDonald's Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles 6th Edition, CRC Press, 2011). Examples of such models are shown in Figure 2. The diagram in Figure 2 shows the three basic configurations of the Windkessel model, respectively. The first and oldest model, published by German physiologist Otto Frank at the end of the 19th century, includes only two elements (Figure 2 (2WM)) in the form of compliance (C=dV / dP, the ratio of the change in volume to the change in intravascular pressure) and resistance (mean blood pressure to the mean blood flow), modeling the distensibility of the aorta and the resistance of small peripheral vessels (Frank O. Die Grundform des Arteriellen Pulses. Zeitschrift fur Biologie. 1899; 37: 483-526). Despite its simplicity, this model is suitable to describe the pressure drop in the aorta and can be used to estimate cardiac output or blood pressure. To extend the applicability of 2WM to the high frequency region, Nicholas Westerhof proposed in 1969 to add a third element, the input inertial resistance (see Westerhof N et al. Journal of Biomechanics. 1969; 2 (2): 121-143). The Windkessel three-component model (Figure 2(3WM)) was able to reproduce realistic blood pressure and flow waveforms and was able to fit experimental data from in vivo measurements. The Windkessel three-component model is probably the most widely used and accepted model to describe the systemic circulation.

[0042] The 3WM model is a major drawback, because it ignores the inertial effect, which was demonstrated by Nikos Stergiopulos in the late 90s of the 20th century (see Stergiopulos N, Westerhof BE, Westerhof N. Total arterial inertance as the fourth element of the Windkessel model. Am J Physiol. 1999; 276(1):H81-88). It was shown that the sudden change in central arterial pressure, which first appears as a sudden increase in blood volume in systole and then as a decrease in blood volume in diastole, results in an inertial effect. As Nikos Stergiopulos argued, in order to take these effects into account, it is necessary to complete the model by adding an inertial term (Δp~dq / dt) (Figure 2 (4WM)).

[0043] This detailed review of the prior art revealed a need for an accurate, non-invasive method that takes into account all relevant influences and uses a model that is patient-specific, and a practical implementation of which would allow the method to be used, for example, during clinical intensive care monitoring.The present invention addresses this need. Summary of the Invention [Means for solving the problem]

[0044] The following description is provided for a better understanding of the principles and advantages of the invention as set forth in the appended claims, and is not intended to be limiting in any sense.

[0045] It is an object of the present invention to provide information about parameters of a central arterial pressure waveform as a function of time using one or more non-invasive measurements. The information provided corresponds to values ​​obtained by invasive measurements. It is further an object of the present invention to provide a blood pressure waveform based on a non-invasive measurement, which corresponds to a blood pressure waveform obtained by invasive measurements. It is further an object of the present invention to provide a shape analysis of a digital volume pulse wave and cardiac output based on one or more non-invasive measurements, which corresponds to an analysis, pulse wave, and cardiac output obtained by invasive measurements.

[0046] This allows the invention to diagnose the state of the human heart solely on the basis of one or more non-invasive measurements, thereby enabling the diagnosis and treatment of elevated blood pressure and / or hypertension, avoiding the risks associated with organizing and performing invasive measurements.

[0047] The present invention is based on a specific approach to converting the measured distal blood pressure waveform into a proximal (central) blood pressure waveform using an inventive model of the Windkessel type.

[0048] One of the key points of the present invention is causality: distal blood pressure, e.g. radial artery pressure, is influenced by central aortic variations, and not vice versa. Regardless of the fact that distal blood pressure is measured, the starting point for building the relationship must be the human heart and the main blood vessels leading to it. Another key point is to use models based on patient-specific data and not to follow universal formulas that do not take into account factors specific to sex, age, or other health conditions. In particular, the present invention does not follow models based on unknown empirical parameters that are in principle only used to improve the fit to experimental data.

[0049] In one aspect, the invention relates to a method for reconstructing central arterial pressure waveform morphology from continuous non-invasive distal blood pressure measurements, which may be realized as a computer-implemented invention.

[0050] The method according to the invention comprises the step of collecting patient-specific demographic and health data that influences pressure pulse wave propagation in the body. Patient-specific data is understood as data obtained from a particular human patient. The patient-specific demographic and health data may include the patient's gender, age, height, general physical fitness rating and / or current medications. Current medications include, but are not limited to, beta-adrenergic blockers, angiotensin-converting enzyme inhibitors and / or antiarrhythmic drugs. In general, all medications that may affect the human heart are considered.

[0051] The method according to the invention also includes non-invasive measurement of the patient's systolic blood pressure, diastolic blood pressure, and heart rate. The term "non-invasive measurement" means a measurement that does not involve any type of surgery and / or significant health risk. In some embodiments, the measurement does not involve the insertion of any probe into the patient's body. In some specific embodiments, the method does not require such measurements. Instead, the method includes estimating the patient's systolic blood pressure, diastolic blood pressure, and heart rate based on continuously recorded distal blood pressure waveforms of the patient.

[0052] The method according to the invention also includes a step of continuously recording a distal blood pressure waveform, such as from a distal artery of the patient, the recording including an entire cardiac cycle of the patient and performed non-invasively. In extreme cases of abnormally slow respiratory rate (bradypnea), half the length of the respiratory cycle can be used. The recording includes measuring the arterial blood pressure with a sensor placed on the radial artery using a method selected from photoplethysmography and / or applanation sphygmography. Furthermore, the recording may be made over a sequence of consecutive cardiac cycles within a single respiratory cycle. The recording may also be made over any number of cardiac or respiratory cycles. The term "non-invasive recording" refers to recording without any type of surgery and / or significant health risks. In some embodiments, the recording does not include the insertion of any probe into the patient's body.

[0053] The method according to the invention also comprises a step of determining the parametrics of a coupled system, said coupled system comprising a lumped parameter model of the central compartment of the blood circulatory system and a lumped parameter model responsible for the distal to proximal transmission. The two-model structure of the coupled system ensures an accurate conversion of the measured distal blood pressure into a proximal (central) blood pressure, so that the results obtained can be used in subsequent analysis and diagnosis of the human heart with an accuracy close to that obtained using invasive measurement methods.

[0054] Further features and advantages will become more apparent from the detailed description of the non-limiting embodiments and the accompanying drawings. It is to be understood that, unless otherwise stated, all of the embodiments and features thereof described in this specification and claims can be combined in any order and number to form new embodiments that become part of this disclosure. [Brief description of the drawings]

[0055] The invention will now be described in more detail with reference to the accompanying drawings. [Figure 1] FIG. 1 shows a schematic diagram of a method according to one embodiment of the present invention. [Diagram 2] Figure 2 shows a schematic block diagram of basic variants of mono-compartment lumped parameter models of the blood circulatory system: the two-element (2WM) by Otto Frank, the three-element (3WM) by Nicholas Westerhof, and the four-element (4WM) by Nikos Stergiopulos. [Diagram 3] FIG. 3 illustrates a single section (CRL) of a functional building block according to a preferred embodiment of the present invention, and a multiple section (n?CRL) of the Windkessel model in generalized form. [Figure 4] FIG. 4 shows a lumped parameter model and functional building blocks of the central compartment of the blood circulatory system. [Diagram 5] FIG. 5 shows two exemplary complete records of continuous blood pressure waveform measurements of patients obtained in a clinical trial using an invasive method and a non-invasive method according to the present invention. [Figure 6] FIG. 6 shows reconstructions of central arterial pressure waveforms against a reference signal from an intra-aortic catheter in resting and hyperemic states. [Figure 7] FIG. 7 illustrates the convergence process of model parameters using local and global minimization algorithms according to an embodiment of the present invention. [Figure 8] FIG. 8 shows an assessment of the accuracy of the reconstruction of systolic and diastolic blood pressure obtained using an embodiment of the method of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0056] In a preferred embodiment of the present invention, reconstruction of the central arterial pressure waveform from distal non-invasive measurements is performed based on the multi-compartment lumped parameter model building blocks shown in Figure 3. As shown, the multi-compartment model is implemented using the CRL function building block (CRL-Compliance (C i ), resistance (R i ), inertia (L iThe input and output parameters are linked by a pair of equations for blood pressure (p) and blood flow (q).

[0057]

number

[0058]

number

[0059]

number

[0060]

number

[0061]

number

[0062]

number

[0063]

number

[0064]

number

[0065] Practical implementations of distal-to-proximal transfer functions of the form (11), (12), and especially (15), include the i ,R i ,L i}, precise knowledge of blood flow is required. Here, for n = 1 compartment model (1-CRL), the model needs to be completed by an auxiliary {q0} blood flow relation, whereas for n compartments, {q0,q1,...q n-1 Note that the relationship between the distal and central blood pressures must be completed by the relationship: {\distal\left\right}. Mapping distal and central blood pressure requires knowledge of the central compartment of the circulation, or at least knowledge of central blood flow, and in some circumstances, knowledge of blood flow as well.

[0066] In a preferred embodiment of the present invention, the center {q0} or the partitions {q0, q1, ... q n-1}The blood flow is represented by a Windkessel model of adjacent lumped parameters. Thus, in contrast to many of the aforementioned prior art attempts to seek a predetermined transfer function, the present invention does not assume structural rigor of the transfer relationship, but evolutionary laws that provide the necessary relationships to find the missing hemodynamic state quantities. This is one of the key ideas of the present invention, in which the present invention abandons the concept that requires treating the distal measurement location as part of the peripheral circulation region. The central compartment model provides only the missing relationships of (11) and (12), or (15), and does not analyze the blood distribution from the aorta to the peripheral regions. It should be noted that if the missing blood flows were calculated separately and used to construct the transfer function of the blood pressure waveform, a completely uncoupled and decoupled phenomenological model would be obtained; i.e., the central compartment is distributed to the peripheral parts, and the peripheral parts are connected weekly (if such a situation exists) to the central compartment.

[0067] In a preferred embodiment of the present invention, at least one lumped parameter functional building block CRL (shown in FIG. 4(a)) is used to build a central compartment model of the blood circulatory system. This lumped parameter functional building block CRL may have a valve, which is a diode modeling a heart valve. This diode should be understood as a model of unidirectional flow, according to the terminology used in the art. This model may consist of either a single closed loop circuit (FIG. 4(d)), or two closed loop circuits (FIG. 4(c)). In these embodiments, the central compartment model includes at least two CRL functional building blocks: the first functional building block represents large and medium-sized elastic blood vessels exhibiting large inertio-elastic effects, and the second functional building block represents resisto-capacitive effects. In the preferred embodiment of FIG. 4(c), the model is a (C i-1 =C pa , R i =R pa , L i =L pa The inertial and elastic CRL building blocks are determined by Ci-1 =C pv , R i =R pv , L i = 0) and the circulation analogue of the left heart (C i-1 =C sa , R i =R sa , L i =L sa The inertial and elastic CRL building blocks are C i-1 =C sv , R i =R sv , L i In yet another preferred embodiment of the present invention, a sufficiently accurate central compartment model can be constructed based on a single closed loop circuit of the systemic circulation (as shown in FIG. 4(d), the inertial and elastic CRL functional building block has a resistance-capacitance CRL building block with C = 0). i-1 =C sa , R i =R sa , L i =L sa The resistive-capacitive CRL function building block is determined by C i-1 =C sv , R i =R sv , L i = 0). These embodiments of the present invention are shown in diagrams (c) and (d) of FIG. 4, and the equations included in these embodiments follow equations (8) and (9) disclosed herein. The embodiment of FIG. 4(c) is more accurate than the embodiment of FIG. 4(d). However, the embodiment of FIG. 4(d) can provide results much faster than the embodiment of FIG. 4(c) while maintaining sufficient accuracy. In a further embodiment providing more details on the central compartment of the blood circulatory system, L i is not equal to "0." In another embodiment, the time-varying elastance concept (E) used in Figures 4(c) and 4(d) is partially or wholly replaced by the myocardial fiber stress and strain concept (MF).

[0068] In order to reproduce physiological conditions, the closed loop circuit of any central compartment described herein preferably forms a self-excited oscillator. Such a closed loop circuit is therefore completed by components that mimic the hemodynamic action of the heart. The appropriate use of boundary conditions is an alternative to a self-excited oscillator. To incorporate this oscillator, the following must be taken into account: The internal structure of the heart is four chambers (left and right atria, and left and right ventricles) in the form of cavities surrounded by endocardium, epicardium, and a much larger volume of myocardium, made of fibrous walls (see Barrett K et al. (2019) Ganong's Review of Medical Physiology, McGraw-Hill Education, Pappano AJ, Wier WG (2019) Cardiovascular Physiology, Elsevier, Klabunde RE (2018) Cardiovascular Physiology Concepts, Lippincott Williams & Wilkin). The two atria receive blood returning to the heart from the body tissues and the lungs, while the ventricles pump blood to the lungs and all other organs. Each chamber has a valve to maintain one-way blood flow. Each valve opens and closes (more or less) passively in response to the difference in blood pressure on either side of the valve. Thus, the blood flow through the valves, and therefore the outflow of blood from the chambers, can be described (according to the building blocks shown in Figure 4(b)) by the following equation:

[0069]

number

[0070]

number

[0071] In a preferred embodiment of the present invention, the above-mentioned cardiac chamber pressure-volume relationship is formulated using the variable elastance concept (Suga H. (1969) Time course of left ventricular pressure-volume relationship under various enddiastolic volume. Jpn Heart J. 1969; 10(6): 509-15, Walley KR (2016) Left ventricular function: time-varying elastance and left ventricular aortic coupling. Critical Care 20(270): 1-11, Bozkurt S (2019) Mathematical modeling of cardiac function to evaluate clinical cases in adults and children. PloS One. 2019; 14(10): e0224663, Li W (2020) Biomechanics of infarcted left ventricle: a review of modelling. Biomedical Engineering Letters. 10(3): 387-417) and can be expressed by the following equation.

[0072]

number

[0073]

number

[0074] In another embodiment of the present invention, the cardiac chamber pressure-volume closure relationship in the closed-loop lumped parameter central compartment model can be directly expressed by the myocardial fiber stress and strain concept (MF) (Mirota K (2008) Constitutive Models of Vascular Tissue. Solid State Phenomena. Vol. 144, 100-105, Avazmohammadi R et al. A Contemporary Look at Biomechanical Models of Myocardium. Annual Review of Biomedical Engineering. 2019 Jun 4; 21: 417-442, Voigt JU, Cvijic M (2019) 2-and 3-Dimensional Myocardial Strain in Cardiac Health and Disease. JACC Cardiovasc Imaging. 12 (9): 1849-1863). In general, for all thin-walled and (nearly) rotationally symmetric geometries, the fiber stress (σ f ) is the change in bulk modulus

[0075]

number

[0076]

number

[0077] FIG. 1 shows an embodiment of the method according to the present invention. This embodiment is used to explain the present invention in detail. The method of this embodiment has five steps: 1) collecting general demographic and health data of the patient; 2) measuring the systolic blood pressure, diastolic blood pressure, and heart rate of the patient by a selected non-invasive method; 3) recording a single block of distal blood pressure waveform within a selected time window; 4) performing parametric identification of a lumped parameter model, including iterative refinement of the model until convergence is reached; and finally, 5) using the improved model to determine central (proximal) blood pressure and proximal blood flow. In another embodiment not shown in FIG. 1, blood pressure values ​​and heart rate are estimated using the measured distal blood pressure waveform. An arbitrary time window is selected so that the method includes at least one complete cardiac cycle. However, in extreme cases of abnormally slow respiratory rate (bradypnea), a length half the length of the respiratory cycle can be used. Those skilled in the art will understand that such a selection can be made using R waves. In a preferred embodiment, an arbitrary time window is selected so that the method includes two, three, four, five or more complete cardiac cycles.

[0078] In a first step, demographic and general medical data of the patient related to the pressure pulse wave propagation in the patient's body are collected. This data is only used in step 4, which is related to the identification of the model's parametrics, but it influences the efficiency of the method. This efficiency can be measured by the time required to reach the result of the method. In one embodiment, data on gender, age, weight, and height are collected (Smulyan H et al. (1998) Influence of body height on pulsatile arterial hemodynamic data. Journal of the American College of Cardiology. 31(5): 1103-9, Christofaro DGD et al. (2017) Relationship between Resting Heart Rate, Blood Pressure and Pulse Pressure in Adolescents. Arquivos Brasileiros de Cardiologia. 108(5): 405-410, Evans JM et al. (2017) Body Size Predicts Cardiac and Vascular Resistance Effects on Men's and Women's Blood Pressure. Front Physiol. 9; 8: 561, Gallo C et al. (2021) Testing a Patient-Specific In-Silico Model to Noninvasively Estimate Central Blood Pressure. Cardiovascular Engineering and Technology.12(2):144-157). Medications may also affect pulse wave propagation within a patient's body. In particular, patients taking drugs from the group of beta-adrenergic blockers (BBLOCK), angiotensin-converting enzyme inhibitors (ACE), and / or antiarrhythmic drugs (AARR) are affected (Harris WS, Schoenfeld CD, Weissler AM (1967) Effects of adrenergic receptor activation and blockade on the systolic preejection period, heart rate, and arterial pressure in man. Journal of Clinical Investigation. 46 (11): 1704-14, Morgan TO et al. (1974) A comparison of beta adrenergic blocking drugs in the treatment of hypertension. Postgraduate Medical Journal. 50 (583): 253-259, Nyberg G (1976) Effect of beta-adrenoreceptor blockers on heart rate and blood pressure in dynamic and isometric exercise. Drugs. 11 SUPPL 1: 185-95, Fitzpatrick MA, Julius S(1985)Hemodynamic effects of angiotensin-converting enzyme inhibitors in essential hypertension:a review.Journal of Cardiovascular Pharmacology.7 Suppl 1:S35-9,Ting CT et all(1993)Arterial hemodynamics in human hypertension.Effects of angiotensin converting enzyme inhibition.Hypertension.22(6):839-46,Jobs A et al.(2019) Angiotensin-converting-enzyme inhibitors in hemodynamic congestion: a meta-analysis of early studies. Clinical Research in Cardiology. 108(11): 1240-1248, Block PJ, Winkle RA (1983) Hemodynamic effects of antiarrhythmic drugs. American Journal of Cardiology. 52(6): 14C-23C, Weiner B (1991) Hemodynamic effects of antidysrhythmic drugs. Journal of Cardiovascular Nursing. 5(4): 39-48). In some embodiments, the method includes a drug therapy, and in other embodiments, the drug therapy includes the drugs listed above. Other drugs may also affect pulse wave propagation in the patient's body, so other drug therapies are included in other embodiments. It should be noted that drug therapy includes various dosing regimens, and any other medical effects. Each of the above mentioned factors can be used individually or in any way combined to obtain a better starting value for (initiating) the model identification process for a particular patient case. A better starting value directly impacts the efficiency of the method according to the invention. In certain embodiments, it is envisaged that only selected patient-specific data is collected, such as, for example, gender, age or selected medication.

[0079] In a second step, in one embodiment, the systolic blood pressure (SYS), diastolic blood pressure (DIA), and heart rate (HR) are estimated based on measurements of the distal blood pressure waveform performed in arbitrary units (AU). Those skilled in the art know how to estimate the systolic blood pressure (SYS), diastolic blood pressure (DIA), and heart rate (HR) values ​​from the measured distal blood pressure waveform. The main objective of this approach is to obtain data from a source independent of the source for obtaining the distal blood pressure waveform, and b) data that allows for the calibration of a lumped parameter model of the central compartment. This is important because measurements of the distal blood pressure waveform are generally assumed to provide information only about the waveform morphology. In this case, it is acceptable to express the results of the non-invasive measurement of the distal blood pressure waveform in any unit, not necessarily in pressure units. Thus, in one embodiment, the non-invasive measurement of the distal blood pressure waveform is performed using arbitrary units (AU). Some devices used to record the distal blood pressure waveform have the capability to independently measure the systolic blood pressure (SYS), diastolic blood pressure (DIA), and heart rate (HR) parameters. For example, if such a device includes an upper arm cuff, the measurement results are used as the source of the data described above. Thus, in another embodiment, in the second step, the non-invasive measurement of the distal blood pressure waveform is made in pressure units, and it may not be necessary to estimate the systolic blood pressure (SYS), diastolic blood pressure (DIA), and heart rate (HR) values, since the systolic blood pressure (SYS), diastolic blood pressure (DIA), and heart rate (HR) are measured independently by the same or different devices. In these embodiments, the measurements including the distal blood pressure waveform can be used directly to calibrate the lumped parameter model of the central compartment. In another embodiment, the entire coupling system can be calibrated, or only one of the models included in the coupling system can be calibrated. In one particular embodiment, the distal to proximal transmission lumped parameter model is calibrated using measurements from, for example, the radial artery, the aorta, and the patient's demographic and medical data.

[0080] In the third step, a non-invasive continuous recording of the distal blood pressure waveform is made. Most devices offering non-invasive measurement capabilities scale the distal blood pressure waveform to a value expressed in pressure units (usually expressed in millimeters of mercury, denoted here as mmHg or mm Hg), but this is not necessary for the present invention. The measured distal blood pressure can be expressed in arbitrary units (denoted here as AU), since it is the form of the signal itself that is important. Considering the limitations and preferences of clinical practice in measuring continuous non-invasive blood pressure (CNBP), in a preferred embodiment of the present invention, it is envisaged that the distal blood pressure measurement is made in the radial artery. Recommended measurement methods include finger cuff photoplethysmography and / or applanation sphygmography. The recording and analysis of distal blood pressure can be performed within a single block window encompassing a predetermined number of complete cycles corresponding to the systolic and diastolic operation of the heart. In a preferred embodiment of the invention, a window of width corresponding to the length of the respiratory cycle is analyzed (Rodriguez-Molinero A (2013) Normal respiratory rate and peripheral blood oxygen saturation in the elderly population. Journal of the American Geriatrics Society. 61(12):2238-2240, Park C, Lee B (2014) Real-time estimation of respiratory rate from a photoplethysmogram using an adaptive lattice notch filter. Biomedical Engineering Online. 17;13:170, Scholkmann F, Wolf U (2019) The Pulse-Respiration Quotient: A Powerful but Untapped Parameter for Modern Studies About Human Physiology and Pathophysiology. Front Physiol. 9;10:371). In other embodiments, including extreme cases of abnormally slow respiratory rates (bradypnea), a length half the length of the respiratory cycle is used.

[0081] In the fourth step, a step of identifying the model parametrics is performed. The model structure for which the parameters are identified is defined by a coupled system having two different models: a lumped parameter model of the central compartment of the blood circulation system and a lumped parameter model of the transmission from distal to proximal. In this step, the whole step of identifying the model parametrics is performed by three sub-steps. First, the proximal blood pressure and blood flow are calculated using the closed-loop central compartment of the blood circulation system lumped parameter model. It should be noted that in the basic and preferred embodiment, the central compartment of the blood circulation system model has the systemic circulatory system and the pulmonary circulatory system (i.e., the left and right cardiac cycles) in order to obtain the most detailed and accurate results. However, the central compartment of the blood circulation system model can be limited to only the systemic circulatory system without significantly compromising the prediction accuracy. The initial values ​​of the empirical parameters mentioned above are provided using the demographic and general medical data of the patient collected in the first step of the method. More precisely, the initial values ​​of the empirical parameters are calculated by referring to the literature, if necessary, together with the demographic and general medical data of the patient. Furthermore, the distal to proximal blood pressure approximation is determined according to a distal to proximal lumped parameter model. Since both models forming a coupled system are analyzed in a coupled manner, the error function is defined as follows, where β is the estimated model parameter set:

[0082]

number

[0083] In a preferred embodiment of the invention, the provision of the empirical parameters mentioned above is formulated as an optimization task based on a minimization method (see Villaverde AF et al. (2019) Benchmarking optimization methods for parameter estimation in large kinetic models. Bioinformatics. 35(5): 830-838, Kreutz C (2019) Guidelines for benchmarking of optimization-based approaches for fitting mathematical models. Genome Biol. 20(1): 281, Schmiester L (2020) Efficient parameterization of large-scale dynamic models based on relative measurements. Bioinformatics. 36(2): 594-602). Thus, the error function (21) is used as the loss function of the optimization algorithm. This approach of providing empirical parameters is much more flexible than the known classical parametric identification approaches. Thus, in a preferred embodiment of the invention, the design of the method comprises a local search (minimization) algorithm on the mathematical structure of the coupled system. Its task is to directly control the selection of empirical parameter values ​​in the fourth step of the method, and to gradually improve the quality of the central arterial pressure prediction. In a preferred embodiment of the invention, three local search (minimization) algorithms can be used alternatively or in any combination.In one embodiment, the relatively stable and moderately complex Nelder-Mead method can be used (Nelder J, Mead R (1965) A simplex method for function minimization. Computer Journal, 7(4): 308-313, Gao F, Han L (2010) Implementing the Nelder-Mead simplex algorithm with adaptive parameters. Computational Optimization and Applications, 51: 1, 259-277). This method does not require the calculation of derivatives and uses only the value of the objective function as an n-dimensional simplex, which is then transformed geometrically. In another embodiment, the local search (minimization) algorithm uses Sequential Least Squares Programming methods (see Kraft D (1988) A software package for sequential quadratic programming. DFVLR, Braunschweig, Koln) and / or Broyden-Fletcher-Goldfarb-Shanno (Zhu C, Byrd RH, Nocedal J (1997). Algorithm 778. L-BFGSB: Fortran routines for large scale bound constrained optimization. ACM Transactions on Mathematical Software, 23(4), 550-560). Both of these methods solve Newton-type equations.

[0084]

number

[0085]

number

[0086] The fifth and final step is used to calculate the central arterial pressure value and / or the proximal blood flow based on the results of the parametric identification step. For the avoidance of doubt, the parametric identification step has a coupling system with empirical parameters refined for a particular patient. The values ​​calculated in this way can be output in any suitable manner and format. Knowledge of the proximal (central) arterial pressure and the proximal flow in a given time window allows the waveform morphology of the central arterial pressure to be reconstructed. This is useful for the diagnosis and treatment of elevated blood pressure and / or hypertension.

[0087] Various embodiments of the method of the present invention were evaluated in detail. The following shows the validation results of the embodiment implementing the model of FIG. 4(c) with X=E. The method of the present invention was validated based on the medical experimental results of a multi-center non-randomized clinical trial. The demographic data, health data, and measurement results of the patients are summarized in Table 1 below (measurements are expressed as mean ± standard deviation).

[0088] [Table 1] In general, all hospitalized patients in this study had clinical signs of coronary artery disease. For each patient, continuous noninvasive blood pressure measurements (CNBP) were performed in the radial artery at a sampling rate of 100 Hz. In addition, an independent noninvasive measurement was performed in the brachial artery using an oscillometric method.

[0089] Since the patients underwent invasive diagnosis, the results of blood pressure measurements with an intra-aortic catheter were used as reference data. The blood pressure signals obtained by invasive measurements were sampled at 200 Hz. As an example, Figure 5 shows the complete blood pressure recordings of two patients (cases A and B), the upper part being the non-invasive recording (performed using finger cuff photoplethysmography) and the lower part being the invasive recording (performed using an intra-arterial catheter).

[0090] As the records for Case B show, there are clear signs of arrhythmia (see Heart Rate Irregularities and Dysrhythmias). According to the Clinical Case Report Form (CRF), this patient was treated with AARR (antiarrhythmic) medications.

[0091] Figure 6 shows the results of reconstruction of central arterial pressure against the reference data for the two cases shown in Figure 5. The reconstruction results are quite satisfactory in both cases, but the adverse effect of arrhythmia in case B is clearly visible.

[0092] Reconstruction of the central arterial pressure can be performed using a local minimization algorithm or a global minimization algorithm, or a combination of both types of minimization algorithms.

[0093] A preferred embodiment of the central arterial pressure reconstruction method is based on local minimization algorithms, including, for example, the Nelder-Mead method, SLAQP, and / or L-BFGSB. In Figure 7, we can see the convergence of the minimization process achieved using the Nelder-Mead method, SLAQP, or L-BFGSB (arrhythmic patient, case B). It can be clearly seen that each local minimization algorithm successfully completed the task of reconstruction. The errors measured as MAE (mean absolute error) were 0.1196667, 0.1203333, and 0.1203333, respectively. Similarly, the calculated results of RMSE (root mean square error) were 0.1401587, 0.1407667, and 0.1408015. However, the nature of the convergence was quite different. In the case of the Nelder-Mead algorithm, the iterative process converged stably, but was very time-consuming. In contrast, SLAQP and L-BFGSB converged more rapidly, although additional effort was required to suppress numerical oscillations and achieve convergence.

[0094] Alternatively or additionally, one or more global minimization algorithms can be used. The lower part of FIG. 7 shows the convergence results using two global minimization algorithms, AMPGO and SHGO. A final summary of the results obtained by the method of the invention implementing the global minimization algorithms is shown in FIG. 8. This figure summarizes the systolic and diastolic blood pressure values ​​(left and right sides of the figure, respectively) obtained from the invasive measurements and the calculations implementing the method of the invention. The upper part of the figure contains a correlation plot of all the cases constituting the clinical trial, shown in a window including 5 cardiac cycles, and the lower part of the figure contains the Bland-Altman plot (Tukey mean difference plot) of the cases. The more than 250 validation examples shown in FIG. 8 clearly show a very good agreement between the results obtained invasively and those obtained using the method of the invention. It is noted that one or more other global minimization algorithms can be used in the method together with the AMPGO and / or SHGO algorithms or as an alternative thereto.

[0095] In another embodiment, various combinations of local and global minimization algorithms are used to determine, for example, systolic and diastolic blood pressures. These various combinations are used to meet specific implementation requirements. In these embodiments, one or more local minimization algorithms (e.g., Nelder-Mead, SLAQP, and / or L-BFGSB) can be used before, after, and / or as layers of one or more global minimization algorithms (e.g., AMPGO and / or SHGO). In some embodiments, only one or more local minimization algorithms can be used. These algorithms are computationally less intensive and therefore can provide faster results. Also, embodiments including only these algorithms can be implemented on a portable device with low computational power. In another embodiment, the portable device is used to communicate with a server configured to perform the overall computationally intensive calculations. These communications can be over a local network or the Internet.

[0096] Any computational steps or substeps of the methods of the present invention can be performed using a computer or a computer program. In some specific embodiments, part of the computations, or the entirety, are performed using a computer program stored on a computer or any type of storage device, or both. In another embodiment, part of the computations for the method, or the entirety, can be performed remotely, for example, using a cloud-based infrastructure, which can include the use of the Internet or a local network.

[0097] All the measurement methods and minimization algorithms described herein are intended to define the parameters of the present invention and provide specific results for comparison with clinical trials, and are exemplary and not limiting in any way. Terms such as "including" or "comprising" are not limiting, for example, if an element A contains another element B, the element A may contain other elements or elements in addition to element B. The use of the singular or plural forms is not intended to limit the scope of the present disclosure, for example, a portion of a description indicating that an element A contains an element B discloses an embodiment in which multiple elements B are contained in an element A, and multiple elements A are contained in an element B, and further, multiple elements A contain multiple elements B. Many other embodiments will be apparent to those skilled in the art upon review of the present disclosure. Thus, the scope of the present invention should be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

Claims

1. 1. A method for reconstructing central arterial pressure waveform morphology in a human patient from continuous non-invasive distal blood pressure measurements, comprising: (4) identifying the parametrics of the coupled system using the patient's specific demographic and health data, and using the patient's estimated systolic blood pressure, diastolic blood pressure, and heart rate; the patient-specific demographic and health data affecting pressure pulse wave propagation within the patient; the coupled system includes a central compartment lumped parameter model of the blood circulatory system and a distal-to-proximal transmission lumped parameter model; wherein estimating the patient's systolic blood pressure, diastolic blood pressure, and heart rate is performed using the patient's recorded distal blood pressure waveform; recording the patient's distal blood pressure waveform continuously within a time window that includes at least one complete cardiac cycle or one half-cycle of a complete respiratory cycle of the patient; The recording is performed non-invasively. Identifying the parametric; (5) calculating central arterial pressure and proximal blood flow using the results of the step of determining the parametric parameters; A method comprising:

2. 10. The method of claim 1, wherein the patient-specific demographic and health data includes gender, age, height, general physical fitness rating, and / or current medications; The method, wherein the currently taking medication comprises a beta-adrenergic blocker, an angiotensin-converting enzyme inhibitor, and / or an antiarrhythmic drug.

3. 10. The method of claim 1, wherein the continuous blood pressure waveform recording is made at a distal artery and / or comprises measuring arterial blood pressure with a sensor placed on the radial artery using a method selected from photoplethysmography and / or applanation sphygmography.

4. 10. The method of claim 1, wherein the time window includes a sequence of consecutive complete cardiac cycles within a single respiratory cycle, the sequence comprising two, three, four, five, or more consecutive cardiac cycles of the patient within the single respiratory cycle.

5. 2. The method of claim 1, wherein the central compartment lumped parameter model comprises at least one lumped parameter functional building block (CRL); the at least one lumped parameter functional building block (CRL) is selected from the group consisting of a vascular compartment functional building block (CRL) and a heart chamber functional building block; The vascular compartment functional building block (CRL) has the following structure: The blood flow rate (q i-1 ) of the previous partition functional building block and the blood pressure (p i-1 ) of the previous partition functional building block pass through the resistance (R i ) of the current partition functional building block and then through the inertial resistance (L i ) of the current partition functional building block, thereby generating the blood flow rate (q i ) of the current partition functional building block and the blood pressure (p i ) of the current partition functional building block; The resistance (R i ) of the current partition functional building block and the inertial resistance (L i ) of the current partition functional building block are connected in series; The blood flow rate (q i-1 ) of the preceding compartment functional building block and the blood pressure (p i-1 ) of the preceding compartment functional building block are affected by the compliance (C i-1 ) of the preceding compartment functional building block connected in parallel to the blood vessel compartment functional building block (CRL) before passing through the resistance (R i ) of the current compartment functional building block, The heart chamber functional building blocks have the following structure: The blood flow rate (q i-1 ) of the previous compartment functional building block and the blood pressure (p i-1 ) of the previous compartment functional building block pass through the resistance (R i ) of the current compartment functional building block and then through a diode (valve) that models a heart valve, thereby generating the blood flow rate (q i ) of the current compartment functional building block and the blood pressure (p i ) of the current compartment functional building block; The resistor (R i ) of the current compartment functional building block and the diode (valve) modeling the heart valve are connected in series; The structure is such that the blood flow rate (q i-1 ) of the anterior compartment functional building block and the blood pressure (p i-1 ) of the anterior compartment functional building block are affected by an anterior compartment time-varying elastance concept (E i-1 ) or anterior compartment myocardial fiber stress and strain concept (MF i-1 ) connected in parallel to the heart chamber functional building block before passing through the resistance (R i ) of the current compartment functional building block. method.

6. 6. The method of claim 5, wherein the central compartment lumped parameter model comprises at least two of the lumped parameter functional building blocks (CRLs); The first functional building block represents large and medium-sized elastic vessels that exhibit large inertial and elastic effects; The second functional building block represents the resistive-capacitive effect. method.

7. 7. The method of claim 6, wherein the central compartment lumped parameter model comprises a right heart circulatory structure in the form of a superposition of the lumped parameter functional building blocks (CRLs), The inertia and elasticity functional building block is C ｉ-１ =C ｐａ , R ｉ =R ｐａ , L ｉ =L ｐａ is determined by The resistive-capacitive functional building block is C ｉ－１ =C ｐｖ , R ｉ =R ｐｖ , L ｉ = 0, method.

8. 6. The method of claim 5, wherein the central compartment lumped parameter model has the following structure: The venous systemic blood pressure (p sv ) and venous systemic blood flow (q sv ) pass through the right atrium (R.A.), generating right ventricular blood pressure (p RV ) and right ventricular blood flow (q t ); The right ventricular blood pressure (p RV ) and the right ventricular blood flow (q t ) pass through the right ventricle (R.V.), generating a pulmonary artery circulation blood pressure (p pa ) and a pulmonary artery circulation blood flow (q pa ); The pulmonary arterial circulation pressure (p pa ) and the pulmonary arterial circulation blood flow rate (q pa ) pass through the pulmonary arterial circulation (pa) and then through the pulmonary venous circulation (p v ), thereby generating the pulmonary venous circulation pressure (p pv ) and the pulmonary venous circulation blood flow rate (q pv ), and the pulmonary arterial circulation (pa) and the pulmonary venous circulation (p v ) are connected in series, The pulmonary venous circulation blood pressure (p pv ) and the pulmonary venous circulation blood flow (q pv ) pass through the left atrium (LA) to generate a left ventricular blood pressure (p LV ) and a left ventricular blood flow (q m ); The left ventricular blood pressure (p LV ) and the left ventricular blood flow (q m ) pass through the left ventricle (LV), thereby generating an aortic systemic blood pressure (p sa ) ​​and an aortic systemic blood flow (q sa ); The aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​pass through the aortic systemic circulation (sa) and then through the venous systemic circulation (sv), thereby generating the venous systemic blood pressure (p sv ) and the venous systemic blood flow rate (q sv ), and the aortic systemic circulation (sa) and the venous systemic circulation (sv) are connected in series. having a structure, The right atrium (RA) has the venous systemic blood pressure (p sv ) and the venous systemic blood flow rate (q sv ) that pass through a right atrial resistance (R RA ) and then through a tricuspid valve (atrioventricular valve) (t.v.), the right atrial resistance (R RA ) and the tricuspid valve (atrioventricular valve) (t.v.) are connected in series, and the venous systemic blood pressure (p sv ) and the venous systemic blood flow rate (q sv ) are affected by a right atrial time-varying elastance concept (E RA ) or a right atrial myocardial fiber stress and strain concept (MF RA ) that are connected in parallel to the right atrium (RA) before passing through the right atrial resistance (R RA ), The right ventricle (R.V.) has the right ventricular blood pressure (p RV ) and the right ventricular blood flow rate (q t ) that pass through a right ventricular resistance (R RV ) and then pass through a pulmonary valve (p.v.), the right ventricular resistance (R RV ) and the pulmonary valve (p.v.) being connected in series, and the right ventricular blood pressure (p RV ) and the right ventricular blood flow rate (q t ) being affected by a right ventricular time-varying elastance concept (E RV ) or a right ventricular myocardial fiber stress and strain concept (MF RV ) that are connected in parallel to the right ventricle (R.V.) before passing through the right ventricular resistance (R RV ); The pulmonary artery circulation (pa) has the pulmonary artery circulation pressure (ppa) and the pulmonary artery circulation blood flow rate (qpa), which pass through a pulmonary artery circulation resistance (Rpa) and then through a pulmonary artery circulation inertial resistance (Lpa), the pulmonary artery circulation resistance (Rpa) and the pulmonary artery circulation inertial resistance (Lpa) being connected in series, and the pulmonary artery circulation pressure (ppa) and the pulmonary artery circulation blood flow rate (qpa) being affected by a pulmonary artery circulation compliance (Cpa) connected in parallel to the pulmonary artery circulation (pa) before passing through the pulmonary artery circulation resistance (Rpa), The pulmonary venous circulation (pv) has the pulmonary artery circulation pressure (ppa) and the pulmonary artery circulation blood flow rate (qpa) that pass through a pulmonary venous circulation resistance (Rpv), the pulmonary venous circulation resistance (Rpv) is connected in series with the pulmonary artery circulation pressure (ppa) and the pulmonary artery circulation blood flow rate (qpa), and the pulmonary artery circulation pressure (ppa) and the pulmonary artery circulation blood flow rate (qpa) are affected by a pulmonary venous circulation compliance (Cpv) that is connected in parallel with the pulmonary venous circulation (pv) before passing through the pulmonary venous circulation resistance (Rpv), The left atrium (LA) has the pulmonary venous circulation pressure (p pv ) and the pulmonary venous circulation blood flow rate (q pv ) that pass through a left atrial resistance (R LA ) and then through a mitral valve (atrioventricular valve) (m.v.), the left atrial resistance (R LA ) and the mitral valve (atrioventricular valve) (m.v.) are connected in series, and the pulmonary venous circulation pressure (p pv ) and the pulmonary venous circulation blood flow rate (q pv ) are affected by a left atrial time-varying elastance concept (E LA ) or a left atrial myocardial fiber stress and strain concept (MF LA ) that are connected in parallel to the left atrium (LA) before passing through the left atrial resistance (R LA ), The left ventricle (L.V.) has the left ventricular blood pressure (p LV ) and the left ventricular blood flow rate (q m ) which pass through a left ventricular resistance (R Lv ) and then through an aortic valve (ventricular valve) (a.v.), the left ventricular resistance (R Lv ) and the aortic valve (ventricular valve) (a.v.) being connected in series, and the left ventricular blood pressure (p LV ) and the left ventricular blood flow rate (q m ) being affected by a left ventricular time-varying elastance concept (E Lv ) or a left ventricular myocardial fiber stress and strain concept (MF Lv ) which are connected in parallel to the left ventricle (L.V.) before passing through the left ventricular resistance (R Lv ), The aortic systemic circulation (sa) has aortic systemic blood pressure (p sa ) ​​and aortic systemic blood flow rate (q sa ) ​​which pass through aortic systemic circulation resistance (R sa ) ​​and then through aortic systemic circulation inertial resistance (L sa ), the aortic systemic circulation resistance (R sa ) ​​and the aortic systemic circulation inertial resistance (L sa ) ​​are connected in series, and the aortic systemic blood pressure (p sa ) ​​and aortic systemic blood flow rate (q sa ) ​​are affected by aortic systemic circulation compliance (C sa ) ​​connected in parallel to the aortic systemic circulation (sa) before passing through the aortic systemic circulation resistance (R sa ), The venous systemic circulation (sv) has the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​that pass through a venous systemic circulation resistance (R sv ), the venous systemic circulation resistance (R sv ) is connected in series with the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ), and the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​are affected by a venous systemic circulation compliance (C sv ) that is connected in parallel with the venous systemic circulation (sv) before passing through the venous systemic circulation resistance (R sv ). method.

9. 6. The method of claim 5, wherein the central compartment lumped parameter model comprises a single closed loop circuit of the systemic circulation; The inertia and elasticity functional building block is C ｉ－１ =C ｓａ , R ｉ =R ｓａ , L ｉ =L ｓａ is determined by The resistive-capacitive functional building block is C ｉ－１ =C ｓｖ , R ｉ =R ｓｖ , L ｉ = 0, method.

10. 6. The method of claim 5, wherein the central compartment lumped parameter model has the following structure: The venous systemic blood pressure (p sv ) and venous systemic blood flow (q sv ) pass through the left atrium (LA), generating left ventricular blood pressure (p LV ) and left ventricular blood flow (q m ); The left ventricular blood pressure (p LV ) and the left ventricular blood flow (q m ) pass through the left ventricle (LV), thereby generating an aortic systemic blood pressure (p sa ) ​​and an aortic systemic blood flow (q sa ); The aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​pass through the aortic systemic circulation (sa) and then through the venous systemic circulation (sv), thereby generating the venous systemic blood pressure (p sv ) and the venous systemic blood flow rate (q sv ), and the aortic systemic circulation (sa) and the venous systemic circulation (sv) are connected in series. having a structure, The left atrium (LA) has the venous systemic blood pressure (p sv ) and the venous systemic blood flow rate (q sv ) that pass through a left atrial resistance (R LA ) and then through a mitral valve (atrioventricular valve) (m.v.), the left atrial resistance (R LA ) and the mitral valve (atrioventricular valve) (m.v.) are connected in series, and the venous systemic blood pressure (p sv ) and the venous systemic blood flow rate (q sv ) are affected by a left atrial time-varying elastance concept (E LA ) or a left atrial myocardial fiber stress and strain concept (MF LA ) that are connected in parallel to the left atrium (LA) before passing through the left atrial resistance (R LA ), The left ventricle (L.V.) has the left ventricular blood pressure (p LV ) and the left ventricular blood flow rate (q m ) which pass through a left ventricular resistance (R Lv ) and then through an aortic valve (ventricular valve) (a.v.), the left ventricular resistance (R Lv ) and the aortic valve (ventricular valve) (a.v.) being connected in series, and the left ventricular blood pressure (p LV ) and the left ventricular blood flow rate (q m ) being affected by a left ventricular time-varying elastance concept (E Lv ) or a left ventricular myocardial fiber stress and strain concept (MF Lv ) which are connected in parallel to the left ventricle (L.V.) before passing through the left ventricular resistance (R Lv ), The aortic systemic circulation (sa) has the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​which pass through an aortic systemic circulation resistance (R sa ) ​​and then through an aortic systemic circulation inertial resistance (L sa ), the aortic systemic circulation resistance (R sa ) ​​and the aortic systemic circulation inertial resistance (L sa ) ​​being connected in series, and the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​are affected by an aortic systemic circulation compliance (C sa ) ​​connected in parallel to the aortic systemic circulation (sa) before passing through the aortic systemic circulation resistance (R sa ), The venous systemic circulation (sv) has the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​that pass through a venous systemic circulation resistance (R sv ), the venous systemic circulation resistance (R sv ) is connected in series with the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ), and the aortic systemic blood pressure (p sa ) ​​and the aortic systemic blood flow rate (q sa ) ​​are affected by a venous systemic circulation compliance (C sv ) that is connected in parallel with the venous systemic circulation (sv) before passing through the venous systemic circulation resistance (R sv ). method.

11. 6. The method of claim 5, wherein the central compartment lumped parameter model satisfies the following conditions: [Equation 25] A method comprising:

12. 6. The method of claim 5, wherein the central compartment lumped parameter model has a heart chamber pressure-volume relationship formulated using a variable elastance concept (E).

13. 6. The method of claim 5, wherein the central compartment lumped parameter model has a cardiac chamber pressure-volume relationship formulated using myocardial fiber stress and strain concepts (MF).

14. 2. The method of claim 1, wherein the step (4) of identifying the parameters of the coupled system comprises: calculating initial values ​​of empirical parameters of the combined system using demographic and health data of the patient; solving the equations of the coupled system using initial values ​​of the empirical parameters to calculate an approximation of the central arterial pressure and / or the proximal blood flow within a selected time window; (4c) iteratively refining the empirical parameter values ​​of the coupled system for constant or varying values ​​of the central arterial pressure and / or the proximal blood flow until convergence is reached; A method comprising:

15. 15. The method of claim 14, wherein the step of calculating initial values ​​of the empirical parameters of the coupled system is performed only for a central compartment lumped parameter model of the blood circulatory system.

16. 15. The method of claim 14, wherein the empirical parameters comprise compliance, resistance and / or inertia and / or parameters of the time-varying elastance concept or parameters of the myocardial fiber stress and strain concept.

17. 15. The method of claim 14, wherein the iterative refinement step comprises at least one minimization algorithm; A method wherein each of the minimization algorithms is selected from the group consisting of a local minimization algorithm and a global minimization algorithm.

18. 18. The method of claim 17, wherein the local minimization algorithm is selected from the group consisting of Nelder-Mead, recursive least squares programming, and Broyden-Fletcher-Goldfarb-Shanno.

19. 17. The method of claim 16, wherein the global minimization algorithm is selected from the group consisting of: adaptive memory programming for global optimization, and global optimization by simplicial homology.

20. 17. The method of claim 16, wherein the iterative refinement step comprises a refinement step that implements the global minimization algorithm followed by a refinement step that implements a local minimization algorithm.

21. 10. A computer readable storage medium having instructions that, when executed by a computer, cause the computer to perform the steps of the method of claim 1.

22. 1. A system for reconstructing central arterial pressure waveform morphology from continuous non-invasive distal blood pressure measurements, comprising: a means for non-invasively and continuously recording blood pressure waveforms from a human patient; a measuring means for non-invasively measuring the patient's systolic blood pressure, diastolic blood pressure, and heart rate; a computer configured to perform the steps of the method of claim 1; A system having:

Images