Method for evaluating hemodynamic response to adenosine receptor agonist stimulation, evaluation system thereof, and computer readable medium

Abstract

A method, computer readable medium, and system for evaluating the hemodynamic response to adenosine receptor agonist stimulation in a human patient are disclosed. The starting point of the method is non-invasive pressure data in the resting state. The said data is used for pharmacokinetic / pharmacodynamic modeling to obtain an individual patient-specific response to adenosine receptor stimulation by an agonist. This response is used to calibrate a lumped parameter Windkessel-type model of blood circulation. The calibrated model allows the calculation of hemodynamic parameters in the hyperemic state, in particular the flow and pressure values ​​at both coronary artery ostia. The hemodynamic parameters allow in silico coronary vascular diagnosis. The effectiveness of the developed method has been confirmed in a multicenter clinical trial.

Description

[Technical field]

[0001] The present invention relates to a method for assessing hemodynamic response to adenosine receptor agonist stimulation in a human patient, a computer-readable medium containing instructions that, when executed by a computer, cause the computer to carry out said method for assessing hemodynamic response to adenosine receptor agonist stimulation in a human patient, and a system for assessing hemodynamic response to adenosine receptor agonist stimulation in a human patient. In particular, the present invention relates to estimating hemodynamic parameters of the coronary circulation of a patient in a hyperemic state by modeling the individual patient's specific response to adenosine receptor stimulation by an agonist. The present invention allows for the diagnosis and treatment of coronary insufficiency. In particular, it allows for the diagnosis and treatment of coronary heart disease. [Background technology]

[0002] According to the National Center for Health Statistics, 5.6% of US adults (defined as age 18 and older) had coronary heart disease (CHD) in 2018. Men are more likely to have CHD than women, with 7.4% of US adults having CHD compared with 4.1% of women. Risk increases with age, with 15.5% of people aged 65-74 and 23.9% of people aged 75 and older having CHD (Villarroel MA, Blackwell DL, Jen A (2019) Tables of Summary Health Statistics for US Adults: 2018 National Health Interview Survey, National Center for Health Statistics, Accessed: 2021.12.17). In 2018, 365,744 people died from coronary heart disease in the United States, accounting for approximately 13% of all deaths recorded in 2018 (see Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System: public use data file documentation: mortality multiple cause-of-death micro-data files, accessed 17 December 2021). In 2017, 623,276 people died from ischemic heart disease in the European Union, accounting for approximately 12% of all deaths recorded in 2017 (see Eurostat (2021) Causes of death - deaths by country of residence and occurrence [HLTH_CD_ARO], accessed 17 December 2021).

[0003] The diagnostic pathway for coronary insufficiency is highly complex and multifaceted. Except for the initial stage, each of the following key stages has progressively more reliability, which may be related to the increased invasiveness of the subsequent key stages and the increased risk for patients undergoing the diagnostic pathway (see Scanlon PJ et al. (1999) ACC / AHA Guidelines for Coronary Angiography: Executive Summary and Recommendations, Circulation, 99(17), p. 2345-2357; ESC Scientific Document Group (2020) 2019 ESC Guidelines for the determination and management of chronic coronary syndromes, Eur Heart J, 41(3), p. 407-477). The diagnostic accuracy pyramid has been comprehensively shown to have its apex and gold standard, fractional coronary reserve in the form of fractional flow reserve (FFR) (see Johnson NP et al. (2016) Continuum of Vasodilator Stress from Rest to Contrast Medium to Adenosine Hyperemia for Fractional Flow Reserve Assessment, JACC Cardiovasc Interv, 9(8), p. 757-767). FFR is described as the base ratio of distal and proximal pressures, and measurements of each pressure must be performed in hyperemic conditions. As written in an excellent paper by the discoverers of the FFR ratio, Nico Pijls and Bernard De Bruyne, "meaningful measurements of coronary pressure can only be performed in maximal hyperemic conditions" (see Pijls NH, Bruyne BD(1997)Coronary Pressure, Springer-Science+Business Media). This quote does not only apply to FFR, but also to the accuracy of any CHD diagnostic method. The accuracy of ratios such as FFR and hyperemia ratio comfortably exceeded the 95% threshold.At the same time, the calculated rate at rest is only slightly above 80%, while visual assessment, including angiography, is less than 70%. As a general rule, reliable functional assessment is only possible in hyperemic conditions.

[0004] The hyperemic state in functional assessment is achieved by stimulation of adenosine receptors (AR) by drugs. There are many over-the-counter drugs that act as adenosine receptor agonists. Adenosine as a drug is sold under the trade names Adenocard, Adenocor, Adenic, Adenoco, Adeno-Jec, and Adenoscan. An example of the FDA dosage and administration of Adenoscan is that the recommended dose for adults is 140 μg / kg / min, administered as a continuous peripheral intravenous infusion for 6 minutes (taken from the US Food and Drug Administration. Labeling for FDA-approved drugs - NDA 020059. Accessed: 2021.12.17). Regadenoson is sold under the trade names Lexiscan or Rapiscan, Binodenoson is sold under the trade name CorVue, and Apadenoson is sold under the trade name Stedivaze. An example of the dosage of these products is 5 ml of Lexiscan (0.4 mg of regadenoson) injected intravenously within 10 seconds, followed by a radiopharmaceutical and saline flush (taken from the US Food and Drug Administration. Labeling of FDA approved drug - NDA 022161. Access date: 2021.12.17). According to the DrugBank database, phase III of clinical trials for apadenoson has ended and phase III of clinical trials for binodenoson has been completed (see DrugBank Online. Access date: 2021.12.17).

[0005] Adenosine receptors are transmembrane proteins that belong to the P1 class of G protein-coupled purinergic receptors. G proteins are heterotrimeric protein complexes composed of three subunits, α, β, and γ. The stable dimeric complex of this protein is composed of α and β subunits. The adenosine receptors are classified as A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A19, A119, A111, A112, A13, A14, A15, A16, A17, A18, A19 ...1, A111, A111, A111, A112, A13, A13, A14, A15, A16, A17, A18, A19, A19, A111, A111, A111, A111, A111, A111, A111, A111 2A , A 2BEach receptor is composed of seven transmembrane domains. The number of amino acids differs for each receptor, but the A 2A Adenosine receptors are the longest (see Borea et al. (2018) Pharmacology of Adenosine Receptors: The State of the Art, Physiological Reviews, 98 (3), p. 1591-1625; Sheth S et al. (2014) Adenosine receptors: expression, function and regulation, International Journal of Molecular Sciences, 15(2), p. 2024-2052). All transmembrane domains consist of 21-28 amino acids that form a helix. The C-terminal side of the adenosine receptor is located in the cytoplasmic region, and the N-terminal side is located on the extracellular side of the cell membrane (Vincenzi M, Bednarska K, Le?nikowski ZJ (2018) Comparative Study of Carborane - and Phenyl-Modified Adenosine Derivatives as Ligands for the A 2A and A3Adenosine Receptors Based on a Rigid in Silico Docking and Radioligand Replacement Assay, Molecules (Basel, Switzerland), 23(8), p. 1846).

[0006] Adenosine receptors are found in the nervous system, cardiac system, pulmonary system, digestive system, urinary system, and immune system (see Borea et al. (2018) Pharmacology of Adenosine Receptors: The State of the Art, Physiological Reviews, 98 (3), p. 1591-1625). Each type of adenosine receptor (A1, A2, A3, A4, A5, A6, A7, A8, A9, A10, A11, A12, A13, A14, A15, A16, A17, A18, A19, A21, A22, A23, A24, A25, A26, A27, A28, A29, A30, A31, A32, A33, A34, A35, A36, A37, A38, A39, A41, A42, A43, A 2A , A 2BActivation of adenosine receptors A1 and A3 can affect cardiac function. The A1 adenosine receptor has the highest affinity for adenosine. Activation of this receptor slows the heart rate, constricts blood vessels, and inhibits the release of neurotransmitters. Adenosine receptor A 2A A is the largest of all adenosine receptors. 2A AR mRNA is most abundant in the heart. 2A When AR is activated, cAMP levels increase, causing vasodilation and inhibition of platelet aggregation. 2B A high concentration of adenosine is required to activate A. 2B Adenosine receptor expression is low. 2B Activation of the AR leads to increased cAMP levels, vasodilation, and cardiac constriction. A3 adenosine receptors are most abundant in the liver and less abundant in the aorta and heart. Although their exact location in the heart is unknown, activation of these receptors has a cardioprotective effect. Activation of the A3AR also leads to inhibition of adenylyl cyclase (see Shryock JC, Belardinelli L. (1997) Adenosine and adenosine receptors in the cardiovascular system: biochemistry, physiology, and pharmacology. Am J Cardiol . 19;79(12A):2-10; Wilson C, Mustafa SJ (2009) Adenosine Receptors in Health and Disease. Springer Nature; Guieu R et al. (2020) Adenosine and the cardiovascular system: The good and the bad. J. Clin. Med. 9(1366):1-21).

[0007] Most of the adenosine receptor agonists are purine nucleoside derivatives, adenosine or xanthosine. Alkylxanthine derivatives are the prototype antagonists of the adenosine receptor. Several moieties of adenosine can potentially interact with the five nitrogen atoms of the adenine group of the amino acid of the receptor (i.e., N1, N3, N6, N7 and N9) and the three hydroxyl groups of the ribose moiety (i.e., 2', 3' and 5'). There are many different agonists for the adenosine receptor. Substitution of adenosine can change the selectivity of the receptor. Substitution at the N6 position with arylalkyl, cycloalkyl and alkyl groups changes the selectivity of the A1 adenosine receptor. Substitution at the 2 position of adenosine with (thio)ethers, alkynes and secondary amines has led to the synthesis of new A1 adenosine agonists. 2A Adenosine-selective analogs are formed. Substitution of the N6 position of adenosine with selected aryl groups results in the formation of A 2B The affinity for the adenosine receptors is improved. Improved selectivity for the A3 adenosine receptor can be achieved by substituting with an N6-benzyl or substituted benzyl group. Adenosine receptor antagonists can be obtained in a variety of ways. High affinity and selectivity for the A1 adenosine receptor can be achieved by modifying the xanthine core structure at position 8 to an aryl or cycloalkyl. A 2A Adenosine receptor selectivity can be achieved by modifying the 8-position of the xanthine with an alkene. 2B Adenosine receptor selectivity can be achieved by modifying the xanthine core structure at position 8 with an aryl group. A number of non-xanthine classes have been defined as A3 adenosine receptor antagonists (Muller CE, Jacobson KA (2011) Recent developments in adenosine receptor ligands and their potential as novel drugs, Biochimica et Biophysica Acta (BBA) - Biomembranes, 1808(5), p. 1290-1308). 2AAdenosine A has an orthosteric binding site that can be found through its involvement in binding agonist and antagonist residues (Carpenter B, Lebon G (2017) Human Adenosine A 2A Receptor: Molecular Mechanism of Ligand Binding and Activation, Frontiers in Pharmacology, 8, p. 898). Caffeine is a 2A It is a natural antagonist of adenosine receptors. Studies have shown that caffeine affects the cardiovascular system, but the results of those studies are inconclusive. The cardiovascular response after caffeine ingestion depends on many factors, including the amount ingested, the time of ingestion, frequency, liver metabolism, and the level of absorption. The effects of caffeine also vary depending on gender, and the differences may be related to the concentrations of steroid hormones (see Muller CE, Jacobson KA (2011) Recent developments in adenosine receptor liposine and their potential as novel drugs, Biochimica et Biophysica Acta (BBA) - Biomembranes, 1808 (5), p. 1290-1308). Based on information from the Protein Data Bank, caffeine binds to the A receptor using atoms O13 and N7 and hydrogen bonds, and atom N3 and cation-pi interactions. 2A Regadenoson binds to adenosine receptors. 2AIt is an adenosine receptor agonist. Regadenoson is metabolized more slowly than adenosine in plasma. Regadenoson may increase coronary blood flow in a dose-dependent manner (Muller CE, Jacobson KA (2011) Recent developments in adenosine receptor liposine and their potential as novel drugs, Biochimica et Biophysica Acta (BBA)- Biomembranes, 1808 (5), p. 1290-1308).

[0008] Other Selection A 2A Adenosine receptor agonists include binodenoson (BND) and apadenoson (APA). Apadenoson is more selective than binodenoson. In this comparison, regadenoson is the least selective. APA and BND have equal affinity for adenosine receptors. The average onset of binding action is 1-2 minutes. Apadenoson's duration of action is longer than bidadenoson's, 10-20 minutes. Binodenoson's duration is less than 5 minutes. The dosage of APA and BND is calculated according to the patient's weight. Both adenosine agonists are introduced by intravenous bolus administration (see Elkholy KO et al. (2021) Regadenoson Stress Testing: A Comprehensive Review with a Focused Update, Cureus, 13(1), p. e12940).

[0009] Adenosine receptors are involved in the regulation of vascular tone, a process that is not fully understood. 2A Adenosine receptor activation has been shown to regulate vascular tone, but 2B It has also been demonstrated that adenosine receptors are involved in the relaxation of small arteries. Adenosine receptors interact with each other to regulate coronary vasculature. Regulation of coronary microcirculation via adenosine can be either endothelium-dependent or endothelium-independent. 2ANitric oxide is involved in the regulation of basal tone via adenosine receptors. It also mediates adenosine-mediated A 2A Part of AR activation and A in reactive hyperemia 2A AR-K ATP It is also part of the line (see Zhang Y et al. (2021) Adenosine and adenosine receptor-mediated action in coronary microcirculation, Basic Research in Cardiology, 12(116)).

[0010] Due to the importance and role of adenosine receptors, the pharmacokinetics and pharmacodynamics of their endogenous and exogenous agonists and antagonists have been the subject of study. To date, adenosine has been the most studied and investigated. Kern et al. have tested adenosine as a measure of coronary vasodilator reserve. During the study, adenosine was infused intravenously for 3 minutes at doses of 50, 100, and 150 μg / kg / min. These studies included 34 patients, 17 of whom had coronary artery disease (CAD) and 17 of whom did not (Kern MJ et al. (1991) Intravenous adenosine: continuous infusion and low dose bolus administration for determination of coronary vasodilator reserve in patients with and without coronary artery disease, Journal of the American College of Cardiology, 18(3), p. 718-29). Alexopoulos et al. found that increasing the dose of adenosine d / P a Thirty-eight patients received a standard dose of adenosine (140 μg / kg / min) via the femoral vein and had a P d / P aThe evaluation was repeated. Eight patients refused to participate in the high-dose adenosine trial. Thirty patients received high-dose adenosine (200 μg / kg / min) and had no significant improvement in P d / P aThe evaluation was repeated. Patients had stable coronary artery disease or presented with acute coronary syndrome under certain conditions. Mean systolic blood pressure was reduced to a greater extent after high-dose adenosine 200 μg / kg / min compared to standard adenosine dose of 140 μg / kg / min. Mean diastolic blood pressure was reduced to lower values ​​after high-dose adenosine compared to treatment with standard adenosine dose. Heart rate (HR) was increased to a greater extent after high-dose adenosine infusion (see Alexopoulos D et al. (2016) Effect of High (200 μg / kg per Minute) Adenosine Dose Infusion on Fractional Flow Reserve Variability, Journal of the American Heart Association, 5). The hemodynamic effect of adenosine was measured using cardiovascular magnetic resonance imaging in a study conducted by Thomas et al. Adenosine was administered intravenously to 25 healthy patients for 6 minutes at a dose of 140 μg / kg / min. The results showed a slight decrease in mean systolic blood pressure after adenosine injection, but no change in mean diastolic blood pressure, and a significant increase in heart rate after adenosine injection (see Thomas D et al. (2017) Effects of adenosine and regadenoson on hemodynamics measured using cardiovascular magnetic Resonance Imaging, Journal of Cardiovascular Magnetic Resonance, 19). Another study, in which researchers infused a regular dose of adenosine for 6 minutes and studied its effects on hemodynamics, was conducted by Mishra et al. This study involved 348 healthy volunteers. The results obtained showed that the changes in mean systolic and mean diastolic blood pressure were very small.Only heart rate was significantly increased (according to Mishra R et al. (2005) Quantitative relationship between hemodynamic changes during intravenous adenosine infusion and the amplitude of coronary hyperemia - Implications for myocardial perfusion Implications, Journal of the American College of Cardiology, 45, p. 553-8). Lee et al. tested the use of adenosine in 15 healthy volunteers as an alternative to exercise testing (known as pharmacological stress testing). Adenosine was infused intravenously for 6 minutes at a dose of 140 μg / kg / min. Consistent with previous tests, there was a small decrease in mean systolic blood pressure. The mean diastolic blood pressure was slightly decreased and the heart rate was significantly increased (for details, see Lee JH et al. (1994) Biokinetics of thallium-201 in normal subjects: comparison between adenosine, dipyridamole, dobutamine and exercise, Journal of nuclear medicine: official publication, Society of Nuclear Medicine, p. 535-41). Experimental results show that it is not possible to mathematically model the effect of adenosine receptor stimulation, which is necessary to estimate the effect of adenosine on cardiac hemodynamics, based on simple relationships. In all tests performed, the heart rate increased, but at different values. At the same time, the difference in HR between the group with coronary artery disease (CAD) and the group without CAD was not significant. The results for systolic blood pressure were equivocal. In each test, there were differences in the values ​​of systolic blood pressure at baseline and peak hyperemia after adenosine injection.These results showed that systolic blood pressure decreased after adenosine infusion, whereas in the experiment performed by Brown et al., systolic blood pressure did not decrease. (See Brown L et al. (2021) A comparison of standard and high dosage adenosine protocols in routine vasodilator stress cardiovascular magnetic resonance: dosage effects hyperemic myocardial blood flow in patients with serious left ventricular systolic impairment, Journal of Cardiovascular Magnetic Resonance, 23.) Systolic blood pressure either did not change after adenosine infusion (group of 16 patients with heart failure (HF) ≥ 40% left ventricular ejection fraction (LVEF) - 124 ± 16 at rest, 124 ± 11 at stress) or increased (group of 20 patients with HF LVEF < 40% - 119 ± 16 at rest, 122 ± 18 at stress). The results showed either a decrease in diastolic blood pressure or no change in diastolic blood pressure values. Vasu et al. tested adenosine, regadenoson, and dipyridamole in 15 healthy volunteers. The relative potency of these vasodilators was assessed using quantified stress and resting myocardial perfusion. Results were obtained using cardiovascular magnetic resonance (CMR). Stress tests were performed within a few days after administration of a 400 μg bolus of regadenoson, 0.56 mg / kg dipyridamole, and 140 μg / kg / min adenosine. Each test began with the recording of a resting perfusion image. After 20 min, vasodilatory stress images were recorded at peak times, i.e., 3–4 min after adenosine injection, 4 min after dipyridamole injection, and 70 s after regadenoson injection. Myocardial blood flow (MBF) and myocardial perfusion reserve (MPR) were quantified by fully quantitative model-constrained deconvolution. Higher values ​​of MBF and heart rate (HR) during stress were recorded after regadenoson injection compared with those recorded with dipyridamole or adenosine.After adjusting the stress MBF values ​​to the HR values, there was no difference between adenosine and regadenoson, but there was still a difference between dipyridamole and regadenoson. The unadjusted MPR was higher for regadenoson than for adenosine and dipyridamole. After adjusting the stress MPR values ​​to the HR values, there was no difference between regadenoson and adenosine, but there was still a difference between regadenoson and dipyridamole and between adenosine and dipyridamole (see Vasu S et al. (2013) Regadenoson and adenosine are equal vasodilators and are superior than dipyridamole- a study of first pass quantitative perfusion cardiovascular magnetic resonance, Journal of cardiovascular magnetic resonance: official journal of the Society for Cardiovascular Magnetic Resonance, 15, p. 85). The mean results of the various clinical trials for systolic blood pressure (SYS), diastolic blood pressure (DIA), heart rate (HR), and coronary artery disease (CAD) are given in the cited literature.

[0011] Non-invasive diagnostic methods for circulatory failure based on computational fluid dynamics (CFD) technology are gradually gaining trust in the medical field, despite some issues. The current results of validation studies do not provide clear recommendations (Nakanishi R, Budoff MJ (2016) Noninvasive FFR derived from coronary CT angiography in the management of coronary artery disease: technology and clinical update, Vasc Health Risk Manag, 12, p. 269-78, Kruk M et al. (2016) Workstation-based calculation of CTA-based FFR for intermediate stenosis, JACC Cardiovasc Imagin, 9(6), p. 690-9, Chung JH et al. (2017) Diagnostic performance of a novel method for fractional flow reserve computed from noninvasive computed tomography angiography (NOVEL-FLOW Study), Am J Cardiol, 120(3), p. 362-368, Wang ZQ et al. (2019) Diagnostic accuracy of a deep learning approach to calculate FFR from coronary CT angiography, J Geriatr Cardiol, 16(1), p. 42-48, Fujimoto S et al.(2019) Diagnostic performance of on-site computed CT-fractional flow reserve based on fluid structure interactions: comparison with invasive fractional flow reserve and instantaneous wave-free ratio, Eur Heart J Cardiovasc Imaging, 20(3), p. 343-352, Tang CX et al. (2020) CT FFR for ischemia-specific cad with a new computational fluid dynamics algorithm: A Chinese multicenter study, JACC Cardiovasc Imaging, 13(4), p. 980-990, Nozaki YO et al. (2021) Comparison of diagnostic performance in on-site based CT-derived fractional flow reserve measurements, Int J Cardiol Heart Vasc, 35, p. 1-5)。.

[0012] Even when limited to in silico evaluation platforms, the sensitivity, understood as true positive rate (TPN), and the specificity, understood as true negative rate (TNR), reach satisfactory values ​​in a single clinical trial (TPR / TNR=93.2% / 82.9%, followed by 74% / 67% and 85% / 79%). The results obtained in similar trials can be questionable in the so-called "gray zone". This can be better understood using an example. For example, a few years ago, a team from Imperial College London published a review article that showed that "for vessels with FFR-CT values ​​of <0.60, 0.60-0.70, 0.70-0.80, 0.80-0.90, and ≥0.90, the diagnostic accuracy of FFR-CT was 86.4% (95% CI, 78.0%-94.0%), 74.7% (95% CI, 71.9%-77.5%), 46.1% (95% CI, 42.9%-49.3%), 87.3% (95% CI, 85.1%-89.5%), and 97.9% (95% CI, 97.9%-98.8%), respectively" (Cook CM et al. (2017) Diagnostic Accuracy of Computed Tomography-Derived Fractional Flow Reserve: A Systematic Review, JAMA Cardiol, 2(7), pp. 803-810).

[0013] This review article has generated controversy and debate. Above all, it can be concluded that the authors' assessment was not very accurate, since a 50% accuracy is the diagnostic accuracy equivalent to flipping a coin. This level of accuracy is not satisfactory.

[0014] This ambivalent view of in silico techniques is likely due to the fact that their development is still at a relatively early stage. Most of the developments of in silico techniques are still in the research phase and need more time to be translated into commercially developed products. Indeed, there are still many elements of these developments that need to be improved. The most promising approach concerns the invasive fractional coronary flow reserve (FFR), as FFR is currently described as the gold standard test. Nico Pijls and Bernard De Bruyne, authors of the already mentioned FFR index concept, write in their research summary in the paper: "...there are three reasons why coronary pressure measurements have not been useful for blood flow assessment in the past. First, the first prerequisite for reliable intracoronary pressure recording is the use of very thin pressure monitoring guidewires. Second, pressure measurements are only meaningful at maximum coronary hyperemia. Third, it is not the gradient that determines myocardial perfusion, but the residual distal coronary pressure." (Pijls NHJ, de Bruyne B (2000) Coronary Pressure, Springer-Science+Business Media, p.52). Probably, in the coming years, there will be development of recommendations for the improvement of in silico techniques. Today, the first recommendation concerns measuring instruments, which in computer technology should be translated as an increase in the resolution of medical images and discrete geometric models. This recommendation is well established without raising any major doubts. Moreover, the second recommendation concerns improvements towards the description of very important physiological situations. In particular, those that may increase or decrease the reliability of the measurements and, as a result, cause the coronary insufficiency state to not be fully indicated. The third recommendation seems expected and is also uncontroversial. This recommendation has to do with what should be evaluated or measured as a diagnostic indicator. As with invasive procedures, in silico, fractional flow reserve is calculated by relating distal and proximal pressures. Curiously, this proximal pressure, while essential for all clinicians, is treated differently by everyone working with in silico techniques. The above detailed prior art review indicates a need for reliable computational methods for assessing the status of the coronary artery system of a human patient. The invention disclosed and claimed herein addresses this need. Summary of the Invention [Means for solving the problem]

[0015] 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. The present invention approaches the problem of assessment of the coronary system by adequately describing the hyperemic state. The inventors have concluded that a reliable assessment of the function of the coronary system can only be performed if it is known how the coronary system behaves under hyperemic conditions. Since a hyperemic state is induced by stimulation of adenosine receptors using an adenosine receptor agonist, an assessment of the patient's hemodynamics after exposure to the adenosine receptor agonist is necessary for a reliable assessment of function. The present invention models the patient's response to adenosine receptor agonist stimulation.

[0016] In one aspect, the invention relates to a method for assessing the hemodynamic response to adenosine receptor agonist stimulation in a human patient, which method may be embodied as a computer-implemented invention.

[0017] The method according to the invention involves estimating the response of one or more of these adenosine receptors to stimulation by one or more agonists, using a pharmacokinetic / pharmacodynamic (PK / PD) model to arrive at parameters of systolic and diastolic blood pressure, and heart rate under hyperemic conditions, which is performed in a step called mapping.

[0018] To map resting hemodynamic parameters describing a resting human patient to hyperemic hemodynamic parameters describing a hyperemic human patient using said pharmacokinetic / pharmacodynamic model, the pharmacokinetic / pharmacodynamic model must be a nonlinear model. The pharmacokinetic / pharmacodynamic model must also include Michaelis-Menten, Hill-Langmuir, Black-Leff, multi-pathway and / or multi-species blood-tissue exchange models (for details, see e.g. Bassingthwaighte JB, Wang CY, Chan IS (1989) Blood-tissue exchange via transport and transformation by capillary endothelial cells, Circ Res, 65(4):997-1020; Kroll K, Deussen A, Sweet IR (1992) Comprehensive model of transport and metabolism of adenosine and S-adenosylhomocysteine ​​in the guinea pig heart, Circ Res, 71(3):590-604). The use of at least one model, for example Hill-Langmuir, is sufficient for the present invention. Using a combination of models may provide better results in some cases.

[0019] The hemodynamic parameters at rest include the patient's systolic blood pressure, the patient's diastolic blood pressure, and / or the patient's heart rate. In one embodiment, the hemodynamic parameters at rest consist only of the patient's systolic blood pressure, the patient's diastolic blood pressure, and the patient's heart rate. For example, if the hemodynamic parameters at rest consist only of the patient's systolic blood pressure at rest, the patient's diastolic blood pressure at rest, and the patient's heart rate at rest, the hemodynamic parameters at hyperemia consist only of the patient's systolic blood pressure at hyperemia, the patient's diastolic blood pressure at hyperemia, and the patient's heart rate at hyperemia.

[0020] These resting parameters can be measured directly from the patient in a non-invasive manner or obtained from non-invasive measurements previously made on the patient. Alternatively, the resting hemodynamic parameters can be estimated using a non-invasively recorded continuous patient pressure waveform. The waveform may or may not be recorded simultaneously with the resting hemodynamic parameters. To reliably estimate the resting hemodynamic parameters, the waveform must be recorded for a time window that includes at least one full cardiac cycle of the patient. The more time windows, the better the results. To be able to estimate the resting hemodynamic parameters, the waveform must represent a resting state and not be a waveform recorded for a hyperemic state. In another embodiment, the resting hemodynamic parameters further include a non-invasively recorded continuous resting patient pressure waveform.

[0021] In one embodiment, the PK / PD model also calculates the hyperemic patient pressure waveform using non-invasively recorded continuous resting patient pressure waveforms, which is less advantageous in terms of calculation pace but is advantageous because the hyperemic patient pressure waveform can be used in the parameter identification step of the method.

[0022] The method according to the invention then comprises obtaining a model describing the hyperemic patient. This is achieved by performing parameter identification of the lumped parameter model using the hyperemic hemodynamic parameters describing the hyperemic human patient. In this step, the empirical parameters of the lumped parameter model are adjusted using the hyperemic hemodynamic parameters describing the hyperemic human patient. The empirical parameters of the lumped parameter model and the whole model then describe the hyperemic patient. In an embodiment, the empirical parameters include parameters of resistance, compliance, inertia, time-varying elasticity concepts and / or myocardial fiber strain / stress concepts. In other embodiments, the empirical parameters further include parameters of myocardial vascular interactions, parameters of blood circulation system (BCS) elements, parameters of ventricular pressure volume (HPV) elements and / or parameters of coronary blood flow (CBF) elements. The empirical parameters can include other additional parameters that are used to adjust the model to the patient and / or that are beneficial in terms of the stability of the model.

[0023] The lumped parameter model must be of Windkessel type, i.e. the model must be built using variations of the blocks developed by Windkessel. The lumped parameter model must at least partially describe the hemodynamics of the patient. This must be understood to include the description of relevant parts of the blood circulatory system of a human patient or the entire blood circulatory system. The level of description of the hemodynamics is adjusted to the information that needs to be provided by the present invention.

[0024] The lumped parameter model representing the hyperemic state obtained using the method according to the invention allows a computer diagnosis of the patient. In particular it allows the calculation of flow parameters which may include pressure, flow and / or cardiac time intervals, including but not limited to cardiac cycle duration and ejection time. Said flow parameters are hyperemic transients which can be used to obtain a reliable diagnosis of a human patient. In particular, said transients can be used to diagnose and treat coronary heart disease. Computer diagnosis is advantageous since it avoids all the drawbacks of subjecting the patient to an invasive procedure.

[0025] The method may include the collection or use of demographic and health data that affect circulatory hemodynamics in the patient's body. The data specifically concerns data that affect pulse propagation in the patient's body. The demographic and health data includes the patient's sex, age, height, general fitness rating, and / or current medications. Current medications include, but are not limited to, beta blockers, angiotensin converting enzyme inhibitors, and / or antiarrhythmic drugs. Current medications may include any medication that may affect the human heart. The data may include data obtained from a database and may be data for any patient. The data is used to obtain a first approximation of the values ​​of the empirical parameters of the lumped parameter model. This reduces the time required to arrive at the result of the method. It may also be beneficial in terms of stability of the calculations. In one embodiment, this is done by finding a relationship between the empirical parameters and the data. The first approximation speeds up the calculations and the method quickly and stably arrives at a lumped parameter model that describes the human patient in a hyperemic state. In another embodiment, when collecting or acquiring data from patients suffering from coronary artery disease, the method arrives at a lumped parameter model that more quickly and stably describes a human patient in a hyperemic state, since the empirical parameters of the model are very close to the parameters describing the patient in the hyperemic state that needs to be diagnosed.

[0026] The method according to the invention may also include the step of non-invasively recording the patient's pressure waveform. This waveform should represent the patient's resting state. The patient's pressure waveform is recorded for a selected time window including at least one cycle of the patient's heart. The patient's pressure waveform may be recorded continuously. It is preferably recorded for the radial artery. Preferred techniques for recording the patient's pressure waveform are finger cuff photoplethysmography and / or applanation tonometry. By non-invasive recording is meant recording without any type of surgery and / or significant health risks. In some embodiments, said recording does not involve the insertion of a probe into the patient's body.

[0027] An important development provided by the present invention is a reliable approach to modeling of myocardial vascular interaction (MVI), which is part of the present three-component model. This interaction is necessary to improve the calculation of blood flow present in the coronary arteries. The present three-component model provides a reliable description of the hemodynamics in hyperemic conditions. It also covers the description of the hemodynamics in patients with complex pathologies. The present three-component model provides adaptability and good computational performance.

[0028] In another aspect, the invention relates to a computer-readable storage medium comprising instructions that, when executed by a computer, cause the computer to perform the steps of the method according to the embodiments disclosed herein.

[0029] In another aspect, the present invention relates to a system for assessing the hemodynamic response of a human patient to adenosine receptor agonist stimulation, the system comprising a measuring means for non-invasively measuring resting hemodynamic parameters of the human patient, the resting hemodynamic parameters including the systolic blood pressure of the patient, the diastolic blood pressure of the patient, and / or the heart rate of the patient, and a computer system adapted to perform the steps of the method defined in any embodiment of the present invention. In this aspect, simultaneous measurement and diagnosis are possible. This is beneficial in that the patient arrives at a diagnosis sooner. Alternatively, this allows for a separation of the diagnostic location where the patient is and the location where the calculations are performed. This is beneficial in many ways, for example, because some calculations may require more computer power that cannot be provided at the diagnostic location. Another advantage is that measurements can be made using devices with less computing power (e.g., handheld devices) compared to configurations where the same device is used for calculations and measurements / recording. In one embodiment, the system for assessing hemodynamic response to adenosine receptor agonist stimulation in a human patient as defined in the present disclosure includes a computer system as defined in the present disclosure, which further comprises a recording means for non-invasively and continuously recording the resting pressure waveform of the human patient. This aspect allows for a wider range of simultaneous measurements / recordings and diagnostics, which is further beneficial with respect to the above embodiment.

[0030] In one embodiment, a system for evaluating hemodynamic response to adenosine receptor agonist stimulation in a human patient as defined herein, the computer system as defined herein comprises at least one computer adapted to map resting hemodynamic parameters describing the human patient in a resting state to hyperemic hemodynamic parameters describing the human patient in a hyperemic state using a pharmacokinetic / pharmacodynamic model, and at least one computer adapted to perform parameter identification of a lumped parameter model using the hyperemic hemodynamic parameters describing the human patient in a hyperemic state to arrive at a lumped parameter model describing the human patient in a hyperemic state, wherein the at least one computer adapted to map resting hemodynamic parameters describing the human patient in a resting state to hyperemic hemodynamic parameters describing the human patient in a hyperemic state using a pharmacokinetic / pharmacodynamic model is configured to communicate directly or indirectly with the at least one computer adapted to map resting hemodynamic parameters describing the human patient in a resting state to hyperemic hemodynamic parameters describing the human patient in a hyperemic state using a pharmacokinetic / pharmacodynamic model. The hyperemic hemodynamic parameters describing the hyperemic human patient are used to perform parameter identification of the lumped parameter model to arrive at a lumped parameter model describing the hyperemic human patient. This system of the present invention allows for separation of the locations where the various calculations are performed, which allows for greater flexibility and adaptability to the conditions at the location of the patient being diagnosed. This may allow for faster diagnosis.

[0031] Further features and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments and the accompanying figures. It should be understood that the embodiments presented in the detailed description and claims can be combined in any order and number to produce new embodiments that form part of the disclosure, unless otherwise specified. The description includes numerous references to prior art documents, especially scientific journals. The full disclosure of these references is part of this disclosure. [Brief description of the drawings]

[0032] The present invention will now be described in more detail with reference to the accompanying drawings. [Figure 1] FIG. 1 illustrates an embodiment of the present invention that also implements optional steps. [Diagram 2] Figure 2 shows a block diagram of the generalized form of the three-element model (A) and the three basic building blocks of lumped parameter models that can be used to construct the blood circulation system (B), ventricular pressure volume (C), and coronary blood flow (D). [Diagram 3] FIG. 3 is a detailed block diagram of one embodiment of the present invention of the three-component model used to determine hemodynamics. [Figure 4] FIG. 4 shows the PK / PD fit results of a patient transitioning from a resting state to a hyperemic state showing typical responses obtained with an embodiment of the present invention. [Diagram 5] FIG. 5 shows the PK / PD approximation results of the transition from resting state to hyperemic state in patients exhibiting a hump reaction obtained with an embodiment of the present invention. [Figure 6] FIG. 6 shows the results of a regression model used for selected internal empirical parameters of a three-component model describing a patient in a hyperemic state in an embodiment of the present invention. [Figure 7] FIG. 7 shows a comparison of pressure values ​​measured in vivo and calculated in silico under hyperemic conditions using a three-component model according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0033] The present invention aims at the diagnosis of coronary insufficiency and provides developments in the methodology of identifying models and describing in silico techniques. As already mentioned, to achieve a hyperemic state, it is necessary to use specific pharmacological adenosine receptor stimulation (commonly the ADORA2 receptor is targeted) in clinical practice. Stimulation is performed using adenosine, regadenoson, binodenoson or apadenoson administered by intracoronary bolus (ic) or intravenous infusion (iv). The agonist binding effect is manifested in the form of an adenosine receptor signaling pathway cascade, which ultimately leads to vasodilation of the coronary arteries. As a result, coronary artery resistance (CVR) is minimized and coronary blood flow is increased to the maximum achievable. The present invention approaches this in two aspects: reaching the coronary blood flow itself and tracking the changes that occur during the hyperemic state.

[0034] In an embodiment of the invention, the hemodynamic model is defined by a three-component model (3CM) of coronary hemodynamics (see FIG. 2A). The model is a coupled model of the blood circulation system (BCS), the ventricular pressure volume (HPV), and the coronary blood flow (CBF). Each element is created using one or more function blocks, which must be specific for a given element (for BCS, HPV, and CBF, respectively, we use CRL, ERv, and RCpRp, which are depicted in FIG. 2B, FIG. 2C, and FIG. 2D, respectively). The ERv function block includes a valve, which is a heart valve modeling diode. This diode should be understood as a model of unidirectional flow, according to the terminology used in the field.

[0035] In the following description, detailed relationships in the form of equations are provided for each block. For example, the parameter of pressure (p) may or may not match the pressure (p) in one equation with that of another. Which pressure should be used in each equation can be derived, for example, from FIG. 3. Each block can be used multiple times with a particular element, so the form of the equation used is justified. Describing all possible configurations with their respective relationships would unnecessarily extend the description and may impair understanding of the invention due to its complexity.

[0036] For a CRL function block, the flow parameter relationships between the inputs and outputs can be expressed using the following basic equations:

number

[0037] where C, R, and L are compliance, resistance, and inertance, respectively.

[0038] In one embodiment, the systemic (left heart circuit) and pulmonary (right heart circuit) circulation models have the form of at least two CRL functional blocks connected in series for each circuit (i.e. n-compartment n-CRL, where n=2,3,4,...). This is beneficial in that it arrives at a model that provides a reliable description without requiring too high computational power to provide results. Furthermore, the heart model is formed by a two-chamber block describing the atria and ventricles, separately for the left and right blood circulation sides (Figure 2C). The ventricular hemodynamics can be described by the system of differential equations shown below:

[0039]

number

[0040] where the symbol <> denotes Macaulay brackets. To close this system of equations, it is preferable, and in some cases necessary, to know the changes in blood pressure ((V)p) in all four chambers with the periodic changes in systolic and diastolic volumes. Unlike in the vascular tree, where the capacitors are assumed to satisfy C(t)≠const, in the present invention the changes in compliance are significant and do not have a passive nature. In a preferred embodiment of the present invention, the vulnerability of the chambers is described by a time-varying elastance concept (E) as the inverse of the time-varying capacitor (see Suga H (1969) Time course of left ventricular pressure-volume relationship under variety enddiastolic volume, Jpn Heart J, 10(6), p. 509-15). This is of the form:

number

[0041] where V is the unloaded volume defined as the end-systolic chamber relationship p=f(V). To practically realize the relation (3), an approximation E(t) is required that is compatible with the use of cardiac physiology. In a preferred embodiment of the present invention, the approximation is performed using a periodic double Hill function (see Stergiopulos N et al. (1996) Determinants of step volume and systolic and diastolic aortic pressure, Am J Physiol, 270(6 Pt 2), p. H2050-9) according to the following formula:

[0042]

number

[0043] Where:

number

[0044] Here, in formula (3), the minimum value E min , maximum value E max , the dimensionless coefficient A, the sigmoid coefficient n, and the normalized value between t n =(t%T) / t max (modulo operator%, cardiac cycle T, time t at which the elastance function reaches its maximum value max ) is defined as

[0045] To be adequately described, the model needs to be complemented with a pV relationship describing the myocardium. In one embodiment, this relationship is represented using a time-varying elastance (E), and in another embodiment, by implementing the concept of myocardial fiber strain / stress (see Mirota K (2008) Constitutive Models of Vascular Tissue, Solid State Phenomena, 144, p. 100-105; Avazmohammadi R et al. (2019) A Contemporary Look at Biomechanical Models of Myocardium, Annual Review of Biomedical Engineering, 21, p. 417-442; Voigt JU, Cvijic M (2019) 2- and 3-Dimensional Myocardial Strain in Cardiac Health and Disease, JACC Cardiovasc Imaging, 12(9), p. 1849-1863). The fiber strain / stress concept defines the fiber stress (S f ) is assumed to be proportional to the bulk modulus, 1 / V dp / dV~S f , so after integration, the cavity pressure (p) and fiber stress (σ f ) is expressed as follows:

[0046]

number

[0047] The myocardial fiber strain / stress concept (MF) is much more sophisticated compared to the time-varying elastance concept (E), because MF covers the deformation and strain of myocardial fibers, whereas the time-varying elastance concept (E) has the advantage in terms of computational speed.

[0048] The final element of the three-element model (shown in Figure 2A) used to determine coronary hemodynamics is the coronary blood flow (CBF) element. This element is composed of the RCpRp functional block shown in Figure 2D. The relationship between its inputs and outputs can be expressed by the following general equation:

[0049]

number

[0050] Here, p zf is the pressure at zero flow in the coronary artery blood flow.

[0051] In a preferred embodiment of the present invention, the pressure p zfis assumed to be 14.8±7 mmHg for patients with stable angina, 22.5±9.1 mmHg for patients with non-Q wave myocardial infarction, and 37.1±1.9 mmHg for patients with Q wave myocardial infarction. (See Nanto S et al. (1996) Zero flow pressure in human coronary circulation, Angiology, 47(2), p. 115-22; Van Herck PL et al. (2007) Coronary microvascular dysfunction after myocardial infarction: elevated coronary zero flow pressure both in the infarcted and in the remote myocardium is mostly related to left ventricular filled pressure, Heart, 93(10), p. 1231-7.) The assumptions help speed up the calculations and shorten the time needed for diagnosis.

[0052] Myocardial Vascular Interaction (MVI) is used to describe the characteristics of the coronary circulation. Coronary blood flow is a function of the arterial information and the outlet conditions, which is a function of MVI. To obtain more reliable values ​​of coronary blood flow, MVI is necessary, but at the same time, no clear and reliable modeling strategy for MVI is known in the art. The development of a modeling strategy for MVI is a very important component of the system developed by the inventors of the present invention. Pressure p C , p RThis shows the effectiveness of MVI modeling and is fundamental for the MVI model (Boileau E et al. (2015) One-dimensional modelling of the coronary circulation. Application to noninvasive quantification of fractional flow reserve (FFR), Computational and Experimental Biomedical Sciences: Methods and Applications, 21, p. 137-155, Bruinsma T et al. (1988) Model of the coronary circulation based on pressure dependence of coronary resistance and compliance, Basic Res Cardiol, 83, p. 510-524, Epstein S et al. (2015) Reducing the number of parameters in 1D arterial blood flow modeling: less is more for-patient - specific simulations, American Journal of Physiology, Heart and Circulatory Physiology, 309(1), p. H222-H234, Mohrman D et al. (2013) Cardiovascular physiology, McGraw-Hill, Lange NY et al. (2013) Cardiovascular physiology, Elsevier, Zamir M (2005) The physics of coronary blood flow, Springer-Verlag).If we focus only on passive effects, which are inherent in the hyperemic state, the values ​​of pressure are controlled by two mechanisms: (i) interstitial, cavity-induced extracellular pressure C. EP =μ1p v (t)and (ii) shortening-induced intracellular pressure SIP=μ2E n (t)(as described in Algranati D et al. (2010) Mechanisms of myocardium-coronary vessel interaction, American Journal of Physiology. Heart and Circulatory Physiology, 298(3), p. H861-H873, Mynard JP et al. (2014) Scalability and in vivo validation of a multiscale numerical model of the left coronary circulation, American Journal of Physiology. Heart and Circulatory Physiology, 306(4), p. H517-H528, Westerhof N et al. (2006) Cross-talk between cardiac muscle and coronary vasculature, Physiological Reviews, 86(4), p. 1263-1308). Thus, focusing only on the passive effects inherent to the hyperemic state, pressure values ​​are controlled by two mechanisms: (i) interstitial, cavity-induced extracellular pressure C EP =μ1p v (t)(ii) Shortening-induced intracellular pressure SIP = μE n(t) (as shown in Algranati D et al. (2010) Mechanisms of myocardium-coronary vascular interaction, American Journal of Physiology. Heart and Circulatory Physiology, 298(3), p. H861-H873; Mynard JP et al. (2014) Scalability and in vivo validation of a multiscale numeric model of the left coronary circulation, American Journal of Physiology. Heart and Circulatory Physiology, 306(4), p. H517-H528; Westerhof N et al. (2006) Cross-talk between cardiac muscle and coronary vasculature, Physiological Reviews, 86(4) p.1263-1308). Thus, in a preferred embodiment of the present invention, the coronary flow limiting pressure is calculated as follows:

number

[0053] where p is the interstitial and intraluminal pressure, k R,C is the resistance and compliance coefficient of the coronary artery tree, CEP is the lumen-induced extracellular pressure, SIP is the shortening-induced intracellular pressure, μ is the lumen-induced extracellular pressure coefficient, μ is the intracellular pressure due to shortening, and E nshows the normal elasticity of Equation 4A. This embodiment provides a reliable description necessary for a reliable diagnosis, while at the same time having balanced computational requirements. Using the above functional blocks CRL, ERv, and RCpRp (shown in Figures 2B, 2C, and 2D, respectively), a complete hemodynamic model can be built. An overview of the complete hemodynamic model is shown in Figure 2A, and a more detailed structure of this model according to a preferred embodiment of the invention is shown in Figure 3. To make the model useful for a method that is a practical implementation, it is necessary to complete it with a set of empirical parameters. The most straightforward approach to achieving this goal is to have access to a large number of publications.Research into blood flow modelling over the past decades has resulted in thousands of papers published in scientific journals as well as many books on the subject (e.g. Rideout VC (1991) Mathematical and computer modelling of physiological systems, Prentice Hall; Spaan JAE (1991) Coronary blood flow mechanics, distribution, and control, Springer Science+Business Media Dordrecht; Hoppensteadt FC, Peskin Ch (1992) Modelling and Simulation in Medicine and the Life Sciences, Springer-Verlag; Ottesen JT, Olufsen MS, Larsen JK (2004) Applied Mathematical Models in Human Physiology, SIAM; Zamir M (2005) The Physics of Coronary Blood Flow, Springer Science+Business Media; van Meurs WL (2011) Modelling and Simulation in Biomedical Engineering: Applications in Cardiorespiratory Physiology, Willem van Meurs, McGraw Hill).

[0054] This approach may offer a small opportunity to account for patient-specific variability, as long as it successfully overcomes the problem of explicit method formulation.

[0055] In one embodiment of the present invention, selected (preferably easy to measure) anatomical or hemodynamic parameters are determined or measured individually for all analyzed cases, and then said anatomical or hemodynamic parameters can be used to calibrate the empirical parameters of the model (see Kirk JA et al. (2013) A priori identifiability analysis of cardiovascular models, Cardiovasc Eng Technol, 4(4), p. 500-512; Stergiopulos N et al. (1996) Determinants of strike volume and systolic and diastolic aortic pressure, Am J Physiol, 270(6 Pt 2): p. H2050-9). In a preferred embodiment of the present invention, it is a continuous non-invasive measurement using photoplethysmography.

[0056] The problem solved by the present invention is approached in a different way in the state of the art. Below we briefly introduce strategies that form the state of the art, which are not used in the present invention, and which are important for a proper understanding of the present invention.

[0057] The state of the art includes strategies based on statistical approaches. The simplest approach would be to estimate the variability of patient-specific empirical parameters using population-based statistical rules that are appropriate for demographic and general medical data (Burattini R, Di Salvia PO (2007) Development of systemic arterial mechanical properties from infancy to adulthood interpreted by four-element windkessel models, J Appl Physiol (1985), 103(1), p. 66-79, McVeigh GE et al. (1999) Age-related abnormalities in arterial compliance identified by pressure pulse contour analysis: aging and arterial compliance, Hypertension, 33(6), p. 1392-8, Kassab GS (2019) Coronary Circulation, Anatomy, Mechanical Properties, and Biomechanics, Springer Science+Business Media, Westerhof BE et al. (2020) Pressure and Flow Relations in the Systemic Arterial Tree Throughout Development From Newborn to Adult, Front Pediatr, 8, p. 251).

[0058] A very easy hemodynamic parameter to measure is the pressure at the level of the main arteries. At the same time, pressure is a direct reference data for the model prediction results. In this respect, auscultatory methods using a stethoscope and a sphygmomanometer or oscillometric methods can be used.

[0059] Estimation of cardiac output is then performed based on echocardiographic assessment (attempts in this regard are described in Lang RM et al. (2015) Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging, J Am Soc Echocardiogr , 28(1), p. 1-39 and Dumesnil JG et al. (1995) A new, simple and precise method for determining ejection fraction by Doppler echocardiography, Can J Cardiol , 11(11), p. 1007-14) or on computed tomographic assessment of the cardiac anatomy.

[0060] Based on knowledge of the mass or volume of the cardiac chambers, it is estimated according to the myocardial blood flow (MBF) [ml / min 100g] (see Anderson HV et al. (2000) Coronary artery flow velocity is related to lumen area and regional left ventricular mass, Circulation, 102(1), p. 48-54) and expressed as:

number

[0061] Alternatively, it is possible by direct estimation of myocardial oxygen consumption MVO2 [ml O2 / min 100g] (a very extensive review of potentially available estimates is provided in Baller D et al. (1979) Validity of myocardial oxygen consumer parameters, Clin Cardiol, 2(5), p. 317-27). It has been confirmed that a strategy of selecting empirical parameters of the circulatory model has been used, as in the preferred embodiment of Fig. 3. This strategy involves selecting (calibrating) the empirical parameters of the model according to available partial characteristics of the hemodynamics (pressure, myocardial blood flow, etc.). This methodology allows a direct association with the real hemodynamic state, but has major limitations.

[0062] Unfortunately, the above strategy is insufficient for its use in the diagnostic functional evaluation of coronary artery disease. Its drawback is the lack of access to some parameters of the target hemodynamic state. If the model predictions relate to a hyperemic state, its parameters should be set during hyperemia and not at rest. In fact, it is possible to measure resting pressures (SYS, DIA) and heart rate (HR), and there are no difficulties or complications with the continuous use of pressure waveforms. However, it is unreasonable to administer adenosine or other vasodilators used in pharmacological stress tests during the measurement of SYS, DIA, and HR.

[0063] The prior art offers various other strategies to address the problem of the present invention. In the simplest approach, one can try to reproduce the hyperemic effect by assuming arbitrary changes in the main hemodynamic parameters. For example, one can assume that the coronary resistance is reduced by 4 times the constant, the aortic blood pressure is reduced by about 10%, and the heart rate is increased by about 10% (disclosed in Taylor CA (2012) Method and system for patient-specific modeling of blood flow. US Patent No. 8157742B2; Taylor CA (2021) Method and system for image processing to determine blood flow. US Patent No. 20210282860A1). As shown by an extensive review of clinical trial results, there is a very high risk that the results obtained with these prior art approaches are far from reality. The main reason is the large patient-specific variability here and the lack of direct correlation with demographic or general medical data.

[0064] The large patient-specific variability in response to adenosine receptor stimulation was observed long before models were developed to quantify that variability. It is therefore worthwhile to refer to earlier prior art approaches in searching for proposals to explain the mechanism of this phenomenon, which will allow a better understanding of the details of the present invention.

[0065] To test the dose-response kinetics of intravenously and intracoronarily infused adenosine, Wilson et al. conducted a study in 39 patients. First, coronary angiography was performed to diagnose the patients' chest pain syndrome. The 39 patients were divided into two groups, 8 of whom had microcirculatory abnormalities associated with ventricular hypertrophy of unknown etiology, and 31 had atypical chest pain due to angina, mild or normal epicardial coronary stenosis, normal coronary flow reserve (CFR), and normal left ventricular function. The researchers tested the kinetic response to doses of adenosine administered by intravenous infusion, intracoronary infusion, and intracoronary bolus administration. Single doses of intracoronary bolus ranged from 2 to 16 μg, intracoronary infusions ranged from 10 to 240 μg / min, and intravenous infusions ranged from 35 to 140 μg / kg / min. Coronary blood flow velocity (CBFV) was measured using a 3F Doppler coronary catheter. The maximum change in CBFV was calculated as the ratio of peak CBFV after an intracoronary bolus of adenosine or papaverine to resting CBFV (?CBFV). The index of change in total coronary resistance was calculated as follows:

number

[0066]

number

[0067] Intracoronary adenosine safely induced maximal coronary vasodilation, and intravenous adenosine infusion at 140 μg / kg / min produced near-maximal hyperemia (see Wilson RF et al. (1990) Effects of adenosine on human coronary arterial circulation, Circulation, 82 (5), p. 1595-1606). Another important observation was that different doses of adenosine had different effects on the results of the studies.

[0068] The researchers, led by Sparv, were investigating the effect of increasing adenosine doses on fractional cardiac reserve (FFR), mean arterial pressure (MAP), heart rate (HR), and discomfort (measured by visual analog scale (VAS)). 87 patients participated in the study (10 patients were excluded due to atrioventricular block after adenosine injection, and 2 withdrew due to significant discomfort). During the FFR evaluation, patients received a standard dose of 140 μg / kg / min followed by a high dose of adenosine intravenously at 220 μg / kg / min. A separate patient group was also analyzed, which included 43 patients who consumed caffeine within 6 hours of FFR. The results showed no significant changes in FFR results, MAP, or HR between standard and high adenosine doses. The only difference between standard and high adenosine doses was that patients experienced increased discomfort after the high dose adenosine administration. During the study, MAP and HR were monitored and recorded as baseline at 30-second intervals. Statistical analysis was performed using the Wilcoxon matched-pairs signed-rank test and linear regression model correlation of adenosine dose. Bland-Altman plots were created to show agreement, and for the continuous hemodynamic variables of MAP and HR, area under the curve (AUC) was analyzed (see Sparv D et al. (2017) Assessment of increasing intravenous adenosine dosage in fractional flow reserve, BMC Cardiovascular Disorders, 17, p. 1-9).

[0069] Another approach was developed by Robert Wilson and his group. Robert Wilson and his colleagues at the University of Minnesota Medical School provided not only a multifaceted adenosine kinetic assessment but also a total coronary resistance index, TCRI (see Equation 9a). This is certainly similar to the linear resistance laws that appear in many areas of physics (e.g. Ohm's law and Hagen-Poiseuille's law). His work was an attempt to understand this problem more deeply, but did not lead to any breakthroughs. Nevertheless, it is exciting that this work presented a slightly different strategy for empirical parameter selection of modern hemodynamic circulation models for hyperemic conditions. That is, instead of questionable ad-hoc assumptions, one can try to build hypothetical structural relationships between blood flow and heart rate as they may change in specific cases (see Sharma P et al. (2012) A framework for personalization of coronary flow computes during rest and hyperemia, Annu Int Conf IEEE Eng Med Biol Soc, p. 6665-8; Sharma et al. (2019) Non-invasive functional assessment of coronary artery stenosis including simulation of hyperemia by changing resting microvascular resistance. US Patent No.10373700B2; Itu LM et al. (2020) Framework for personalization of coronary flow computes during rest and hyperemia. US Patent No.10622110B2).

[0070] The coronary circulatory resistance can be defined in analogy with the systemic circulatory resistance (usually magnitude 9...20 mmHg min / l) as follows:

number

[0071] The TRCI index is expressed here as a quotient of constant resistances for the resting and hyperemic states as follows:

number

[0072] This allows the calculation of the hyperemic resistance parameter R based on the resting resistance parameter. hiper =TRCI·R rest It is clear that the effectiveness of this strategy depends on the determination of the TRCI index value.

[0073] TRCI can also be determined to include a correction for HR, see Equation 11, which has the following form (as presented by Sharma P et al., 2012, which is based on the clinical trial results of McGinn AL. et al. (1990) Interstudy variability of coronary flow reserve. Influence of heart rate, arterial pressure, and ventricular preload, Circulation, 81(4), p. 1319-30):

number

[0074] Strategies based on the TRCI index concept can be effective in some cases in terms of providing flexibility or reproducing the effects of patient-specific variability. However, this strategy can only be effectively applied if the set of empirical parameters in the model is small. In the example above (specifically, Sharma P et al. (2012)), the recalculation related to the transition from resting state to hyperemia is related only to resistance. In other words, this strategy only works for selected models and not for models designed to describe more complex medical cases. The individual medical parameters of the patient cannot be omitted.

[0075] Various embodiments of the present invention utilize a different strategy compared to the ones mentioned above, in particular a different one from the use of population-based statistical data or TRCI. The strategy of the present invention is based on a mapping between resting and hyperemic states performed by selected hemodynamic parameters (SYS, DIA, etc.) that are fitted to the results provided by a pharmacokinetic / pharmacodynamic (PK / PD) model of adenosine metabolism.

[0076] The key approaches to PK / PD modeling of adenosine metabolism are described in the following paragraphs. In the field of PK / PD modeling of adenosine metabolism, a classic study by a team from the Institute of Clinical Medicine, University of Siena, represents a key point. Blardi et al. performed tests of adenosine kinetic analysis 1 min after adenosine infusion of 2.5, 5 and 10 mg (38, 79 and 48 μg / kg, respectively) and after a 200 μg / kg infusion over 10 min, followed by a 400 μg / kg infusion over 10 min. 22 healthy volunteers were divided into 4 groups: Group A (6 volunteers, receiving 2.5 mg intravenous adenosine over 1 min infusion), Group B (6 volunteers, receiving 5 mg intravenous adenosine over 1 min infusion), Group C (5 volunteers, receiving 10 mg intravenous adenosine over 1 min infusion), and Group D (5 volunteers, receiving 200 μg / kg intravenous adenosine over 10 min infusion followed by 400 μg / kg intravenous adenosine over 10 min infusion). The researchers took blood samples at the time of adenosine infusion (time 0) and at selected time points (times 1, 3, 5, 7, 10, and 15 min for groups A, B, and C; and time 0, 5, 10, 15, 20, 21, 25, and 30 min for group D). One minute after adenosine infusion, adenosine concentrations increased, mean clearance decreased, and half-life increased. After 20 min, adenosine plasma concentrations peaked compared to concentrations reached after an infusion of 5 mg adenosine administered over 1 min, but the mean clearance and half-life were different. Adenosine concentrations returned to baseline 5–15 min after infusion was stopped. Adenosine plasma concentrations were examined by HPLC. Pharmacokinetic analysis between adenosine concentrations after a given period and baseline concentrations was performed. Adenosine concentrations C(t) in groups A, B, and C were calculated as follows:

number

[0077] where t is time, A and β are parameters estimated by nonlinear regression (see Blardi P et al. (1993) Pharmacokinetics of exogenous adenosine in man after infusion, European journal of clinical pharmacology, 44 (5), p. 505-507).

[0078] Although Blardi's work differs from modern PK / PD models used in the field, it dates from the early 1990s and serves as a very important reference point.

[0079] Another very interesting prior art regarding the quantification kinetics of adenosine is the study by Davis et al. ATP channel) and N ω We sought to define the effect of -Nitro-L-arginine methyl ester (L-NAME), which inhibits nitric oxide synthesis. Adenosine was administered by intracoronary infusion (using doses of 10–1000 μg / min). Dogs were first properly prepared and then anesthetized using a mechanical ventilator. A left tracheotomy was performed and catheters were placed in the descending aorta and left atrium. A Konigsberg micromanometer was placed in the left ventricle through the left ventricular apex. A Transonic Systems transit-time ultrasound flow probe and hydraulic occluder were placed in the proximal left circumflex coronary artery or the left anterior descending coronary artery. A 22-gauge angiocatheter connected to small-bore tubing was placed in the artery distal to the flow probe. A bipolar pacing lead was sewn into the left atrial appendage. Under basal conditions, the plasma adenosine concentration required to achieve 50% of the maximum increase in conductance (ED50) increased 3-fold after administration of 1 mg / kg or 10 mg / kg L-NAME. The ED50 increased dose-dependently after administration of glibenclamide. ATPChannel and NO synthesis inhibition were additive with a 15-fold increase in ED50. Glibenclamide and L-NAME caused an increase in systemic pressure and a decrease in coronary conductance. During the study, hemodynamics were monitored in analog form with a sampling rate of 30 seconds. For each, the values ​​of the variables were measured every 30 seconds. Adenosine plasma concentrations were calculated using the Olson method (see Equation 14). The dose-response adenosine data were fitted to a sigmoidal model, expressed as follows:

number

[0080] where x is the logarithm of the adenosine plasma concentration, f(x) is the conductance, and A, B, C, and D are constants (see Davis CA et al. (1998) Coronary vascular responsiveness to adenosine is impaired additively by blockade of nitric oxygen synthesis and a sulfuronylurea, Journal of the American College of Cardiology, 31 (4), p.816-822). In this review, the typical sigmoid adenosine receptor response model is directly referenced.

[0081] A further prior study was performed by Tune et al.. They analyzed the role of adenosine in coronary exercise hyperemia in 10 dogs. Catheters were placed in the aorta and coronary sinus. A flow probe was placed in the circumflex coronary artery. Cardiac interstitial adenosine concentrations were estimated from coronary venous plasma concentrations and arterial venous plasma concentrations using a mathematical model. Coronary blood flow (CBF), myocardial oxygen consumption, heart rate (HR), and aortic pressure were measured at rest and during exercise on a treadmill with and without adenosine receptor blockade. During exercise, the dogs showed a 4.2-fold increase in oxygen consumption, a 3.8-fold increase in CBF, and a 2.5-fold increase in HR, with no change in mean blood pressure (MBP). Coronary venous adenosine plasma concentrations changed little during exercise. Estimated interstitial adenosine concentrations were well below the threshold for coronary dilation. Oxygen consumption or CBF at rest and during exercise remained largely stable after adenosine receptor blockade. Coronary venous adenosine concentrations and estimated interstitial adenosine concentrations did not change to compensate for receptor blockade. Adenosine arterial plasma concentration [ADO] was calculated using the following formula:

number

[0082] where [ADO] is the adenosine concentration and Hct is the hematocrit. The Hill equation dose-response adenosine data was then fitted using the following equation:

number

[0083] where q is the coronary blood flow, q min is the measured coronary blood flow at rest, q maxis the measured maximum coronary blood flow, [A] is the calculated intraarterial adenosine concentration, ED50 is the 50% effective dose, and H is the Hill ratio (see Tune JD et al. (2000) Adenosine is not responsible for local metabolic control of coronary blood flow in dogs during exercise, American Journal of Physiology-Heart and Circulatory Physiology, 278 (1), p. H74-H84).

[0084] The present invention takes a different approach to PK / PD modeling, which is explained below. To achieve biochemical control, nature has developed a very precise regulatory mechanism in the form of enzyme-catalyzed reactions. Apart from the substrate S and product P, a biocatalyst E appears in the setup, whose activity is delicately adjusted according to the needs and circumstances. This relationship can be summarized in its simplest form as a chemical equation:

number

[0085] It is reasonable to assume that the concentration of the enzyme-substrate complex (ES) during the reaction does not change significantly. Similarly, one can assume that an equilibrium occurs between the concentrations of the complex and the substrate. k1[E][S]-k -1 [ES]=k2[ES] Combining all this gives a simple relationship that describes the dependence of the rate of product formation on substrate accessibility.

number

[0086] The above equation was discovered by Leonor Michaelis and Maud Menten while studying fermentation kinetics. Over time, this equation has become very useful in explaining widely understood enzyme-catalyzed processes (see Rodwell VW et al. (2018) Harpers Illustrated Biochemistry, McGraw-Hill Medical). For complex biochemical systems, the number of reactants is large and reactions often form cascade systems. A good example is the purinergic receptor (especially the adenosine receptor). The reactions of these receptors are usually carried out in parallel with the use of n catalytic centers, so we get the following equation (following Mirota K (2013) Hemodynamic aspects of atherogenesis):

number

[0087] Here, K n =K1 K2 ... K n When assessing the effect of an agonist A binding to one of the n catalytic centers of a receptor R, we need to arrive at [S] = [AR]. Therefore, according to the Michaelis-Menten model, the following equation is obtained:

number

[0088] Combining both equations, we define the transducer ratio (τ = [R0] / K A ) gives a new equation describing the contract catalytic effect (following the original work by Black JW, Leff P (1983) Operational models of pharmacological agonism. Proc R Soc Lond B Biol Sci. 220(1219):141-62).

number

[0089] This is a generalization of Equations 16a and 16b.

[0090] The transition effect between resting and hyperemic is not limited to the regulation of vasomotor tone of active coronary arteries. In fact, the observed and investigated changes in coronary circulatory resistance are only a specific type of manifestation of the effects caused by different cardiotropic forms (chronotropic, inotropic, relaxant, etc.). In a preferred embodiment, the transition from baseline to hyperemic state is considered as a pure cardiotropic effect due to the binding of purinergic receptor agonists and is modeled by a cooperative binding relationship of the following form:

number

[0091] where X is an arbitrary measure reflecting the change in ligand concentration and K denotes the apparent dissociation constant. This equation represents one approach to a PK / PD model. In another embodiment, the net cardiotrophic effect is modeled by the Black-Leff operational model of pharmacological agonism, which has the following form: (Adapted from Black JW, Leff P (1983) Operational models of pharmacological agonism, Proc R Soc Lond B Biol Sci, 220(1219), p. 141-62, Leff P, Martin GR, Morse JM (1985) Application of the operational model of agonism to establish conditions when functional antagonism may be used to estimate agonist dissociation constants, Br J Pharmacol, 85(3), p. 655-63)

number

[0092] where X and K are again the assumed values ​​of the concentration change and dissociation constant, and τ is the ratio that defines the efficacy of the agonist (see Jambhekar SS, Breen PJ (2012) Basic Pharmacokinetics, Pharmaceutical Press; Frigyesi A, Hossjer O (2006) Estimating the parameters of the operational model of pharmacological agonism, Stat Med, 25(17), p. 2932-45).

[0093] All embodiments of the invention have been tested. The method according to the invention has been verified based on the medical experimental results of a multicenter non-randomized clinical trial. The study was planned to be carried out on 60 participants. Finally, 67 patients were screened and 48 patients were enrolled. The collected patient demographic and health data that affect the circulatory hemodynamics in the body are shown in Table 1 (parameters are expressed as mean ± standard deviation). For this purpose, in particular data that affect the pressure pulse wave propagation in the patient's body. [Table 1]

[0094] For each enrolled patient, continuous noninvasive blood pressure measurements (CNBP) were performed at the radial artery with a sampling rate of 100 Hz. In addition, an independent noninvasive measurement was performed at the brachial artery using an oscillometric method. Because patients underwent invasive diagnostics, the results of intra-aortic catheter pressure measurements were used as reference data. Pressure signals from intra-aortic catheter pressure measurements were sampled at a rate of 200 Hz.

[0095] Each patient enrolled in the clinical trial was diagnosed with coronary heart disease, but the clinical trial itself required CHD to be confirmed by functional assessment, which is shown in Table 2. [Table 2]

[0096] This specific selection of the validation group (only those diagnosed with coronary heart disease) reflects the focus of the subject application of the present invention.

[0097] Figure 4 shows the results of one embodiment of the method according to the invention, incorporating a model based on the Hill-Langmuir or Black-Leff equations. Both types of equations produce similar results and can be used interchangeably. In the background is a recording of a 3-minute left coronary artery catheterization and the predicted systolic (downward triangles) and diastolic (upward triangles) pressure changes plotted with a solid line. In the resting state, the mean systolic pressure was SYS = 122.7 ± 3.0 mmHg (coefficient of variation cv = 2.4%) and the diastolic pressure was DIA = 57.9 ± 1.51 mmHg (cv = 2.6%). Adenosine was administered exactly 43.7 seconds and stable hyperemia was reached at 108.44 seconds. The pressure then decreased to SYS = 92.5 ± 1.5 mmHg (cv = 1.6%) and DIA = 32.1 ± 0.6 mmHg (cv = 1.8%). Essentially, regardless of the model (8) or (9) used, the prediction results followed the same pattern as shown by the SYS and DIA pressure lines (matching). The mean residual in the systolic value area for model (8) was only 1.6 mmHg, not exceeding the range of -5.2 to 12.2 mmHg, and for diastolic pressure it was 0.4 in the range of -3.7 to 3.3 mmHg. Similarly, for model (9), the mean residual for systolic pressure was 1.3 mmHg in the range -6.0 to 13.1 mmHg, and for diastolic pressure it was 0.3 in the range -3.6 to 3.2 mmHg. It is also worth noting that case A shown in Figure 4 represents the response to adenosine receptor stimulation with a typical S-shaped form. Statistically, this shape is observed in about 57% of cases in clinical conditions, so a little more than one in two patients. In about 39% of patients (i.e., two out of five patients who became hyperemic), a completely different response pattern is observed. In these patients, the response pattern is hump-shaped and sigmoidal with overlapping ridges of fluctuation (see Johnson NP (2015) Repeatability of Fractional Flow Reserve Despite Variations in Systemic and Coronary Hemodynamics, JACC Cardiovasc Interv, 8(8), p. 1018-1027).

[0098] Figure 5 shows the second response pattern to adenosine. At rest, pressures SYS = 125.5 ± 2.1 mmHg (cv = 1.7%), DIA = 71.9 ± 1.2 mmHg (cv = 1.7%). Indeed, much higher residual values ​​are clearly seen in the transition phase, and the model used does not significantly affect them. In the first model (equation 18), the systolic pressure is -3.7 (-13.9 ... 9.5) mmHg, and the diastolic pressure is -1.6 (-7.4 ... 5.8) mmHg. In the second model (equation 19), the values ​​are -3.6 (-13.9 ... 10.0) mmHg and -1.6 (-7.5 ... 6.7) mmHg, respectively. The residual values ​​of the hump-shaped residuals are higher in the transition phase, but do not negatively affect the predictive quality of both models for the state of maximum stable hyperemia.

[0099] Methods involving models formulated using equations (18) and (19) above provide the potential for effective mapping from the resting state to expected states under hyperemic conditions using the results of in vivo measurements made while the patient is in a resting state (e.g., central arterial pressure measured non-invasively using photoplethysmography).

[0100] In a preferred embodiment of the present invention, a regression model can be used that provides initial values ​​for the internal empirical parameters of the three-component model (3CM) (Heijmans RDH, Magnus JR (1986) Consistent maximum-likelihood estimation with dependent observations. The general (non-normal) case and the normal case, J Econ, 32, p. 253-285, Escobar J (2012) Time-varying parameter estimation under stochastic perturbations using LSM, J Math Control Inform, 29(2), p. 235-258, Banks HT, Hu S, Thompson WC (2014) Modeling and inverse problems in the presence of uncertainty, CPC Press, Boca Raton, Mao Z (2020) Model validation and uncertainty quantification, Springer International Publishing). This model was used to process the data of the study.

[0101] As explained above, clinical experimental population data were used for parameter identification. These were used to estimate values ​​for the experimental parameters of the 3CM. A single parameter identification run results in a model that can be used for patient-specific calculations. This is particularly suitable for routine use, where a fast and accurate starting value needs to be given for calculations for a particular case.

[0102] Figure 6 shows the predicted results of selected empirical parameters of the 3CM when the 3CM has a structure as shown in Figure 3. The horizontal axis of each graph represents the values ​​of the model parameters obtained as a result of the calibration (parameter identification) compared to the values ​​obtained using invasive measurements (performed in vivo for hyperemic states). The round symbols indicate the exact location of the values ​​obtained using in vivo measurements. The calibration was performed individually for each patient participating in the clinical trial within an established range of 5 cardiac cycles. When assessing the reproducibility quality of the results of the invasive measurements, the L2 norm of the pressure difference was calculated (in vivo measurements and simulations). The value of the L2 norm was 2.3 mmHg for all cases and cycles (interval at 95% confidence level is 2.1...2.5). The calibration resulted in values ​​of the empirical parameters that represent with a statistically high probability the hyperemic state of a particular patient.

[0103] Then, for each parameter, a regression model was built using demographic and general medical data. The results of this model are plotted on the vertical axis of FIG. 6. Stars indicate the value of a particular empirical parameter obtained using the respective demographic and general medical data. Furthermore, in each graph, the simple regression is shown in solid line and the range with a 95% confidence level is shown in dashed line. FIG. 6 shows only 10 selected empirical parameters out of the dozens of parameters included in 3CM. The same type of marks are used for the same data in FIG. 3 and FIG. 6. It can be seen that the results of the parameter identification of 3CM using demographic and medical data and the results of continuous non-invasive measurements are highly correlated with the results obtained from the calibration performed using invasive data (the value of correlation is shown as a percentage in the title of each plot). The adaptation of the method according to the invention to a particular patient and accounting for the variability of the patient using the patient's demographic and medical data, etc. is also clearly shown in FIG. 6. The use of regression models is entirely optional in the present invention and does not allow to obtain more accurate results. Regression models running on statistical invasive data can provide a very good starting point for the method of the invention, which can speed up the conclusion of the model's parameter identification by several times, and therefore the patient-specific parameters calculated by the model can be reached very quickly.

[0104] FIG. 7 shows selected results obtained with an embodiment of the method of the invention. The results are shown using two extreme boundary cases. This selection allows a comparison. Both selected cases concern hyperemic conditions, case A on the left and case B on the right. The first case showed the most typical and most common reaction of the systemic circulation to the administration of adenosine, known from textbooks and elsewhere. The second case (case B) is very unusual and peculiar. Case B is the only case of this kind in the pool selected for clinical trials. This case was deliberately selected for comparison as a benchmark of the invention in the most extreme environment. Nevertheless, in both cases, a very good agreement was obtained between the results obtained using the method of the invention and those obtained from invasive measurements. The agreement was evaluated by calculating the Perason's correlation coefficient, whose value reached 99% (case A) and 98% (case B). The plots of the variations expressed as a function of time of the calculated pressure (solid line in the graph) and the invasively measured pressure (circular points) also showed an excellent agreement. A larger deviation in the maximum range (approximately 10 mmHg) was observed in the fourth cycle of case B. This deviation was expected due to the extreme nature of case B.

[0105] The method of the present invention, in one embodiment, comprises five steps, as shown in Figure 1. In the first step, demographic and general medical data of the patient are collected. As mentioned before, this step is not essential to the method, but may help to speed up the calculations. These data are only used in step 4 of the method, which relates to parameter identification of the model, but their importance relates to the efficiency of the method. First, data on sex, age, weight, and / or height are collected (in one embodiment, the influence is estimated from: 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, particularly those including beta-adrenergic blockers (BBLOCKs), angiotensin-converting enzyme inhibitors (ACEs) and / or antiarrhythmic drugs (AARRs). (In one embodiment, the effect is estimated from: 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 Aet 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)). .

[0106] In some embodiments, a drug therapy is included in the method, and in other embodiments, the drug therapy includes the drugs listed above. Other drugs may also affect the pulse wave propagation in the patient's body, and as a result, other embodiments include other drug therapies. Drug therapy includes different dosing regimes as well as additional medical effects of drug therapy. Each of the above elements can be used individually or in any combination to obtain a better starting value (starting value) of the model identification process for a particular patient case. A better starting value directly affects the efficiency of the method according to the invention. In certain embodiments, it is conceivable that only selected data is collected, such as, for example, gender, age, or selected drug. The more data used, the better the starting value obtained.

[0107] As shown in FIG. 1, in the second step of the method of the invention, a pressure waveform of the patient at rest is obtained. This step is also not essential to the invention, but may be useful in the parameter identification step. Alternatively or additionally, this step may be used to estimate the main hemodynamic parameters for the purpose of the third step. Taking into account the limitations and preferences of clinical practice in measuring continuous non-invasive blood pressure (CNBP), in a preferred embodiment of the invention, the pressure measurement is performed in the radial artery. In one embodiment, the pressure measurement is performed using finger cuff photoplethysmography and / or applanation tonometry. The advantage of using said measurement techniques is that they are simple to use and the invention works with these techniques as well. The recording and further analysis of the continuous pressure waveform is performed within a window that includes a signal block covering one or more complete cycles of the patient's heart. Each cycle corresponds to the systolic-diastolic movement of the heart. In a preferred embodiment of the invention, a window of width corresponding to the length of the respiratory cycle can be analyzed (following Rodriguez-Molinero A (2013) Normal ventilation rate and peripheral blood oxygen saturation in the older population. Journal of the American Geriatrics Society. 61(12):2238-2240; Park C, Lee B (2014) Real-time Estimation of Breather 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).This embodiment is beneficial as it allows for a more accurate estimation of the parameters.

[0108] As shown in FIG. 1, in the third step of the method of the invention, the main hemodynamic parameters of the patient in the hyperemic state are determined. This is done using a mapping process that transposes the parameters representative of the resting state to the parameters representative of the hyperemic state. The mapping of the parameters obtained using the results of the non-invasive measurements made in the resting state is done using the PK / PD model using Equations 18 and 19. In one embodiment, the mapping can be done only for the main hemodynamic parameters including systolic (SYS) and diastolic (DIA) pressures and heart rate (HR). In another embodiment, further hemodynamic parameters can be used. In one embodiment, the PK / PD model is not used directly for the experimental parameters of the 3CM (having the structure shown in FIG. 3). In another embodiment, no waveforms are used in the PK / PD model, and the waveforms in the hyperemic state are provided by the 3CM. This embodiment is much more time-efficient compared to the embodiment in which the waveforms are calculated using the PK / PD model.

[0109] In the fourth step, parameter identification of the 3CM of lumped parameters of hemodynamics is performed (Figure 3). For this purpose, regression models can be used. Regression models provide a probabilistic quantitative description. Regression models can be used to provide starting values ​​of the 3CM. In one embodiment, regression models are not used in methods that require additional time to complete the calculations. This additional time can be in the region of tens of hours. In an embodiment using regression models, the result is reached faster and in that embodiment parameter identification is a more numerically stable process. In another embodiment, parameter identification is performed independently and in parallel with adenosine receptor stimulation. Adjustment of the values ​​of the empirical parameters using the hyperemic hemodynamic parameters can be performed using any suitable mathematical method.

[0110] In a fifth step, one or more flow parameters in the systemic, pulmonary and / or coronary circulation are calculated by a three-component model (3CM). Said flow parameters include flow rates and pressures at locations corresponding to the used structure of the 3CM. The specific structure of the 3CM is not very important for ventricular flow parameters (each with two atria and two ventricles), but for the blood circulation system (BCS), the differences can be important. The embodiment shown in FIG. 3 has two compartments for the systemic circulation and two compartments for the pulmonary circulation of the BCS model, providing flow rates and pressures only for these compartments (precisely the arterial and venous reservoirs). In different embodiments, there are n-compartment models where n is used greater than 2. These embodiments provide a more detailed description at the various levels of the circulatory system. In a preferred embodiment of the invention, the coronary blood flow parameters are calculated for the left (LCA) and right (RCA) arterial inlets.

[0111] The calculation steps or sub-steps can be implemented using a computer or a computer program. In certain embodiments, some or all of the calculations are performed using a computer program stored on a computer or any type of memory device, or both. In another embodiment, some or all of the calculations for the purposes of the method can be performed remotely, such as using a cloud-based infrastructure, including the use of the Internet or a local network, at a location separate from where the patient is located.

[0112] All the measurement methods and algorithms described herein are intended to define the parameters of the invention and provide specific results that can be compared to clinical trials, but are by no means limiting and are exemplary. Words such as "comprise" or "comprise" are not limiting, for example, if an element A includes another element B, element A can include other elements or elements in addition to element B. The use of the singular or plural forms does not limit the scope of the disclosure, for example, a portion of the description indicating that element A includes element B also discloses an embodiment in which multiple elements B are included in one element A, multiple elements A are included in one element B, and multiple elements A include multiple elements B. Many other embodiments will be apparent and clear to those skilled in the art upon review of the contents of this disclosure. Thus, the scope of the invention should be determined with reference to the appended claims and the full scope of equivalents to which such claims are entitled.

Claims

1. A method for assessing the hemodynamic response of a human patient to adenosine receptor agonist stimulation, comprising: - mapping hemodynamic parameters of said human patient at rest to hemodynamic parameters of said human patient during hyperemia using a pharmacokinetic or pharmacodynamic model, wherein the pharmacokinetic / pharmacodynamic model is a nonlinear model including a Michaelis-Menten model, a Hill-Langmuir model, a Black-Leff model, a multi-pathway model, and / or a variety of blood-tissue exchange models; the resting hemodynamic parameters include the patient's systolic blood pressure, the patient's diastolic blood pressure, and / or the patient's heart rate, and the resting hemodynamic parameters are non-invasively measured or estimated using non-invasively recorded continuous patient pressure waveforms during the resting state; the non-invasively recorded continuous patient pressure waveform includes a time window including at least one full cycle of the human patient in the resting state; - performing parameter identification of a lumped parameter model of said human patient using the hyperemic hemodynamic parameters of said human patient to arrive at a lumped parameter model of said human patient in a hyperemic state, wherein the lumped parameter model is of Windkessel type that at least partially describes the hemodynamics of the human patient.

2. 10. The method of claim 1, further comprising calculating flow parameters using the lumped parameter model of the human patient in the hyperemic state; The method wherein the flow parameters include pressure, flow, waveform, and / or cardiac intervals, including but not limited to cardiac cycle duration and ejection time.

3. 10. The method of claim 1, wherein the resting hemodynamic parameters of the human patient further comprise a non-invasively recorded continuous resting patient pressure waveform.

4. 2. The method of claim 1, wherein the parameter identification comprises: - estimating values ​​of empirical parameters of a lumped parameter model of said human patient using demographic and health data that influence pressure pulse wave propagation within said human patient; - using the hyperemic hemodynamic parameters of said human patient to improve the values ​​of the empirical parameters of a lumped parameter model of said human patient.

5. 5. The method of claim 4, wherein the demographic data and the health data are collected for patients diagnosed with coronary heart disease.

6. 5. The method of claim 4, wherein the demographic data and the health data include gender, age, height, general fitness rating and / or current medications, including but not limited to beta-blockers, angiotensin-converting enzyme inhibitors and / or antiarrhythmic medications.

7. 2. The method of claim 1, wherein the Black-Leff model is represented by the following formula: [Equation 20] Including, where Z is the pharmacological effect, and Z ｍ is the maximal response, τ is the ratio that defines the efficacy of the agonist, [A] is the concentration of the agonist, and K Ａ is the agonist-receptor dissociation constant, and n is a factor determining the steepness of the curve.

8. 10. The method of claim 1, wherein the Hill-Langmuir model is represented by the following formula: [Equation 21] Including, where Z is the pharmacological effect, and Z ｍａｘ is the maximum response, and Z ｍｉｎ is the minimum response, X is an arbitrary measure reflecting the change in agonist concentration, K is the apparent dissociation constant, and n is a factor determining the steepness of the curve.

9. 10. The method of claim 1, wherein the Black-Leff model is represented by the following formula: [Equation 22] Including, where Z is the pharmacological effect, and Z ｍａｘ is the maximum response, and Z ｍｉｎ is the minimum response, X is an arbitrary measure reflecting the change in agonist concentration, K is the apparent dissociation constant, n is a factor determining the steepness of the curve, and τ is a ratio defining the effectiveness of the agonist.

10. 10. The method of claim 1, wherein the non-invasively recorded continuous patient pressure waveform is recorded on the radial artery.

11. 4. The method of claim 3, wherein the non-invasively recorded continuous resting patient pressure waveform is recorded over the radial artery.

12. 10. The method of claim 1, wherein the non-invasively recorded continuous patient pressure waveform is recorded using a method selected from photoplethysmography and / or applanation tonometry.

13. 4. The method of claim 3, wherein the non-invasively recorded continuous resting patient pressure waveform is recorded using a method selected from photoplethysmography and / or applanation tonometry.

14. 10. The method of claim 1, wherein the time window comprises at least one complete respiratory cycle.

15. 10. The method of claim 1, wherein the lumped parameter model of the human patient comprises: a blood circulatory system (BCS) element, said BCS element comprising one, two or more CRL functional blocks; a cardiac chamber pressure volume (HPV) element, said HPV element comprising one, two or more ERv functional blocks; a coronary blood flow (CBF) component, said CBF component being a three-component model including one, two or more RCpRp functional blocks; Here, the CRL function block: - Passing the CRL functional block inlet compartment flow (q in ) and CRL functional block inlet compartment pressure (p in ) through the CRL functional block distal resistance (R) and then through the CRL functional block inertance (L) to produce an outlet CRL functional block flow (q out ) and outlet CRL functional block pressure (p out ); and The distal resistance (R) of the CRL functional block and the inertance (L) of the CRL functional block are connected in series, and the inlet compartment flow rate (q in ) of the CRL functional block and the inlet compartment pressure (p in ) of the CRL functional block are affected by the compliance (C) of the CRL functional block connected in parallel to the CRL functional block before the distal resistance (R) of the CRL functional block, Here, the ERv function block is - Passing the inlet compartment flow rate (q in ) of the ERv functional block and the inlet compartment pressure (p in ) of the ERv functional block through the distal resistance (R) of the ERv functional block and then through the heart valve model diode (valve) of the ERv functional block to generate the outlet flow rate (q out ) of the ERv functional block and the outlet pressure (p out ) of the ERv functional block; and The distal resistance (R) of the ERv functional block and the heart valve model diode (valve) of the ERv functional block are connected in series, and the inlet compartment flow rate (q in ) of the ERv functional block and the inlet compartment pressure (p in ) of the ERv functional block are also affected by the time-varying elastic modulus concept (X=E) of the ERv functional block or the myocardial fiber stress and strain concept (X=MF) of the ERv functional block, where the time-varying elastic modulus concept (E) of the ERv functional block and the myocardial fiber stress and strain concept (MF) of the ERv functional block include a structure connected in parallel to the ERv functional block before the distal resistance (R) of the ERv functional block; Here, the RCpRp function block is - Passing the RCpRp functional block inlet compartment flow rate (q in ) and the RCpRp functional block inlet compartment pressure (p in ) through the RCpRp functional block proximal resistance (R 0 ), then through the RCpRp functional block distal resistance (R), and then applying the RCpRp functional block myocardial vascular interaction (MVI) pressure for the resistance artery (p R ) to generate the RCpRp functional block outlet flow rate (q out ) and the RCpRp functional block outlet pressure (p out ); and The inlet proximal resistance (R 0 ) of the RCpRp functional block, the outlet distal resistance (R) of the RCpRp functional block, and the myocardial vascular interaction (MVI) pressure of the RCpRp functional block for a resistance artery (p R ) are connected in series, and the inlet compartment flow rate (q in ) of the RCpRp functional block and the inlet compartment pressure (p in ) of the RCpRp functional block are also affected by the compliance (C) of the RCpRp functional block, and then by the myocardial vascular interaction (MVI) pressure of the RCpRp functional block for a compliant artery (p c ), where the compliance (C) of the RCpRp functional block and the myocardial vascular interaction (MVI) pressure of the RCpRp functional block for a compliant artery (p c ) are connected in parallel with the RCpRp functional block, and the proximal resistance (R 0 ) of the RCpRp functional block ) and a distal resistance (R) of the RCpRp functional block, wherein the compliance (C) of the RCpRp functional block is connected in series with a myocardial vascular interaction (MVI) pressure of the RCpRp functional block for a compliant artery (p c ), and the compliance (C) of the RCpRp functional block is between the myocardial vascular interaction (MVI) pressure of the RCpRp functional block for a compliant artery (p c ) and the RCpRp functional block.

16. 16. The method of claim 15, wherein the myocardial vascular interaction (MVI) comprises calculating a coronary flow limiting pressure; [Equation 23] where p is the interstitial pressure and k Ｒ,Ｃ are the coefficients of the resistive and capacitive parts of the coronary artery tree, and μ １ is the cavity-induced extracellular pressure coefficient, and μ ２ is the intracellular pressure induced by the shortening, and E ｎ is the normal elastic way.

17. 16. The method of claim 15, wherein each RCpRp functional block has the formula: [0000] Fulfilling Here, p ｚｆ is the pressure at zero coronary blood flow, and / or Here, each CRL function block satisfies the following equation: [Equation 25] where C, R, and L are compliance, resistance, and inertance, respectively.

18. 18. The method of claim 17, wherein the pressure at zero flow is equal to 14.8±7 mmHg for patients with stable angina, 22.5±9.1 mmHg for patients with non-Q wave myocardial infarction, or 37.1±1.9 mmHg for patients with Q wave myocardial infarction.

19. 16. The method of claim 15, wherein the lumped parameter model of the human patient includes a systemic circulation model and a pulmonary circulation model, the systemic circulation including a left heart circuit and the pulmonary circulation including a right heart circuit, each circuit having at least two of the CRL functional blocks connected in series.

20. 16. The method of claim 15, wherein the lumped parameter model of the human patient comprises: - Passing the venous systemic pressure (P SV ) and venous systemic flow (q SV ) through the right atrium (R.A.) to generate right ventricular pressure (p RV ) and right ventricular flow (q t ); - passing the right ventricular pressure (p RV ) and right ventricular flow (q t ) through the right ventricle (RV) to generate arterial pulmonary circulation pressure (p pa ) and arterial pulmonary circulation flow (q pa ); - passing the arterial pulmonary circulation pressure (p pa ) and the arterial pulmonary circulation flow (q pa ) through a pulmonary circulation BCS to generate a venous pulmonary circulation pressure (p pv ) and a venous pulmonary circulation flow (q pv ), in which the arterial pulmonary circulation pressure (p pa ) and the arterial pulmonary circulation flow (q pa ) pass through the arterial pulmonary circulation (pa), through an arterial pulmonary circulation reservoir (pr), and then through a series-connected venous pulmonary circulation (pv); - passing said venous pulmonary circulation pressure (P PV ) and said venous pulmonary circulation flow (Q PV ) through the left atrium (LA) to generate a left ventricular pressure (p LV ) and a left ventricular flow (q m ); - Passing the left ventricular pressure (p LV ) and left ventricular flow (q m ) through the left ventricle (LV) to generate aortic systemic pressure (p sa ) ​​and aortic systemic flow (q sa ); and generating the venous systemic circulation pressure (P SV ) and the venous systemic circulation flow (Q SV ) by passing the aortic systemic pressure (p sa ) ​​and the aortic systemic circulation flow (q sa ) ​​through a systemic circulation BCS, in which the aortic systemic pressure (p sa ) ​​and the aortic systemic circulation flow (q sa ) ​​pass through an arterial systemic circulation (sa), an aortic systemic circulation reservoir (sr), and then a venous systemic circulation (sv) connected in series; HPV includes the right atrium (R.A.), the right ventricle (R.V.), the left atrium (L.A.), and the left ventricle (L.V.); and the lumped parameter model of the human patient also comprises: - Passing the CBF aortic systemic circulation pressure (p sa ) ​​and the CBF aortic systemic circulation flow (q 0 ) through a CBF aortic systemic circulation proximal resistance (R 0 ) and then through the CBF to generate a venous systemic circulation pressure (p sv ), and generating the venous systemic circulation flow (q sv ), the venous pulmonary circulation pressure (p pv ), and the venous pulmonary circulation flow (q pv ), wherein the CBF aortic systemic circulation proximal resistance (R 0 ) and the CBF are connected in series, and the CBF includes a plurality of modules connected in parallel, each module including a CBF compliance (C) connected in parallel with a CBF myocardial vascular interaction (MVI) pressure for a compliance artery (p c ), and the CBF aortic systemic circulation pressure (p sa ) ​​and the CBF aortic systemic circulation flow (q 0 ) pass through the CBF compliance (C) and then through the compliance artery (p wherein the CBF compliance (C) and the CBF-myocardial interaction (MVI) pressure for the compliance artery (p c ) are connected in parallel with the CBF resistance (R) and the CBF-myocardial interaction (MVI) pressure for the resistance artery (p R ), and the CBF resistance (R) is connected in series with the CBF-myocardial interaction (MVI) pressure for the resistance artery (p R ), and the CBF-aortic systemic circulation pressure (p sa ) ​​and the CBF-aortic systemic circulation flow (q 0 ) pass through the CBF resistance (R) and then through the CBF-myocardial interaction (MVI) pressure for the resistance artery (p R ); The right atrium (RA) contains the venous systemic circulation pressure (p sv ) and the venous systemic circulation flow rate (q sv ) passing through the right atrial resistance (R RA ), and then passing through the tricuspid atrioventricular valve (t.v.), where the right atrial resistance (R RA ) and the tricuspid atrioventricular valve (t.v.) are connected in series, and the venous systemic circulation pressure (p sv ) and the venous systemic circulation flow rate (q sv ) are also affected by the right atrial time-varying elasticity (E RA ), where the right atrial time-varying elasticity (E RA ) is connected in parallel to the right atrium (RA) and before the right atrial resistance (R RA ); The right ventricle (R.V.) includes the right ventricular pressure (p RV ) and the right ventricular flow rate (q t ) passing through the right ventricular resistance (R RV ), and then passing through the pulmonary valve (p.v.), where the right ventricular resistance (R RV ) and the pulmonary valve (p.v.) are connected in series, and the right ventricular pressure (p RV ) and the right ventricular flow rate (q t ) are also affected by the right ventricular time-varying elasticity concept (E RV ), where the right ventricular time-varying elasticity concept (E RV ) is connected in parallel to the right ventricle (R.V.) and before the right ventricular resistance (R RV ); The arterial pulmonary circulation (pa) includes the arterial pulmonary circulation pressure (Ppa) and the arterial pulmonary circulation flow rate (qpa) passing through the arterial pulmonary circulation resistance (Rpa), and then passing through the arterial pulmonary circulation inertance (Lpa), where the pulmonary artery circulation resistance (Rpa) and the arterial pulmonary circulation inertance (Lpa) are connected in series, and the arterial pulmonary circulation pressure (ppa) and the arterial pulmonary circulation flow rate (qpa) are also affected by the arterial pulmonary circulation compliance (Cpa) connected in parallel to the arterial pulmonary circulation (pa) before the arterial pulmonary circulation resistance (Rpa); The venous pulmonary circulation (pv) includes the arterial pulmonary circulation pressure (Ppa) and the arterial pulmonary circulation flow rate (qpa) that pass through the venous pulmonary circulation resistance (Rpv) in series, and the arterial pulmonary circulation pressure (ppa) and the arterial pulmonary circulation flow rate (qpa) are also affected by the pulmonary artery circulation compliance (Cpv) that is connected in parallel with the arterial pulmonary circulation (pv) and before the arterial pulmonary circulation resistance (Rpv); The left atrium (LA) contains the venous pulmonary circulation pressure (p pv ) and the venous pulmonary circulation flow rate (q pv ) that pass through the left atrial resistance (R LA ), and then pass through the mitral valve (m.v.), the left atrial resistance (R LA ) and the mitral valve (m.v.) are connected in series, the pulmonary circulation venous pressure (p pv ) and the venous pulmonary circulation flow rate (q pv ) are also affected by the left atrial time-varying elasticity (E LA ), and the left atrial time-varying elasticity (E LA ) is connected in parallel to the left atrium (LA) and before the left atrial resistance (R LA ), The left ventricle (L.V.) includes the left ventricular pressure (p LV ) and the left ventricular flow rate (q m ) passing through the left ventricular resistance (R LV ), and then passing through the aortic ventricular valve (a.v.), where the left ventricular resistance (R LV ) and the aortic ventricular valve (a.v.) are connected in series, and the left ventricular pressure (p LV ) and the left ventricular flow rate (q m ) are also affected by the left ventricular time-varying elasticity concept (E LV ), and the left ventricular time-varying elasticity concept (E LV ) is connected in parallel to the left ventricle (L.V.) and before the left ventricular resistance (R LV ); The aortic systemic circulation (sa) includes the aortic systemic circulation pressure (psa) and the aortic systemic circulation flow rate (qsa) passing through the aortic systemic circulation resistance (Rsa), and then passing through the aortic systemic circulation inertance (Lsa), where the aortic systemic circulation resistance (Rsa) and the aortic systemic circulation inertance (Lsa) are connected in series, and the aortic systemic circulation pressure (psa) and the aortic systemic circulation flow rate (qsa) are also affected by the aortic systemic circulation compliance (Csa) connected in parallel with the aortic systemic circulation (sa) and before the aortic systemic circulation resistance (Rsa); and The venous systemic circulation (sv) includes the aortic systemic circulation pressure (Psa) and the aortic systemic circulation flow rate (qsa) passing through a venous systemic circulation resistance (Rsv), and is connected in series with the venous systemic circulation resistance (Rsv), and the aortic systemic circulation pressure (psa) and the aortic systemic circulation flow rate (qsa) are also affected by a venous systemic circulation compliance (Csv) connected in parallel with the venous systemic circulation (sv) and before the venous systemic circulation resistance (Rsv).

21. A computer-readable storage medium which, when executed by a computer, causes the computer to perform the steps of the method defined in claim 1.

22. A system for assessing the hemodynamic response of a human patient to adenosine receptor agonist stimulation, said system comprising: - measuring means for non-invasively measuring resting hemodynamic parameters of a human patient, wherein said resting hemodynamic parameters include the patient's systolic blood pressure, the patient's diastolic blood pressure, and / or the patient's heart rate; - a computer system adapted to perform the steps of the method defined in claim 1.

23. 23. The system for assessing the hemodynamic response of a human patient to adenosine receptor agonist stimulation as described in claim 22, wherein the computer system further comprises recording means for non-invasively and continuously recording the human patient's resting pressure waveform.

24. The system for assessing the hemodynamic response of a human patient to adenosine receptor agonist stimulation according to claim 22, wherein the computer system comprises: - at least one computer that uses a pharmacokinetic / pharmacodynamic model to map hemodynamic parameters of a human patient at rest to hemodynamic parameters of a human patient during hyperemia; at least one computer that performs parameter identification of a lumped parameter model of a human patient using the hyperemic hemodynamic parameters of the human patient to arrive at a lumped parameter model of the human patient in a hyperemic state; wherein the at least one computer configured to map the resting hemodynamic parameters of a human patient to the hyperemic hemodynamic parameters of the human patient using a pharmacokinetic / pharmacodynamic model is configured to communicate directly or indirectly with the at least one computer configured to perform parameter identification of a lumped parameter model of the human patient using the hyperemic hemodynamic parameters of the human patient to arrive at a lumped parameter model of the human patient in a hyperemic state.

Images