A coronary QFR dynamic calculation method and system fusing individualized multi-dimensional physiological parameters

Abstract

The application discloses a coronary artery QFR dynamic calculation method and system fusing individualized multi-dimensional physiological parameters, and relates to the technical field of medical devices and computer-aided diagnosis. The method comprises the following steps: collecting multi-dimensional real-time physiological parameters of a patient, and constructing individualized blood rheology indexes; calculating individualized blood viscosity based on hematocrit, and correcting microcirculation viscosity; reconstructing a coronary three-dimensional geometric model and extracting anatomical features; dynamically correcting reference microcirculation resistance according to heart function indexes; taking the individualized viscosity and the corrected resistance as boundary conditions into a rapid pressure drop equation, simulating the dynamic changes of blood flow and pressure, and then calculating individualized QFR values and visually outputting the same. The application introduces a patient-specific viscosity and a microcirculation resistance dynamic correction mechanism, and breaks through the limitation of traditional QFR which depends on standardized parameters. The method can adjust blood flow boundary conditions according to real-time physiological states, and realizes accurate quantification of coronary ischemia degree.

Description

Technical Field

[0001] This invention relates to the fields of medical devices and computer-aided diagnostic technology, specifically to a method and system for dynamic calculation of coronary artery QFR that integrates individualized multidimensional physiological parameters. Background Technology

[0002] Quantitative Flow Reserve (QFR) is an emerging coronary functional assessment technique based on computational fluid dynamics (CFD) or rapid pressure drop equations. Compared to the traditional gold standard, fractional flow reserve (FFR), QFR eliminates the need for pressure guidewires and adenosine-induced hyperemia, rapidly calculating flow reserve values ​​solely through three-dimensional reconstruction of ICA images. It offers significant advantages such as minimal invasiveness, shorter processing time, and lower cost, and is gradually gaining widespread adoption in clinical diagnosis.

[0003] However, existing QFR analysis algorithms, in pursuit of computational speed, have oversimplified complex physiological boundary conditions, resulting in the following significant technical bottlenecks when dealing with complex clinical cases: 1. Rigidity of Hemorheological Models (Regarding "Blood Viscosity"): Current QFR calculations typically assume blood is an isotropic Newtonian fluid and set blood viscosity as a fixed constant (usually 3.5 mPa·s or 4.0 mPa·s). This "one-size-fits-all" approach ignores the significant influence of factors such as hematocrit (Hct) and whole blood protein concentration on blood viscosity. In reality, blood is a non-Newtonian fluid. In patients with anemia (significantly reduced viscosity) or polycythemia / hyperlipidemia (abnormally increased viscosity), the actual blood viscosity deviates greatly from the pre-assumed value, directly leading to distorted frictional pressure drop calculation results. 2. Idealization of Microcirculation Boundary Conditions (Regarding "Resistance Models"): Existing algorithms typically rely on Murry's Law or its modified formulas, statically extrapolating distal microcirculation resistance based solely on the proximal vessel geometry (diameter, length), and assuming that all patients' microcirculations can reach the theoretically maximum expansion state (maximum congestion). However, in patients with heart failure, hypertensive heart disease, diabetes, or microcirculatory disorders (CMD), the microcirculatory vascular bed is pathological, and its diastolic reserve is impaired, failing to reach the pre-set low-resistance state. Continuing to use a generic static model leads to an overestimation of blood flow reserve in these patients (false negatives). 3. Staticization of Hemodynamic State (addressing "physiological fluctuations"): To simplify numerical solutions, existing techniques often use mean inlet velocity or mean arterial pressure as steady-state boundary conditions, neglecting the real-time impact of blood pressure fluctuations, instantaneous heart rate (HR), and cardiac output (CO) changes on coronary perfusion pressure gradients during the cardiac cycle. Especially in high-dynamic states (such as tachycardia) or hemodynamically unstable conditions (such as shock or severe valvular disease), static models cannot capture transient flow field characteristics, further amplifying computational errors. In summary, due to the neglect of significant physiological and pathological differences among individual patients, existing QFR techniques based on standardized parameters exhibit significantly reduced diagnostic accuracy in specific patient populations, often resulting in discrepancies between the QFR measured by the pressure guidewire and the actual FFR obtained. Therefore, overcoming the limitations of standardized parameter models and deeply integrating patient-specific physiological and biochemical indicators (such as complete blood count data and real-time hemodynamic parameters) into the QFR calculation model to construct adaptive boundary conditions for accurate functional assessment is a pressing technical challenge in this field. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention is implemented through the following technical solution: Firstly, a method for dynamically calculating coronary artery QFR that integrates individualized multidimensional physiological parameters includes the following steps: S1. Collect multidimensional real-time physiological parameters of the patient, including at least cardiac function indicators and blood rheology data, construct individualized blood rheology indicators, and realize comprehensive, dynamic, and high-precision collection and integration of the patient's physiological state. S2. Based on the hematocrit in the blood rheology data, the individualized whole blood viscosity is dynamically calculated using the modified Einstein equation, and the Valeo-Lingquist effect is introduced to correct the microvascular viscosity within the microvascular range, so as to reflect the real flow state of blood in the microcirculation and improve the physiological accuracy and individualization of blood viscosity calculation. S3. Reconstruct a three-dimensional geometric model of the coronary arteries using coronary angiography images, and extract the geometric features of the blood vessels from the three-dimensional geometric model of the coronary arteries to provide an accurate anatomical basis for subsequent blood flow simulation; S4. Based on the geometric features of the blood vessels, the reference microcirculation resistance of the coronary artery tree is calculated using the scaling law, and the reference microcirculation resistance is dynamically corrected according to the cardiac function indicators to obtain the individualized actual microcirculation resistance, thereby realizing the individualized dynamic correction of microcirculation resistance. S5. Substitute the individualized whole blood viscosity or corrected microvascular viscosity, and the individualized actual microcirculation resistance as boundary conditions into the rapid pressure drop equation, couple the frictional pressure drop along the flow path with the local expansion / separation pressure drop, simulate the dynamic changes in blood flow and pressure distribution during the cardiac cycle, and output the pressure distribution and velocity distribution along the vessel centerline to achieve efficient and stable transient hemodynamic simulation. S6. Calculate the individualized QFR value based on the pressure at the distal end of the stenotic lesion and the average pressure at the coronary ostium, and visualize the output on the three-dimensional geometric model of the coronary artery to complete the accurate assessment of coronary artery function and realize the intuitive three-dimensional visualization of individualized QFR.

[0005] Preferably, S1 specifically includes: The system automatically acquires real-time hemodynamic parameters of patients, including heart rate, systolic blood pressure, diastolic blood pressure, cardiac output, ejection fraction, and body surface area, through clinical information systems or bedside monitoring equipment, and forms a cardiac function vector as the cardiac function index, thus realizing the automated synchronous acquisition of multi-source real-time physiological data. Simultaneously extract blood rheological parameters, including hematocrit, hemoglobin concentration, and total whole blood protein, construct blood state characteristics, establish blood rheological characteristics based on actual test indicators, and improve the individualized realism of viscosity modeling; The cardiac function vector and blood state features were fused and normalized to construct individualized blood rheological indices, which were then integrated to generate an individualized multidimensional physiological parameter dataset for subsequent modeling.

[0006] Preferably, S2 specifically includes: Based on the collected hematocrit values, the modified Einstein viscosity equation is used to calculate individualized whole blood viscosity, replacing the fixed viscosity assumption in the traditional QFR and overcoming the systematic error caused by the standardization assumption. For microcirculatory vascular segments, the system automatically identifies whether a segment belongs to the microvascular range based on the vessel diameter and dynamically adjusts the blood viscosity within that range to achieve dynamic adaptive adjustment of viscosity in the microcirculatory area. By introducing the Valeo-Lingquist effect, the viscosity of microvascular segments with diameters smaller than a set threshold is modified to reflect the non-Newtonian fluid characteristics of blood in microcirculation.

[0007] Preferably, S2 further includes: By setting a blood vessel diameter threshold of 0.3 mm, the system automatically identifies blood vessel segments in the coronary artery tree with diameters smaller than this threshold as microcirculation zones, thereby achieving automatic and accurate division of microcirculation zones. Based on the relative relationship between blood vessel diameter and red blood cell size, a viscosity decay function was used to gradually decay individualized whole blood viscosity, and the corrected viscosity value was used as the rheological boundary condition of the microcirculation region to establish a viscosity continuous decay model based on vessel diameter.

[0008] Preferably, S3 specifically includes: Acquire two coronary angiography images with a projection angle difference of not less than 25 degrees to ensure that the target vessel segment does not overlap and that the contrast agent is adequately filled; The two coronary angiography images are registered and reconstructed in three dimensions using epipolar constraint and edge detection algorithms to generate a three-dimensional geometric model of the coronary artery and a mesh on the surface of the lumen, thus achieving high-precision automatic conversion from two-dimensional images to three-dimensional models. Geometric features are extracted from the three-dimensional geometric model of the coronary artery, including the coordinates of the vessel centerline, the diameter of each segment, the length of the segment, and the stenotic lesion segment.

[0009] Preferably, S4 specifically includes: Based on the reconstructed three-dimensional geometric model of the coronary arteries, the reference microcirculation resistance of the coronary artery tree is calculated according to the scaling law based on the total cross-sectional area of ​​the terminal vessels. The basic resistance is calculated through morphological drive, providing a precise anatomical basis for individualized correction. The myocardial oxygen consumption index is calculated based on the product of the patient's real-time heart rate and systolic blood pressure, and the cardiac function status is assessed based on the cardiac output index, providing key physiological input for dynamic correction. Based on the myocardial oxygen consumption index and cardiac function status, the reference microcirculation resistance is dynamically weighted and corrected through a predefined resistance adjustment function to obtain individualized actual microcirculation resistance. This achieves dynamic individualization of resistance boundary conditions and more realistically reflects the patient's real-time physiological and pathological state.

[0010] Preferably, S4 further includes: A function was constructed to establish an inverse relationship between myocardial oxygen consumption index and microcirculation resistance. When myocardial oxygen consumption increases, the resistance coefficient is automatically reduced to simulate the metabolic vasodilation effect and improve the physiological realism of blood flow assessment under hypermetabolic conditions. A cardiac function status classification mechanism is introduced to automatically increase the microcirculation resistance correction weight for patients with heart failure or low cardiac output, and use the corrected actual microcirculation resistance as a downstream boundary condition for resistance assignment in fluid dynamics solutions. This automatically adjusts the resistance for patients with heart failure, improving the accuracy of blood flow reserve assessment under low cardiac output conditions.

[0011] Preferably, S5 specifically includes: The individualized whole blood viscosity or microvascular viscosity, along with the individualized actual microcirculation resistance, are used as boundary conditions and input into a fluid dynamics solver based on the fast pressure drop equation to achieve high-fidelity fluid dynamics simulation and ensure the physiological authenticity of blood flow calculations. In the time domain, with the cardiac cycle as the time step, the total pressure drop is calculated by coupling the frictional pressure drop along the path with the local expansion / separation pressure drop, and the blood flow and pressure distribution are dynamically simulated to realize real-time dynamic simulation of blood flow and pressure within the cardiac cycle, thereby improving the time domain accuracy of the simulation. It outputs instantaneous pressure and flow velocity data at various points along the centerline of the blood vessel, forming a hemodynamic distribution map within the cardiac cycle, providing comprehensive and dynamic scientific parameter support for clinical assessment and improving the integrity of diagnostic evidence.

[0012] Preferably, S6 specifically includes: The pressure distal to the stenotic lesion and the mean pressure at the coronary ostium are extracted from the hemodynamic distribution results. The ratio of distal pressure to ostial pressure is calculated as an individualized QFR value, which objectively and directly yields the individualized fractional flow reserve, providing a key quantitative indicator for clinical functional assessment. Individualized QFR values ​​are overlaid on the surface of the coronary artery 3D geometric model using color mapping, and numerical annotations are made at key anatomical locations to form a visual assessment report. This report intuitively shows the correspondence between blood flow function and anatomical structure, helping doctors to quickly locate functional abnormal areas and interpret the results.

[0013] Secondly, a coronary artery QFR dynamic calculation system integrating individualized multidimensional physiological parameters is provided to implement the above method, including: The data acquisition and preprocessing module is used to automatically connect to the clinical information system and monitoring equipment, acquire and integrate multi-dimensional physiological parameters such as patient hemodynamic data, cardiac function indicators and blood rheology data in real time, complete data alignment, outlier removal and normalization, provide standardized input for subsequent modeling, and effectively solve the problem of traditional data relying on manual entry. The individualized blood viscosity modeling module calculates individualized whole blood viscosity using the modified Einstein equation based on the hematocrit data of the blood rheology data, and dynamically adjusts and corrects microvascular viscosity according to the vessel diameter to more realistically reflect the non-Newtonian fluid properties of blood. The 3D vessel reconstruction and geometric analysis module, based on coronary angiography images, reconstructs the 3D geometric model of the coronary arteries through epipolar constraint and edge detection algorithms, and automatically extracts the geometric features of the vessels, providing an anatomical basis for blood flow simulation; The adaptive microcirculation resistance construction module is used to combine vascular geometry features and cardiac function indicators to dynamically correct the reference microcirculation resistance, simulate vasomotor regulation under different pathological conditions, and achieve adaptive construction of boundary conditions. The QFR calculation and hemodynamic solution module uses individualized viscosity and individualized actual microcirculation resistance as boundary conditions to solve transient blood flow and pressure distribution based on the rapid pressure drop equation. Finally, it calculates the pressure ratio at the distal end of the stenosis and the coronary ostium, outputs individualized QFR values ​​and three-dimensional visualization results. By coupling individualized boundary conditions and transient solutions, it achieves blood flow simulation that is closer to the real physiological state.

[0014] The present invention has the following beneficial effects: (I) Overcoming the limitations of standardized models and improving assessment accuracy: This invention breaks through the constraints of traditional QFR technology, which relies on fixed parameters, by constructing individualized blood rheology indicators and introducing a dynamic correction mechanism for patient-specific whole blood / microvascular viscosity and microcirculation resistance. It can adaptively adjust blood flow boundary conditions based on the patient's actual physiological and pathological state, effectively avoiding calculation biases caused by ignoring individual differences (such as anemia and heart failure), thereby achieving accurate quantification of the degree of coronary ischemia and significantly enhancing the reliability of diagnosis.

[0015] (II) Achieving high-fidelity transient blood flow simulation, closely resembling real physiological behavior: This invention, by integrating multidimensional real-time physiological parameters and combining them with the non-Newtonian fluid characteristics of microvessels (Valeo-Lingquist effect), can simulate transient changes in blood flow and pressure during the cardiac cycle in the time domain. Combined with microcirculatory resistance automatically adjusted according to vascular geometry and cardiac function, this method can more realistically reproduce the dynamic characteristics of coronary blood flow under complex pathological conditions, providing a simulation basis that closely approximates physiological reality for functional assessment.

[0016] (III) Automated Integration of Multi-Source Data to Improve Clinical Workflow Efficiency: This invention breaks down data barriers between hospital information systems, monitoring equipment, and laboratory information systems, enabling automatic acquisition, timestamp alignment, and quality control of multi-dimensional parameters of hemodynamics, cardiac function, and blood rheology. Through standardized interfaces, high-quality input datasets can be generated within seconds, completely eliminating the tediousness and errors of manual data entry, and greatly improving the convenience of clinical applications and the robustness of data processing. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the workflow of a method and system for dynamic calculation of coronary artery QFR that integrates individualized multidimensional physiological parameters, according to the present invention. Figure 2 This is a data flow diagram of a coronary artery QFR dynamic calculation system that integrates individualized multidimensional physiological parameters, according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0019] Example 1, please refer to Figure 1 This invention provides a technical solution: a method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters, comprising the following steps: S1. Collect multidimensional real-time physiological parameters of patients and automatically integrate hemodynamic and rheological data to construct individualized blood rheological indicators. This step achieves comprehensive, dynamic, and high-precision collection and integration of patients' physiological states, laying a data foundation for accurate modeling. Specifically, this includes: automatically acquiring patients' real-time hemodynamic parameters (heart rate, systolic blood pressure, diastolic blood pressure, cardiac output, ejection fraction, and body surface area) through clinical information systems or bedside monitoring equipment to form a cardiac function vector; simultaneously extracting blood rheological parameters (hematocrit, hemoglobin concentration, and total whole blood protein) from electronic medical records or laboratory information systems to construct blood state features; finally, fusing and normalizing the cardiac function vector and blood state features to generate an individualized multidimensional physiological parameter dataset for subsequent modeling.

[0020] It should be noted that in actual operation: Hemodynamic data acquisition: Relying on standardized data interfaces with the Hospital Information System (HIS) and bedside monitoring equipment, waveform and numerical data such as heart rate, systolic blood pressure, and diastolic blood pressure are received in real time from the monitoring equipment via HL7 or FHIR protocols, with a sampling frequency of no less than 1Hz to ensure data continuity. Cardiac output is acquired through a transthoracic or transesophageal echocardiography module connected to the monitoring system, with measurement errors controlled within a clinically acceptable range of ±15%. Ejection fraction is automatically obtained from recent (within 24 hours) structured echocardiography reports, and body surface area is automatically calculated using the Du Bois formula based on height and weight recorded in the electronic medical record. After data acquisition, timestamp alignment and outlier filtering are performed to form a standardized, time-synchronized cardiac function vector data sequence.

[0021] Hemorheological data acquisition: The system interfaces with the Laboratory Information System (LIS) to automatically retrieve the patient's most recent complete blood count and biochemical test results (sampled within 72 hours prior to coronary angiography) via a standardized interface. Key extracted parameters include: hematocrit (Hct, normal reference range: males 38.8%-50.0%, females 34.9%-44.5%), hemoglobin concentration (Hb, normal reference range: males 13.5-17.5 g / dL, females 12.0-15.5 g / dL), and total whole blood protein (normal reference range: 6.0-8.5 g / dL). For data lacking hematocrit but containing hemoglobin concentration, an approximate estimate is made using the empirical formula Hct ≈ Hb × 3.

[0022] Data fusion and preprocessing: The time-window aligned cardiac function vectors are correlated with static hemorheological parameters. For continuous physiological parameters, z-score standardization based on a large-scale healthy population database is used; for blood parameters, the original measurements are used directly and their units are standardized to the International System of Units (SI), ultimately generating an input dataset of individualized multidimensional physiological parameters.

[0023] S2. Based on hematocrit, individualized whole blood viscosity is dynamically calculated using a modified Einstein equation, and the Valeo-Lingquist effect is introduced to correct for microvascular viscosity. This step replaces the fixed viscosity assumption used in traditional QFR, overcomes the systematic errors caused by the standardization assumption, and significantly improves the accuracy of microcirculation blood flow simulation. Specifically, this includes: for microcirculatory vascular segments, automatically identifying whether they belong to the microvascular range based on the vessel diameter, and dynamically adjusting the blood viscosity accordingly to achieve dynamic adaptive adjustment of viscosity in the microcirculatory region, enhancing the physiological realism of the model in microvascular segments; introducing the Valeo-Lingquist effect to perform secondary correction on the viscosity of microvascular segments with diameters smaller than a set threshold, thereby reflecting the non-Newtonian fluid characteristics and actual flow state of blood in the microcirculation, further conforming to the rheological behavior of blood in microvascular segments, and improving the accuracy of microcirculatory blood flow simulation; in addition, setting a vessel diameter threshold of 0.3 mm, automatically identifying vessel segments with diameters smaller than this threshold in the coronary artery tree as microcirculatory regions, and using a viscosity decay function to gradually decay the whole blood viscosity, using the corrected viscosity value as a rheological boundary condition to achieve a seamless transition from macroscopic to microscopic blood flow characteristics; It should be noted that the specific calculation and dynamic correction process for viscosity is as follows: Personalized whole blood viscosity calculation: Based on real-time acquired patient hematocrit (Hct) values, personalized whole blood viscosity is calculated to replace the traditionally set fixed viscosity value of 3.5 mPa·s. The specific calculation follows the formula: ,in, To calculate the individualized whole blood viscosity, This is the plasma viscosity (a constant, approximately 1.2 mPa·s). This is the red blood cell aggregation coefficient (which can be taken as an empirical value or adjusted according to the whole blood protein concentration). Hematocrit. The system processes the input... The viscosity value is validated (acceptable range is 15.0% to 65.0%), and values ​​exceeding the limit are flagged and calculated using boundary values, while a clinical alert log is generated. This viscosity value serves as a baseline parameter for macroscopic vascular segments (diameter ≥ 0.3 mm).

[0024] Preliminary viscosity reduction in the microcirculation zone: Based on a three-dimensional geometric model, vessel segments with an inner diameter less than 0.3 mm marked along the vessel centerline were classified as microcirculation zones. This threshold was set based on the anatomical size range (0.2-0.35 mm) of precapillary arterioles in the human body. For this region, a linear decay function was used for initial correction based on vessel diameter. .in, To initially correct the viscosity; Individualized whole blood viscosity, The current vascular segment diameter (unit: mm) is used to reflect the decrease in apparent viscosity of blood in microvessels due to the axial migration of red blood cells. At the same time, the original diameter, corrected viscosity value and spatial location information of each microcirculation segment are recorded. Secondary correction of the Valeo-Lingquist effect: For blood vessel segments with a diameter less than 0.2 mm, a secondary correction is performed using a nonlinear decay model based on classical fluid dynamics experimental data. ,in, This is the microvascular blood viscosity value after secondary correction. The attenuation coefficient is set to 0.35. The viscosity was set to an exponent (1.5). This model accurately simulates the physical phenomenon of a significant decrease in apparent blood viscosity when the vessel diameter approaches the size of a red blood cell (approximately 0.008 mm). For vessel segments with a diameter less than 0.05 mm, the viscosity value was restricted to 1.1 to 1.3 times the plasma viscosity. The corrected viscosity value was used as the local rheological boundary condition for the corresponding microvascular segment and directly applied to the mesh element property definition of the subsequent fluid dynamics solver. S3. Reconstruct a three-dimensional geometric model of the coronary arteries using coronary angiography images, and extract the geometric features of the vessels from the three-dimensional geometric model. This step provides a high-precision and efficient anatomical basis and reliable structural input for subsequent blood flow simulation. Specifically, acquire two coronary angiography images with a projection angle difference of no less than 25 degrees to ensure that the target vessel segment does not overlap and that the contrast agent is fully filled, effectively guaranteeing the clarity and integrity of the original images required for three-dimensional reconstruction. Use epipolar constraint and edge detection algorithms to perform three-dimensional registration and reconstruction of the two images, generating a three-dimensional geometric model of the coronary arteries and a mesh on the surface of the lumen, achieving high-precision automatic conversion from two-dimensional images to three-dimensional models, providing accurate anatomical basis for hemodynamic calculations. Finally, extract geometric features such as the vessel centerline coordinates, vessel diameter of each segment, vessel segment length, and stenotic lesion segment from the model, automatically extracting morphological and lesion information to support the accurate setting of individualized boundary conditions. It should be noted that in actual operation: Image quality control and preprocessing: Acquired angiographic images must conform to the DICOM 3.0 standard, with a resolution of no less than 1024×1024 pixels and a grayscale depth of 10-12 bits. To ensure sufficient stereo parallax to reduce the uncertainty of coordinate calculation, the difference in projection angle between two images should be no less than 25°, preferably between 30° and 60°. The system uses automated algorithms to evaluate and ensure the clear outline of the target vascular segment. For image noise and artifacts, adaptive median filtering and background subtraction techniques are used for preprocessing. Within 10 seconds of image loading, the system automatically identifies and records the projection geometry parameters (such as the distance from the X-ray source to the detector set to 100-120cm, the projection angle, and pixel calibration information of 0.15-0.30mm / pixel) and generates a data quality report.

[0025] • High-precision 3D reconstruction: Epipolar geometric constraints and high-precision edge detection algorithms are applied, combined with prior knowledge of coronary artery anatomy. Multi-scale Gabor filtering and the Canny operator are used to extract the vessel lumen contour. Subsequently, the epipolar geometry principle is used to match corresponding contour points. By solving the fundamental matrix and using the RANSAC algorithm to remove mismatched points, a sub-pixel-level precision 3D point cloud is generated (the iterative reconstruction algorithm stably handles vessel segments with a diameter ≥1.0 mm, with errors controlled within 0.1-0.3 mm). After Poisson surface reconstruction and Laplacian smoothing, the point cloud is transformed into a non-uniform rational B-spline surface, ultimately outputting the coronary artery 3D geometric model and surface mesh. The mesh cell size in the stenotic region is automatically refined to 0.05-0.1 mm to ensure accurate representation of complex geometric morphology.

[0026] Geometric feature quantification and extraction: A minimum cost path algorithm is used to optimize the centerline of the entire vessel. Sampling is performed along the centerline at 0.1 mm intervals, and the maximum inscribed circle diameter of the inner wall is calculated on the normal plane, forming a continuous diameter distribution curve. The system automatically identifies continuous segments on the diameter curve that are ≥30% smaller than the proximal reference segment, marks them as stenotic lesions, and records their starting position, length, minimum luminal diameter, and percentage of stenosis. All extracted parameters are output as structured data, serving as definitive anatomical inputs.

[0027] S4. Based on the geometric features of blood vessels, the reference microcirculatory resistance of the coronary artery tree is calculated using the scaling law, and the reference microcirculatory resistance is dynamically corrected according to cardiac function indicators to obtain individualized actual microcirculatory resistance. This step achieves dynamic individualization of resistance boundary conditions, significantly improving the model's adaptability and accuracy under complex pathological conditions. Specifically, based on the reconstructed model, the reference microcirculatory resistance (basic resistance) is calculated using the scaling law based on the total cross-sectional area of ​​the terminal vessels; the myocardial oxygen consumption index is calculated based on the product of the patient's real-time heart rate and systolic blood pressure, and cardiac function is assessed in real time in conjunction with the cardiac output index; based on the myocardial oxygen consumption index and cardiac function, dynamic weighted correction is performed through a predefined resistance adjustment function. In addition, the resistance coefficient is automatically reduced when myocardial oxygen consumption increases to simulate metabolic vasodilation; the resistance correction weight is automatically increased for patients with heart failure or low cardiac output, improving the accuracy of blood flow reserve assessment under low cardiac output conditions.

[0028] It should be noted that the calculation and correction process for microcirculation resistance is as follows: Reference microcirculation resistance calculation: The cross-sectional information of all terminal vessel branches in the model is automatically identified to obtain the diameter of the terminal coronary artery. Based on the scaling law, the reference microcirculation resistance has a specific functional relationship with the diameter of the terminal coronary artery: ; In the formula: For reference microcirculation resistance; The diameter of the terminal coronary artery; The morphological index (typically 2.2-3.0, based on large-scale human anatomical data, to reasonably estimate normal resting-state basal resistance) is used. All parameters required for calculation are automatically measured in millimeters, and the system outputs the result within 30 seconds of 3D reconstruction completion. .

[0029] Cardiac function status parameter extraction: Hemodynamic data (sampling rate ≥1Hz) is acquired in real time via an interface. After time alignment and moving average filtering, the myocardial oxygen consumption index is calculated every 60 seconds, and the cardiac output index is calculated in combination with body surface area. ; ; In the formula: The myocardial oxygen consumption index (the higher the index, the greater the oxygen demand; metabolic expansion should reduce resistance). This refers to real-time aortic systolic blood pressure. Heart rate; For cardiac output; It represents the body surface area.

[0030] Individualized resistance dynamic correction: The reference resistance is corrected using a predefined resistance adjustment function and updated every 30 seconds. ; In the formula: For individualized actual microcirculation resistance; As a factor of myocardial oxygen consumption index, when When the level rises, this factor automatically decreases, mimicking metabolic vasodilation; As a factor of cardiac function status, when This factor automatically increases when the heart rate falls below the normal threshold (heart failure), reflecting compensatory contraction of the microcirculation. Both weighted factors are calibrated based on large-scale clinical data. Final adjusted... As downstream boundary conditions for solving the problem.

[0031] S5. The individualized whole blood viscosity or corrected microvascular viscosity, along with the individualized actual microcirculatory resistance, are substituted into the fast pressure drop equation as boundary conditions. This couples the frictional pressure drop along the vessel walls with the local expansion / separation pressure drop, outputting the pressure and velocity distribution along the vessel centerline. This step achieves efficient and stable transient hemodynamic simulation, providing high spatiotemporal resolution blood flow and pressure data for functional assessment. Specifically, individualized whole blood / microvascular viscosity and corrected actual microcirculatory resistance are used as key boundary conditions and input into a fluid dynamics solver based on the fast pressure drop equation to ensure the physiological realism of the blood flow calculation. In the time domain, the total pressure drop is calculated by coupling the frictional pressure drop along the vessel walls with the local expansion / separation pressure drop, using the cardiac cycle as the time step. By dynamically simulating the real-time evolution of blood flow and pressure, the temporal accuracy of the simulation is improved. Finally, instantaneous pressure and velocity data at various points along the vessel centerline are output, forming a hemodynamic distribution map, providing comprehensive and dynamic parameter support for clinical diagnosis.

[0032] The formula for calculating the total pressure drop is as follows: ; ; ; In the formula: Total pressure drop; Frictional pressure drop (Pa) is the pressure loss caused by the viscosity of blood flowing in a straight pipe section. Darcy friction factor; This refers to the length of the blood vessel segment; Blood density; Blood flow rate; The diameter of the blood vessel; The pressure drop (Pa) caused by local dilation / separation due to vascular stenosis or geometrical abrupt changes is classified as kinetic energy loss. This is the local loss coefficient; It should be noted that, in actual operation, the solver configuration and execution logic are as follows: Solver configuration and boundary conditions: Spatial discretization is performed using the finite volume method, and a second-order implicit scheme is used for time integration to ensure numerical stability and computational efficiency. The inlet boundary is set as a time-varying aortic pressure waveform (generated by interpolation of real-time acquired arterial blood pressure data, with a sampling interval of 0.01 seconds); the outlet boundary adopts a resistance-type boundary condition defined by individualized actual microcirculation resistance to simulate downstream impedance characteristics.

[0033] Mesh independence and residual control: Mesh independence verification is automatically performed during the initialization phase (ensuring that the pressure calculation change at key locations is less than 1% when the number of meshes increases by 10%). Residuals are monitored in real time during calculation, and the convergence criteria for continuity and momentum equations are set to 1×10⁻. 4 A progress log is output every 5 iterations.

[0034] Adaptive time-domain solution: The simulation time window (0.6 to 1.2 seconds) is based on a single cardiac cycle corresponding to the patient's current heart rate. An adaptive time-stepping strategy is employed, with an initial step size of 0.005 seconds, automatically narrowed to 0.001 seconds in regions of drastic flow velocity changes. Within each time step, the frictional pressure drop along the flow path is calculated using the modified Hagen-Poiseuille formula (utilizing individualized viscosity and geometric parameters); the local pressure drop is calculated by multiplying the local energy loss coefficient (obtained from a table) by the square of the flow velocity term, and the transient Navier-Stokes equations are iteratively solved.

[0035] Data Output and Visualization: The solver outputs a full-field flow field snapshot every 0.1 seconds and records instantaneous pressure and flow velocity at the vessel centerline (sampling interval 0.1 mm). The system automatically calculates time-averaged pressure, peak flow velocity, and pressure gradient curves. The data is stored in standard HDF5 format with accompanying metadata. Simultaneously, a two-dimensional color mapping map is generated and projected onto the three-dimensional model, supporting dynamic playback of blood flow evolution during the cardiac cycle.

[0036] S6. Calculate the individualized QFR value based on the pressure distal to the stenotic lesion and the mean pressure at the coronary ostium, and visualize the result on a three-dimensional geometric model of the coronary artery. This step objectively and directly derives the individualized fractional flow reserve, enabling accurate assessment of coronary artery function. Specifically, the ratio of the pressure distal to the stenotic lesion to the mean pressure at the coronary ostium is extracted from the hemodynamic distribution results and calculated as the individualized QFR value. This value is then overlaid on the surface of the three-dimensional geometric model of the coronary artery using a color mapping method, with numerical annotations at key anatomical locations, forming a structured, visualized assessment report. This intuitively demonstrates the correspondence between blood flow function and anatomical structure, assisting clinicians in quickly locating abnormal areas and making reliable decisions.

[0037] It should be noted that the QFR calculation and visualization report generation process is as follows: Pressure extraction and QFR calculation: A measurement point 2 mm distal to the stenotic lesion is located based on the vessel centerline markings. Pressure data from the entire cardiac cycle are extracted and a time-weighted average is calculated to obtain the pressure distal to the stenotic lesion. (Unit: mmHg). The mean pressure at the coronary ostium (left main coronary artery or right coronary ostium) was simultaneously extracted. The formula for calculating the individualized QFR is: The system will perform quality verification on the pressure curve. If there are non-physiological drastic fluctuations in the distal pressure, it will automatically trigger resampling and prompt for verification.

[0038] High-fidelity 3D visualization: A continuous color gradation (preset red QFR≤0.75 to green QFR≥0.90, with a yellow-orange gradient in between) is used to smoothly map QFR values ​​to the vertices of the 3D surface mesh in 256 levels. The system automatically generates arrowed bounding boxes (displaying QFR values, local pressure, and vessel diameter, using a 12-point sans-serif font) at the narrow proximal reference segment, the narrowest point, the distal 2 mm, and the bifurcation point. All operations are accelerated by a dedicated GPU, ensuring the entire process from data extraction to 3D rendering is completed within 3 seconds, supporting interactive rotation and zooming.

[0039] Structured Report Archiving: The system automatically generates assessment reports compatible with the DICOM-SR standard. The content includes patient information, a summary of calculated parameters, a QFR value table (detailing each major vessel segment (left anterior descending artery, left circumflex artery, right coronary artery and its major branches, identifying significant stenosis with QFR ≤ 0.80), image quality scores, and convergence indices. This report can be directly transferred to the electronic medical record system via the hospital's HIS system and a clinically interpretable PDF version is generated. All intermediate calculation data are archived in a standard format to ensure the traceability of results.

[0040] Example 2, as Figure 1 , Figure 2 As shown, based on Example 1, the present invention also provides a coronary artery QFR dynamic calculation system that integrates individualized multidimensional physiological parameters, used to implement the above-mentioned method for coronary artery QFR dynamic calculation that integrates individualized multidimensional physiological parameters, including: The data acquisition and preprocessing module automatically interfaces with clinical information systems and monitoring equipment to acquire and integrate multi-dimensional physiological parameters of patient hemodynamics, cardiac function indicators, and blood rheology in real time. This module automatically performs data alignment, outlier removal, and normalization, providing standardized input for subsequent modeling. This design effectively solves the problem of traditional data reliance on manual entry, achieving a high degree of automation and standardized integration of information, laying a high-quality data foundation for accurate modeling. The personalized blood viscosity modeling module calculates individualized whole blood viscosity based on patient blood parameters (such as hematocrit) using a modified fluid dynamics model (such as the Einstein equation), and dynamically adjusts and corrects the viscosity in the microcirculation zone (i.e., microvascular viscosity) according to the vessel diameter. This module more realistically reflects the non-Newtonian fluid characteristics of blood, breaking the traditional assumption of a universal fixed viscosity, and achieving accurate simulation of rheological properties under different blood conditions, significantly improving the physiological realism of hemodynamic calculations. The 3D vessel reconstruction and geometric analysis module, based on coronary angiography images (such as two-position angiography images), reconstructs a 3D geometric model of the coronary arteries through epipolar constraint and edge detection algorithms, and automatically extracts the geometric features of the vessel centerline, diameter distribution, and stenosis lesions. This module achieves automated, high-precision reconstruction from 2D images to 3D geometric models, providing accurate and reliable anatomical input for blood flow simulation and effectively avoiding subjective errors from manual measurements. An adaptive microcirculatory resistance construction module is used to dynamically adjust the reference microcirculatory resistance by combining the morphological characteristics of the terminal coronary arteries (such as vascular geometry) with the patient's real-time cardiac function indicators, thereby obtaining individualized actual microcirculatory resistance. This module simulates vasomotor regulation under different pathological states and achieves adaptive construction of individualized blood flow boundary conditions. This overcomes the limitations of general static resistance models, can dynamically respond to the patient's physiological and pathological changes, and provides more reasonable boundary conditions for blood flow assessment in complex cases. The QFR calculation and hemodynamic solution module uses individualized whole blood viscosity or microvascular viscosity, along with a corrected individualized actual microcirculatory resistance, as boundary conditions to solve for transient blood flow and pressure distribution based on the rapid pressure drop equation. It ultimately calculates the pressure ratio at the distal end of the stenosis to the coronary ostium, outputting individualized QFR values ​​and 3D visualization results. By coupling individualized boundary conditions with transient solutions, this module achieves a blood flow simulation that more closely approximates real physiological conditions, thus outputting functional assessment results with higher clinical consistency and reliability.

[0041] Example 3, as Figure 1 , Figure 2 As shown, based on Embodiments 1 and 2, the present invention provides a complete operational example of a technical solution: it runs on a high-performance medical workstation equipped with an NVIDIA GeForce RTX 3080 GPU. The system software environment includes a Windows 10 operating system, a Python 3.8 computing backend, and a CUDA-accelerated fluid dynamics solver. The system is connected to the hospital's HIS system via an HL7 protocol interface and to the catheterization lab angiography machine via a DICOM network.

[0042] The coronary QFR calculation method based on individualized hemodynamic parameters described in this embodiment is applied to a patient with anemia and a hyperdynamic state, and specifically includes the following steps: Step S1: Multimodal data acquisition and preprocessing, image data: select two images with different projection angles. 25° coronary angiography (ICA) images (e.g., LAO 45° and RAO 30°) ensure no overlap of the target vessel segment and good contrast filling; Physiological parameter acquisition: The system automatically extracts the patient's most recent (within 24 hours) blood routine data and intraoperative monitoring data from the electronic medical record.

[0043] Example of specific parameters in this embodiment: Hematocrit (Hct): 32% (suggesting mild anemia); Total protein (TP) in whole blood: 65 g / L; Intraoperative real-time heart rate (HR): 85 bpm; Systolic blood pressure (Psys): 130 mm Hg, Diastolic blood pressure (Pdia): 70 mm Hg; Parameter preprocessing: Calculation of mean arterial pressure .

[0044] Step S2: Construct a personalized blood rheology model (Viscosity Modeling). Traditional QFR assumes a blood viscosity of [missing value]. In this embodiment, the effective individualized whole blood viscosity of the patient is dynamically calculated based on the Hct value obtained in step S1. The calculation formula uses the modified Einstein viscosity equation: Among them, plasma viscosity Set the baseline value as ; Calculation result: Substituting Hct=32%, the calculation yields... Note: Compared to the general standard value (3.5), the patient's viscosity is reduced by about 18%. If not corrected, the calculated pressure drop value will be too high.

[0045] Step S3: 3D Reconstruction and Geometric Parameter Extraction. The epipolar constraint algorithm is used to perform 3D reconstruction on the two angiographic images, generating a 3D geometric model of the coronary arteries and extracting key geometric parameters. The length of the lesion segment L = 15 mm; Proximal reference diameter Diameter at the narrowest point (Narrowness approximately 56%).

[0046] Step S4: Adaptive Resistance Calculation. The system abandons fixed assumptions about microcirculatory resistance and adjusts boundary conditions according to the patient's cardiac function to calculate the actual microcirculatory resistance. Calculate reference resistance ( ): Based on the total cross-sectional area of ​​the terminal vessels of the coronary tree, the resting reference resistance is calculated using the scaling law; Calculate the myocardial oxygen consumption index (RPP): System settings are normal (reference). The patient's number was 9000. Elevated levels indicate increased myocardial oxygen consumption and a state of mild metabolic dilation in microcirculation. Determine the final boundary resistance ( The drag correction coefficient is calculated using the drag adjustment function. (i.e., resistance reduced by 8%); final actual boundary resistance .

[0047] Steps S5-S6: Fluid dynamics solution and QFR calculation, applying the individualized viscosity from step S2 ( ) and the corrected resistance in step S4 ( Input the fast pressure drop equation based on the Bernoulli and Poiseuille principles: Calculation process: The solver uses a time step of 0.01s to simulate pressure changes during the cardiac cycle; Output results: Calculated mean pressure at the distal end of the stenosis. And finally output the individualized QFR value.

[0048] 3. Comparative Verification of Examples (Case Study) To verify the effectiveness and clinical superiority of this embodiment, a comparative test of three assessment methods was conducted on the aforementioned patient with anemia and hyperdynamic status:

[0049] Verification conclusion: Traditional QFR failed to identify patients' anemia and hyperdynamic state, overestimating blood viscosity and resistance, leading to an inflated frictional pressure drop calculation and false positives (error as high as 0.08). In contrast, the individualized QFR provided by this invention, through dynamic correction of multidimensional physiological parameters, achieves calculation results highly consistent with the gold standard FFR (error only 0.01), accurately correcting misjudgments and effectively avoiding unnecessary stent implantation surgery for patients.

[0050] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters, characterized in that, Includes the following steps: S1. Collect multidimensional real-time physiological parameters of the patient, including at least cardiac function indicators and hemorheological data; S2. Based on the hematocrit in the blood rheology data, the individualized whole blood viscosity is dynamically calculated by modifying the Einstein equation, and the Valeo-Lingquist effect is introduced to correct the microvascular viscosity in the microvessels. S3. Reconstruct a three-dimensional geometric model of the coronary arteries using coronary angiography images, and extract the geometric features of the blood vessels from the three-dimensional geometric model of the coronary arteries; S4. Based on the geometric features of the blood vessels, the reference microcirculation resistance of the coronary artery tree is calculated using the scaling law, and the reference microcirculation resistance is dynamically corrected according to the cardiac function indicators to obtain the individualized actual microcirculation resistance. S5. Substitute the individualized whole blood viscosity or corrected microvascular viscosity, and the individualized actual microcirculation resistance as boundary conditions into the rapid pressure drop equation, couple the frictional pressure drop along the flow path with the local expansion / separation pressure drop, and output the pressure distribution and flow velocity distribution along the vessel centerline. S6. Based on the pressure distribution, extract the pressure at the distal end of the stenotic lesion and the average pressure at the coronary ostium, calculate the individualized QFR value, and visualize the output on the three-dimensional geometric model of the coronary artery.

2. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 1, characterized in that, S1 specifically includes: The patient's real-time hemodynamic parameters, including heart rate, systolic blood pressure, diastolic blood pressure, cardiac output, ejection fraction, and body surface area, are automatically acquired through a clinical information system or bedside monitoring equipment to form a cardiac function vector as the cardiac function index. Simultaneously extract blood rheological parameters, including hematocrit, hemoglobin concentration, and total whole blood protein, to construct blood state characteristics; The cardiac function vector and the blood state features are fused and normalized to construct individualized blood rheology indicators, and integrated to generate an individualized multidimensional physiological parameter dataset.

3. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 1, characterized in that, S2 specifically includes: Based on the collected hematocrit values, the individualized whole blood viscosity was calculated using the modified Einstein viscosity equation. For microcirculatory vascular segments, the system automatically identifies whether a segment belongs to the microvascular range based on the vessel diameter and dynamically adjusts the blood viscosity for segments belonging to the microvascular range. The Valeo-Lingquist effect is introduced to perform a secondary correction on the viscosity of microvascular segments with diameters smaller than a set threshold, resulting in the microvascular viscosity.

4. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 3, characterized in that, S2 further includes: The blood vessel diameter threshold is set to 0.3 mm, and the system automatically identifies blood vessel segments in the coronary artery tree with a diameter smaller than this threshold as microcirculation zones. Based on the relative relationship between blood vessel diameter and red blood cell size, a viscosity decay function is used to gradually decay the individualized whole blood viscosity, and the corrected viscosity value is used as the rheological boundary condition of the microcirculation region.

5. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 1, characterized in that, S3 specifically includes: Acquire two coronary angiography images with a projection angle difference of not less than 25 degrees; The two coronary angiography images were 3D registered and reconstructed using epipolar constraint and edge detection algorithms to generate the 3D geometric model of the coronary artery and the surface mesh of the lumen. Geometric features are extracted from the three-dimensional geometric model of the coronary artery, including the coordinates of the vessel centerline, the diameter of each segment, the length of the segment, and the stenotic lesion segment.

6. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 1, characterized in that, S4 specifically includes: Based on the reconstructed three-dimensional geometric model of the coronary arteries, the reference microcirculation resistance of the coronary artery tree is calculated using the scaling law based on the total cross-sectional area of ​​the terminal vessels; The myocardial oxygen consumption index is calculated based on the product of the patient's real-time heart rate and systolic blood pressure, and the cardiac function status is assessed based on the cardiac output index. Based on the myocardial oxygen consumption index and the cardiac function status, the reference microcirculation resistance is dynamically weighted and corrected using a predefined resistance adjustment function to obtain the individualized actual microcirculation resistance.

7. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 6, characterized in that, S4 further includes: A function is constructed to establish an inverse relationship between the myocardial oxygen consumption index and microcirculation resistance, and the resistance coefficient is automatically reduced when myocardial oxygen consumption increases. A cardiac function status classification mechanism is introduced to automatically increase the microcirculation resistance correction weight for patients with heart failure or low cardiac output.

8. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 1, characterized in that, S5 specifically includes: The individualized blood viscosity, the corrected microvascular viscosity, and the individualized actual microcirculation resistance are used as boundary conditions and input into a fluid dynamics solver based on the fast pressure drop equation. In the time domain, with the cardiac cycle as the time step, the total pressure drop is calculated by coupling the frictional pressure drop along the path with the local expansion / separation pressure drop, and the blood flow and pressure distribution are dynamically simulated. The instantaneous pressure and flow velocity data at each point along the centerline of the blood vessel are output to form a hemodynamic distribution map within the cardiac cycle.

9. The method for dynamic calculation of coronary artery QFR integrating individualized multidimensional physiological parameters according to claim 1, characterized in that, S6 specifically includes: The pressure distal to the stenotic lesion and the average pressure at the coronary ostium are extracted from the pressure distribution results, and the ratio of distal pressure to ostial pressure is calculated as the individualized QFR value. Individualized QFR values ​​are overlaid on the surface of the coronary artery 3D geometric model using a color mapping method, and numerical annotations are made at key anatomical locations to form a visual assessment report.

10. A coronary artery QFR dynamic calculation system integrating individualized multidimensional physiological parameters, used to implement the coronary artery QFR dynamic calculation method integrating individualized multidimensional physiological parameters as described in any one of claims 1-9, characterized in that, include: The data acquisition and preprocessing module is used to automatically connect to the clinical information system and monitoring equipment, acquire and integrate patient hemodynamic data, cardiac function indicators and blood rheology data in real time, and complete data alignment, outlier removal and normalization. The individualized blood viscosity modeling module calculates individualized whole blood viscosity based on the hematocrit of erythrocytes in the blood rheology data using a modified fluid dynamics model, and dynamically adjusts and corrects microvascular viscosity according to the vessel diameter. The 3D vessel reconstruction and geometric analysis module, based on coronary angiography images, reconstructs the 3D geometric model of the coronary arteries through epipolar constraint and edge detection algorithms, and automatically extracts the geometric features of the vessels; An adaptive microcirculation resistance construction module is used to dynamically correct the reference microcirculation resistance by combining the geometric features of the blood vessels with the cardiac function indicators to obtain individualized actual microcirculation resistance. The QFR calculation and hemodynamic solution module is used to use the individualized whole blood viscosity or microvascular viscosity and the individualized actual microcirculation resistance as boundary conditions, solve the transient blood flow and pressure distribution based on the fast pressure drop equation, calculate the individualized QFR value and output a three-dimensional visualization.

Images