A personalized control system for left ventricular assist device

Abstract

The present invention provides a control system and method for personalized treatment of a left ventricular assist device, relating to the field of artificial heart technology. The method comprises: for each cardiac cycle, a Starling-like regulator receives a left atrial pressure signal fed back by a physiological feedback module, and outputs a desired cardiac index based on the patient's body surface area; a pump flow quantification feedback module calculates the pump flow in real time based on the rotational speed and current of the left ventricular assist device, and outputs the feedback cardiac index in combination with the patient's body surface area; calculates the deviation between the desired cardiac index and the feedback cardiac index, inputs the deviation into a controller, and the controller determines the output current increment based on the deviation; and repeats the above process until the control system stabilizes. Ventricular pressure is monitored by the physiological feedback module and the cardiovascular coupling module, and personalized physiological characteristics analysis is performed in combination with the patient's body surface area, so that the pump speed parameters can be automatically optimized according to the patient's real-time physiological state.

Description

Technical Field

[0001] The present application relates to the field of artificial heart technology, and more specifically, to a control system and method for personalized treatment of a left ventricular assist device. Background Art

[0002] Left ventricular assist devices (LVADs) pump blood from the left ventricle into the aorta, thereby relieving the left ventricle's workload and providing stable blood circulation. They play a vital role in the treatment of end-stage heart failure. The development of LVADs has undergone three major technological innovations. The first-generation LVADs used pulsatile flow technology, designed to mimic the beating pattern of the natural heart. However, their complex mechanical structure resulted in a bulky device. The second-generation LVADs innovatively employed mechanical bearing technology, significantly reducing the device's size and improving its mechanical durability. However, this also resulted in mechanical wear and heat generation. The latest generation of LVADs utilizes suspension technology, completely eliminating mechanical contact parts. This breakthrough has reduced the risk of thrombosis and improved blood compatibility, while further reducing mechanical wear. These technological advancements have continuously improved LVADs in terms of miniaturization, reliability, and biocompatibility, providing safer and more effective treatment options for patients with end-stage heart failure. Despite significant progress in LVAD design, existing LVAD control systems often employ fixed speeds or simple feedback adjustments based on a single parameter, which present significant limitations. Typical LVAD control solutions in existing technologies include:

[0003] (1) The Chinese invention patent with publication number CN109793954A discloses a non-differentiated adaptive physiological control method based on a left ventricular assist device (LVAD). First, the preload from the left ventricle is measured and used as the input value of the Frank-Starling-like control system to generate the reference value of the average pump flow and the reference value of the pump flow pulsation value required by the patient. In the signal processing module, a Savitzky-Golay (SG) filter, an extended Kalman filter and a low-pass filter are used to remove noise from the extracted pump speed signal, and at the same time, the required average pump flow measurement value and the flow pulsation value measurement value (Qplus) are estimated. Finally, the signal is input into the PI controller for processing, and the output current signal directly controls the LVAD to achieve the desired pump speed. This solution lacks the ability to adapt to individual differences of patients (such as physiological structure, pathological state) and dynamic physiological state (such as exercise, sleep), and is prone to complications such as ventricular insufficiency or insufficient perfusion.

[0004] (2) The Chinese invention patent with publication number CN115016256A discloses a physiologically adaptive MPC control method for a left ventricular assist device. The method obtains the pressure measurement value LVP from the left ventricle and uses it as the input value of the aortic pressure neural network model. The aortic pressure reference value AOPref required by the patient is obtained through the aortic pressure neural network model. The LVAD motor speed Ps is obtained, and the pump flow Qp is estimated through low-pass filter noise removal and power-pump flow relationship. The safe pump flow Qs is obtained after suction reflux monitoring. The patient's aortic pressure AOP is obtained and AOPref is used as the reference value. LVP, AOP, and Qs are input into the MPC controller as system states for processing. The output current signal I controls the LVAD to achieve a speed that meets the patient's physiological needs. This control strategy has difficulty in quickly responding to sudden changes in physiological parameters and lacks an intelligent decision-making mechanism for multi-parameter fusion.

[0005] To address these issues, there is an urgent need to develop a control system that can integrate multi-dimensional physiological data in real time, dynamically optimize pump speed using intelligent algorithms, and support personalized treatment of left ventricular assist devices to improve clinical safety and treatment adaptability. Summary of the Invention

[0006] To address the above-mentioned issues, the technical solution adopted in this application is a control method for personalized treatment of a left ventricular assist device: for each cardiac cycle, a Starling-like regulator receives the left atrial pressure signal fed back by a physiological feedback module and outputs a desired cardiac index based on the patient's body surface area; a pump flow quantification feedback module calculates the pump flow in real time based on the rotational speed and current of the left ventricular assist device and outputs a feedback cardiac index based on the patient's body surface area; the deviation between the desired cardiac index and the feedback cardiac index is calculated and input into a controller, which determines the output current increment based on the deviation;

[0007] The above process is repeated in each subsequent cardiac cycle until the control system is stable.

[0008] Optionally, the desired cardiac index is calculated according to the following formula:

[0009] ;

[0010] ;

[0011] Where, is the expected cardiac index, is the desired cardiac output, BSA is the body surface area, For height, For weight.

[0012] Optionally, desired cardiac output is calculated according to the following formula:

[0013] ;

[0014] Where, is the desired cardiac output, is the left atrial pressure, The contractility of the simulated myocardium changes the sensitivity of the desired pump flow to left atrial pressure for the scaling factor.

[0015] Optionally, the scaling factor is calculated according to the following formula:

[0016] ;

[0017] Where, represents the myocardial contractility coefficient, represents the ventricular systolic time coefficient, represents the baseline time reference, Indicates heart rate.

[0018] Optionally, the ventricular systolic time coefficient and baseline time reference and cardiac cycle , left ventricular diastolic time and left ventricular systolic time Has the following relationship:

[0019] ;

[0020] ;

[0021] = - .

[0022] Optionally, the controller is designed based on a fuzzy control algorithm, the controller input is the deviation between the desired cardiac index and the fed-back cardiac index and the rate of change of the deviation, and the controller output is the current increment of the left ventricular assist device rotary pump.

[0023] Optionally, the calculation of the current increment includes:

[0024] S1: fuzzy interface;

[0025] The fuzzy subsets of the deviation are divided into:

[0026] Deviation = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB};

[0027] The fuzzy subsets of the deviation change rate are divided into:

[0028] Deviation change rate = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB};

[0029] The fuzzy subset division of phase current is:

[0030] Phase current = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB};

[0031] The domain of deviation and deviation change rate is {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}, and the membership functions are all triangular membership functions;

[0032] S2: Fuzzy rule table for formulating current increment based on deviation and deviation change rate;

[0033] S3: Defuzzification, using the centroid method to calculate the quantized value of the current increment according to the following formula:

[0034] ;

[0035] Where, represents the current increment after defuzzification, Indicates the phase current fuzzy quantity corresponding to the current increment, represents the membership value corresponding to the fuzzy quantity, is the total number of outputs after discretization.

[0036] Optionally, the calculation of the fed-back cardiac index includes:

[0037] Pump flow estimation: The expression for pump flow estimation is as follows:

[0038] ;

[0039] Where, is the estimated pump flow rate, is the rotor inertia, is the rotation speed, is the back electromotive force constant, is the current, is the damping coefficient, 、 is the correlation coefficient;

[0040] Pump flow feature extraction: Pump flow feature extraction refers to extracting the average pump flow from the estimated pump flow data set. The calculation formula for the average pump flow is as follows:

[0041] ;

[0042] Where, is the average pump flow rate, is the total number of samples in one cardiac cycle, j is an index variable, which is used to indicate the sampling point number in one cardiac cycle;

[0043] Feedback cardiac index: The feedback cardiac index is calculated as follows:

[0044] ;

[0045] Where, The cardiac index is fed back.

[0046] Optionally, the left atrial pressure signal fed back by the physiological feedback module is the average left atrial pressure extracted by the physiological feedback module from the cardiovascular coupling system, and is calculated according to the following formula:

[0047] ;

[0048] Where, is the mean left atrial pressure, is the total number of samples in one cardiac cycle, and j is an index variable used to indicate the sampling point number in one cardiac cycle.

[0049] The present application also provides a control system for personalized treatment of a left ventricular assist device, comprising a left ventricular assist device, a cardiovascular coupling system, and a physiological feedback module, and is applicable to any of the aforementioned control methods for personalized treatment of a left ventricular assist device, including:

[0050] Starling-like regulator: used to receive the left atrial pressure signal fed back by the physiological feedback module and output the desired cardiac index based on the patient's body surface area;

[0051] Pump flow quantification feedback module: used to calculate the pump flow in real time based on the rotational speed and current of the left ventricular assist device, and output the feedback cardiac index based on the patient's body surface area;

[0052] Controller: used to determine the output current increment according to the input deviation;

[0053] Physiological feedback module: used to receive the left atrial pressure signal of the cardiovascular coupling system and feed it back to the Starling-like regulator.

[0054] The beneficial effects of the control system and method for personalized treatment of left ventricular assist devices provided in this application are:

[0055] The physiological feedback module uses the left atrial pressure signal to output a desired cardiac index based on the patient's body surface area. The pump flow rate is calculated in real time based on the LVAD's rotary pump speed and current. The feedback cardiac index is then output in conjunction with the patient's body surface area, and the output current increment is determined based on any input deviation. Atrial pressure monitoring via the physiological feedback module and cardiovascular coupling module, combined with personalized physiological analysis of the patient's body surface area, enables precise and adaptive circulatory support. This innovative design enables the LVAD to automatically optimize pump speed parameters based on the patient's real-time physiological state (such as exercise / rest transitions and changes in fluid balance), ensuring adequate cardiac output while avoiding ventricular overloading. This provides personalized blood flow support that better meets physiological needs while significantly reducing the risk of complications such as right heart failure, thromboembolism, and gastrointestinal bleeding, common with traditional LVAD therapy. This provides a safer and more effective mechanical circulatory support solution for patients with advanced heart failure. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art.

[0057] Figure 1 This is a block diagram of a control system for personalized treatment of a left ventricular assist device provided in an embodiment of the present application;

[0058] Figure 2 This is an operation flow chart of the control system provided in the embodiment of the present application;

[0059] Figure 3 This is a graph of current changes during the transition from rest to exercise in different patients provided in the embodiments of the present application;

[0060] Figure 4 This is a graph of pump speed changes during the transition from rest to exercise for different patients provided in the embodiments of the present application;

[0061] Figure 5 This is a graph of cardiac output changes during the transition from rest to exercise in different patients provided in the embodiments of the present application. DETAILED DESCRIPTION

[0062] In order to make the technical problems, technical solutions and beneficial effects to be solved by this application more clearly understood, this application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0063] Example 1

[0064] like Figure 1As shown, the present application provides a control system for personalized treatment of a left ventricular assist device, including a left ventricular assist device, a cardiovascular coupling system and a physiological feedback module, including:

[0065] Starling-like regulator: used to receive the left atrial pressure signal fed back by the physiological feedback module and output the desired cardiac index based on the patient's body surface area;

[0066] Pump flow quantification feedback module: used to calculate the pump flow in real time based on the rotational speed and current of the left ventricular assist device, and output the feedback cardiac index based on the patient's body surface area;

[0067] Controller: used to determine the output current increment according to the input deviation;

[0068] Physiological feedback module: used to receive the left atrial pressure signal of the cardiovascular coupling system and feed it back to the Starling-like regulator.

[0069] The Starling-like regulator mimics the heart's inherent Starling mechanism. The construction process includes the following steps:

[0070] Step 1: Design a Starling-like controller to describe the relationship between the desired cardiac output and left atrial pressure.

[0071] Step 2: Quantify expected cardiac output in combination with individual characteristic parameters as an indicator for individualized assessment and clinical application.

[0072] The desired cardiac output is related to left atrial pressure as follows:

[0073]

[0074] Where, is the desired cardiac output, is the left atrial pressure, is a scaling factor that models the sensitivity of myocardial contractility to altering desired pump flow to left atrial pressure.

[0075] The scaling factor is calculated according to the following formula:

[0076] ;

[0077] Where, represents the myocardial contractility coefficient, represents the ventricular systolic time coefficient, represents the baseline time reference, Indicates heart rate.

[0078] ventricular systolic time coefficient and baseline time reference and cardiac cycle , left ventricular diastolic time and left ventricular systolic time Has the following relationship:

[0079] ;

[0080] ;

[0081] = - ;

[0082] In this embodiment, the ventricular systolic time coefficient , baseline time reference and myocardial contractility The determination is based on the following conditions:

[0083] (1) In the resting state, the ratio of ventricular systole to diastole time is ;

[0084] (2) For healthy adults, At 75bpm, is 1, corresponding to a cardiac output of 5 L / min;

[0085] (3) According to the necessary and sufficient conditions for extreme values, when hour, Reaching a maximum value. When the normal heart rate is 180bpm, Start to decline, that is .

[0086] According to the above conditions, , , .

[0087] Myocardial contractility is related to activity intensity, and changes in heart rate directly reflect activity level. Therefore, a scaling factor was established and heart rate The functional relationship between them is as follows:

[0088]

[0089] The above shows the calculation and value under standard conditions. For each individual, the following adjustments are made:

[0090] For condition (1), the heart rate was monitored and the ratio of ventricular systole to diastole time was collected;

[0091] For condition (2), the reference value of the heart rate and the corresponding cardiac output in the resting state are determined according to the corresponding age group and gender, and then the reference value of the scaling factor is formulated;

[0092] For condition (3), its heart rate maximum is The basis for starting to decline, and then determining The value of .

[0093] Based on the above method, the present application provides a method for calculating the scaling factor, and the calculation of the scaling factor does not require invasive monitoring, which provides convenience for personalized treatment.

[0094] The individual characteristic parameters are body surface area, height, and weight, and the indicator for individualized assessment and clinical application is cardiac index.

[0095] The relationship between body surface area, height, and weight is as follows:

[0096] ;

[0097] Where, BSA is the body surface area (cubic meters), is height (cm), is the weight in kilograms.

[0098] The expected value of cardiac index is calculated as follows:

[0099] ;

[0100] Where, is the expected cardiac index.

[0101] The controller is designed based on an intelligent control algorithm. The input of the controller is the deviation between the desired cardiac index and the feedback cardiac index and the rate of change of the deviation. The output of the controller is the current increment of the left ventricular assist device rotary pump, such as Figure 1 As shown, the expected cardiac index and the fed-back cardiac index are first input into the discriminator, and after calculation by the discriminator, the deviation between the expected cardiac index and the fed-back cardiac index and the rate of change of the deviation are obtained.

[0102] The controller is a fuzzy controller.

[0103] The design steps of the fuzzy controller are as follows:

[0104] Step 1: Fuzzy the interface.

[0105] The fuzzy subsets of the deviation are divided into:

[0106] Deviation = {Negative Large, Negative Medium, Negative Small, Zero Negative, Zero Positive, Positive Small, Positive Medium, Positive Large} = {NB, NM, NS, NZ, PZ, PS, PM, PB}

[0107] The fuzzy subsets of the deviation change rate are divided into:

[0108] Deviation change rate = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB}

[0109] The fuzzy subset division of the motor phase current increment is:

[0110] Phase current = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB}

[0111] The domain of deviation and deviation change rate is {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}, and the membership functions are all triangular membership functions.

[0112] Step 2: Design the fuzzy rule table.

[0113] The fuzzy rules are shown in Table 1.

[0114] Table 1 Fuzzy rule table

[0115]

[0116] Step 3: Defuzzification.

[0117] The center of gravity method is used to calculate the quantitative value of the current increment. The formula is as follows:

[0118] ;

[0119] Where, represents the current increment after defuzzification, represents the fuzzy quantity corresponding to the current increment, represents the membership value corresponding to the fuzzy quantity, is the total number of outputs after discretization.

[0120] The core component of the left ventricular assist device is the rotary pump.

[0121] The model of the rotary pump includes a mechanical dynamics model and a hydraulic model.

[0122] The mechanical dynamics model is expressed as follows:

[0123] ;

[0124] Where, is the rotor inertia, is the rotation speed, is the back electromotive force constant, is the current, is the damping coefficient, 、 is the correlation coefficient, is the pump flow rate.

[0125] The hydraulic model is expressed as follows:

[0126] ;

[0127] Where, is the pump pressure difference, 、 、 is the experimental constant.

[0128] The cardiovascular coupling system is an existing technology used to simulate the hemodynamic characteristics of the cardiovascular system when the left ventricular assist device is in operation, including a cardiovascular lumped parameter model and a pressure reflex model.

[0129] For reference, the doctoral dissertation of Shandong University, "Research on Pulsation Speed ​​Variation and Physiological Control of Continuous Flow Left Ventricular Assist Device" ([1] Liu Hongtao. Research on Pulsation Speed ​​Variation and Physiological Control of Continuous Flow Left Ventricular Assist Device [D]. Shandong University, 2021. DOI: 10.27272 / d.cnki.gshdu.2021.006094.), Chapter 2 Cardiovascular Coupled System Hemodynamic Model provides a detailed modeling method for the hemodynamic model of the cardiovascular coupled system.

[0130] The cardiovascular coupling system referred to in this application is used to provide simulated feedback in in vitro experiments.

[0131] The cardiovascular lumped parameter model is used to simulate the hemodynamics of the cardiovascular system, dividing the blood circulation system into four parts: left heart, right heart, systemic circulation, and pulmonary circulation.

[0132] The baroreflex model is used to simulate the neural regulation of aortic pressure and regulate left ventricular end-systolic elastance, right ventricular end-systolic elastance, systemic peripheral resistance and heart rate in a cardiovascular lumped parameter model.

[0133] The quantitative feedback of pump flow includes pump flow estimation and pump flow feature extraction.

[0134] The expression for pump flow estimation is as follows:

[0135] ;

[0136] Where, is the estimated pump flow rate.

[0137] Pump flow feature extraction refers to extracting the average pump flow from the estimated pump flow data set.

[0138] The average pump flow rate is calculated as follows:

[0139] ;

[0140] Where, is the average pump flow rate, is the total number of samples collected during one cardiac cycle.

[0141] The feedback cardiac index calculation formula is as follows:

[0142] ;

[0143] Where, The cardiac index is fed back.

[0144] The left atrial pressure signal fed back by the physiological feedback module is the average left atrial pressure extracted from the cardiovascular coupling system by the physiological feedback module and is calculated according to the following formula:

[0145] ;

[0146] Where, is the mean left atrial pressure, is the total number of samples in one cardiac cycle, and j is an index variable used to indicate the sampling point number in one cardiac cycle.

[0147] Example 2:

[0148] Figure 2 The present application provides a method for executing the control system of the personalized treatment of the left ventricular assist device within one cardiac cycle. After receiving the patient's body surface area and the left atrial pressure signal fed back by the physiological feedback module, the Starling-like regulator outputs the desired cardiac index. At the same time, the fed-back pump flow is quantified as the fed-back cardiac index through the body surface area. The deviation between the desired cardiac index and the fed-back cardiac index is then calculated, and the deviation is input into the controller. The controller determines the positive or negative output current increment based on the input deviation. If the current increment is positive, the pump speed is high, otherwise the pump speed is reduced. As the pump speed increases, the pump flow increases, otherwise the pump flow decreases. The increase or decrease in pump flow will inevitably cause changes in the hemodynamics of the cardiovascular coupling system, especially the left atrial pressure. The above process is repeated in the next cardiac cycle until the control system stabilizes.

[0149] Example 3:

[0150] Figure 3-Figure 5 The dynamic changes in current, pump speed, and cardiac output during the transition from rest to exercise for patients with different body surface areas (1.64 m2 and 2.23 m2) are detailed. During the initial rest phase (before the 20th second), the system automatically adjusts support levels based on the patient's body surface area: the patient with a smaller body surface area (1.64 m2) receives a baseline cardiac output of 5 L / min, while the patient with a larger body surface area (2.23 m2) receives 7 L / min. This difference in initial settings demonstrates the control system's ability to accurately identify individual patients' physiological needs.

[0151] Starting at the 20th second, as the patient entered the exercise transition phase, the control system demonstrated excellent dynamic regulation. Unlike the fixed speed mode of traditional mechanical pumps, this personalized control system employs a progressive regulation strategy, first gradually increasing the drive current, which in turn drives a gradual increase in pump speed, ultimately achieving an adaptive increase in cardiac output. This regulation process not only avoids drastic hemodynamic fluctuations but also ensures a smooth physiological transition, demonstrating the optimized design of the control algorithm. After approximately 30 seconds of buffering adjustment, the system reached a new steady-state during exercise. At this point, cardiac output increased to 8 L / min for the 1.64 m2 patient and 10.5 L / min for the 2.23 m2 patient. Notably, despite a 36% difference in body surface area between the two patients, the increase in cardiac output during exercise exhibited distinct characteristics (a 60% increase for the 1.64 m2 patient and a 50% increase for the 2.23 m2 patient). This demonstrates that the control system is not simply adjusting proportionally but rather optimizes based on a more complex physiological demand model.

[0152] These results have important clinical significance. First, it verifies that the personalized control algorithm can achieve precise matching of the initial settings based on key parameters such as the patient's body surface area. Second, the smooth transition characteristics exhibited by the system during state transitions can effectively reduce the clinical risks brought about by hemodynamic mutations. In addition, the differentiated improvement in cardiac output during exercise further confirms that the control system can dynamically adapt to the changes in metabolic needs of different patients. These findings provide an important reference for the research and development of a new generation of intelligent ventricular assist devices, indicating that the adaptive regulation system based on physiological feedback can not only provide circulatory support that is more in line with individual needs, but also maintain optimal hemodynamic stability when the patient's daily activity status changes, thereby providing a new technical solution for improving the quality of life and exercise tolerance of patients with heart failure.

[0153] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A personalized control system for a left ventricular assist device, comprising a left ventricular assist device, a cardiovascular coupling system, and a physiological feedback module, characterized in that: include: A Starling-like regulator is configured to receive the left atrial pressure signal fed back by the physiological feedback module and output a desired cardiac index based on the patient's body surface area; Pump flow quantification feedback module: used to calculate the pump flow in real time based on the rotational speed and current of the left ventricular assist device, and output the feedback cardiac index based on the patient's body surface area; Controller: used to determine the output current increment according to the input deviation; Physiological feedback module: used to receive the left atrial pressure signal of the cardiovascular coupling system and feed it back to the Starling-like regulator; The control system achieves the following control: for each cardiac cycle, the Starling-like regulator receives the left atrial pressure signal fed back by the physiological feedback module and outputs a desired cardiac index based on the patient's body surface area; The pump flow quantification feedback module calculates the pump flow in real time based on the rotational speed and current of the left ventricular assist device, and outputs a feedback cardiac index based on the patient's body surface area; calculates the deviation between the expected cardiac index and the feedback cardiac index, and inputs the deviation into the controller, which determines the output current increment based on the deviation; In each subsequent cardiac cycle, the above process is repeated until the control system is stable; The expected cardiac index is calculated according to the following formula: ； ； Where, is the expected cardiac index, is the desired cardiac output, BSA is the body surface area, For height, is weight; The desired cardiac output is calculated according to the following formula: ； Where, is the desired cardiac output, is the left atrial pressure, is the scaling factor, simulating the sensitivity of the myocardial contractility to change the desired pump flow to the left atrial pressure; The scaling factor is calculated according to the following formula: ； Where, represents the myocardial contractility coefficient, represents the ventricular systolic time coefficient, represents the baseline time reference, Indicates heart rate.

2. The personalized control system for a left ventricular assist device according to claim 1, characterized in that: The ventricular systolic time coefficient and baseline time reference and cardiac cycle , left ventricular diastolic time and left ventricular systolic time Has the following relationship: ； ； = - 。 3. The personalized control system for a left ventricular assist device according to claim 1, characterized in that: The controller is designed based on a fuzzy control algorithm. The controller input is the deviation between the desired cardiac index and the fed-back cardiac index and the rate of change of the deviation. The controller output is the current increment of the rotary pump of the left ventricular assist device.

4. The personalized control system for a left ventricular assist device according to claim 3, characterized in that: The calculation of the current increment includes: S1: fuzzy interface; The fuzzy subsets of the deviation are divided into: Deviation = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB}; The fuzzy subsets of the deviation change rate are divided into: Deviation change rate = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB}; The fuzzy subset division of phase current is: Phase current = {negative large, negative medium, negative small, zero negative, zero positive, positive small, positive medium, positive large} = {NB, NM, NS, NZ, PZ, PS, PM, PB}; The domain of the deviation and the rate of change of the deviation is {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}, and the membership functions are all triangular membership functions; S2: Fuzzy rule table for formulating current increment based on deviation and deviation change rate; S3: Defuzzification, using the centroid method to calculate the quantized value of the current increment according to the following formula: ； Where, represents the current increment after defuzzification, Indicates the phase current fuzzy quantity corresponding to the current increment, represents the membership value corresponding to the fuzzy quantity, is the total number of outputs after discretization.

5. The personalized control system for a left ventricular assist device according to claim 3, characterized in that: The calculation of the feedback cardiac index includes: Pump flow estimation: The expression for the pump flow estimation is as follows: ； Where, is the estimated pump flow rate, is the rotor inertia, is the rotation speed, is the back electromotive force constant, is the current, is the damping coefficient, 、 is the correlation coefficient; Pump flow feature extraction: The pump flow feature extraction refers to extracting the average pump flow from the estimated pump flow data set. The calculation formula of the average pump flow is as follows: ； Where, is the average pump flow rate, is the total number of samples in one cardiac cycle, j is an index variable, which is used to indicate the sampling point number in one cardiac cycle; Feedback cardiac index: The feedback cardiac index is calculated as follows: ； Where, The cardiac index is fed back.

6. The personalized control system for a left ventricular assist device according to claim 1, characterized in that: The left atrial pressure signal fed back by the physiological feedback module is the average left atrial pressure extracted from the cardiovascular coupling system by the physiological feedback module, and is calculated according to the following formula: ； Where, is the mean left atrial pressure, is the total number of samples in one cardiac cycle, and j is an index variable used to indicate the sampling point number in one cardiac cycle.

Images