Simulation cycle test system for extracorporeal membrane oxygenation ECMO

Abstract

The invention discloses a simulation cycle test system for extracorporeal membrane oxygenation ECMO, and relates to the technical field of medical instrument extracorporeal test. The system comprises a training data acquisition module for acquiring blood flow characteristic parameters of users with different body characteristic parameters and generating a training sample data set. The prediction model training module establishes a neural network model and trains the blood flow feature extraction model. The physiological feature analysis module predicts blood flow feature parameters by simulating body feature parameters of a user and sets blood exchange flow velocity and pressure according to predicted values. The exchange parameter adjusting module is used for dynamically adjusting the set flow velocity and pressure and acquiring blood quality parameters under the operation condition. The blood quality adjusting module is used for correcting the gas flow of the membrane oxygenator according to the obtained blood quality parameters, the ECMO simulation cycle test is completed, refined gas flow management is achieved, and the oxygenation efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of extracorporeal testing of medical devices, and specifically to a simulated circulation testing system for extracorporeal membrane oxygenation (ECMO). Background Art

[0002] As a means of intensive care treatment, extracorporeal membrane oxygenation (ECMO) is mainly used to support the respiratory and circulatory functions of critically ill patients. When the respiratory and cardiac functions of a patient are severely impaired, ECMO can effectively perform gas exchange from outside the body to maintain life. However, the effectiveness of ECMO depends to a large extent on the real-time monitoring and adjustment of various physiological parameters, including blood flow rate, blood flow pressure, partial pressure of oxygen, partial pressure of carbon dioxide, and blood oxygen saturation, etc. These parameters are affected by individual patient characteristics such as age, weight, basal metabolic rate, and exercise level, resulting in different physiological responses of different patients under the same ECMO settings.

[0003] Currently, the ECMO settings for different patients in clinical practice often rely on the experience of medical staff and some basic physiological models, which may lead to inaccurate parameter settings to a certain extent, thus affecting the use effect. In addition, most of the existing ECMO monitoring systems are statically set and lack the function of adaptive adjustment, failing to respond in a timely manner to changes in the patient's condition, resulting in limitations in use. However, there is currently a lack of corresponding detection systems and evaluation indicators, and it is impossible to evaluate the feasibility, stability, and safety of ECMO during the design, research, and experimental processes, which is not conducive to the comprehensive and objective evaluation of the product. Therefore, developing an intelligent system that can real-time monitor and predict the blood flow characteristics of patients and accurately set the flow rate and pressure according to individual characteristics is an urgent need for the current development of ECMO technology.

[0004] In the prior art, the authorized announcement number CN113440674B discloses a simulation circulation test system for extracorporeal membrane oxygenation (ECMO) and its uses. Specifically, it discloses a simulation circulation test system for extracorporeal membrane oxygenation (ECMO), including a pulmonary circulation simulation module, a flexible heart module, a systemic circulation simulation module, a detection ECMO module, and a control and detection module. The pulmonary circulation simulation module, the flexible heart module, and the systemic circulation simulation module are used to simulate the hemodynamic states of the human body under healthy and different pathological conditions. The control and detection module is used to collect data such as blood pressure, blood flow, and blood oxygen saturation at various points in the test system, and control the opening amplitude of each valve in the test system to simulate various cardiovascular and valve diseases. However, although the system in this solution designs pulmonary circulation, flexible heart, and systemic circulation modules, when simulating complex physiological states, it may not be able to fully reproduce the physiological characteristics of the human body, such as heart rate, respiratory rate, and central venous pressure. This may lead to deviations between the test results and the actual clinical situation, thus reducing the accuracy and effectiveness of the simulation system.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a simulation circulation test system for extracorporeal membrane oxygenation (ECMO) to solve the problems raised in the above background art.

[0007] To achieve the above purpose, the present invention provides the following technical solutions:

[0008] A simulation circulation test system for extracorporeal membrane oxygenation (ECMO) specifically includes:

[0009] A training data acquisition module, which is used to collect blood flow characteristic parameters of several users with different body characteristic parameters as training sample parameters, map the training sample parameters one by one with the corresponding body characteristic parameters of the users, and generate a training sample data set;

[0010] A prediction model training module, which is used to establish a neural network model, use the body characteristic parameter data of the training sample data set as the input of the neural network model, and the corresponding training sample parameter data as labels, and train the neural network model to obtain a blood flow characteristic extraction model;

[0011] A physiological feature analysis module is used to collect the body feature parameters of a simulated user, input them into a blood flow feature extraction model to obtain the predicted values of the blood flow feature parameters of the simulated user, and set the blood exchange flow rate and blood exchange pressure of an extracorporeal membrane oxygenation (ECMO) device according to the predicted values of the blood flow feature parameters. Then, use simulated sample blood to simulate in the set extracorporeal membrane oxygenation (ECMO) device to obtain simulated physiological feature parameters;

[0012] An exchange parameter adjustment module is used to adaptively adjust the set blood exchange flow rate and blood exchange pressure based on the obtained physiological feature parameters to obtain the corrected values of the dynamic blood exchange flow rate and the dynamic blood exchange pressure, and obtain the blood quality parameters when operating according to the corrected values;

[0013] A blood quality adjustment module is used to correct the gas flow rate of the membrane oxygenator by combining the blood quality parameters with the corrected value of the dynamic blood exchange flow rate and the corrected value of the dynamic blood exchange pressure to obtain the gas flow rate of the adaptive membrane oxygenator, and perform control according to the real-time gas flow rate of the adaptive membrane oxygenator, the corrected value of the dynamic blood exchange flow rate and the corrected value of the dynamic blood exchange pressure to complete the simulated circulation test of the extracorporeal membrane oxygenation (ECMO).

[0014] Furthermore, preprocess the collected blood flow feature parameters. The preprocessing is specifically normalization preprocessing. Similarly, the method for normalizing the blood flow pressure is the same. The generation method of the training sample data set is as follows: map the training sample parameters one by one to the body feature parameters of the corresponding user to form a corresponding grid, and record the formed grid as the training sample data set;

[0015] Wherein the blood flow feature parameters include blood flow rate and blood flow pressure, and the body feature parameters include age, weight, basal metabolic rate and exercise level.

[0016] Furthermore, based on the data of the training sample data set, establish a neural network model. Specifically, establish a neural network model through a long short-term memory network (LSTM) model. For the long short-term memory network (LSTM) model, select an activation function and an optimization algorithm. Select the Tanh function as the activation function and select Adam as the optimization algorithm for the LSTM model; The formula of the Tanh function is:

[0017]

[0018] In the formula, f(r) represents the Tanh function, and the independent variable r represents the input weighted sum of the neuron, that is, the result after the input received by the neuron from the previous layer is weighted and summed;

[0019] Meanwhile, set the hyperparameters of the LSTM model. The hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer;

[0020] Among them, the number of network layers is set to a 4-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32;

[0021] The input of the trained blood flow feature extraction model is the user's body feature parameter data including age, weight, basal metabolic rate, and exercise level, and the output is the corresponding predicted blood flow feature parameter values, including the predicted blood flow rate and blood flow pressure values. Among them, the exercise level specifically refers to the average daily exercise time of the user.

[0022] Furthermore, adaptively adjust the set blood exchange flow rate and blood exchange pressure based on the collected physiological feature parameters to obtain the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value. The physiological feature parameters include simulated heart rate, simulated respiratory rate, and simulated central venous pressure. Among them, the formula for calculating the dynamic blood exchange flow rate correction value is:

[0023]

[0024] In the formula, V' z is the dynamic blood exchange flow rate correction value, V for is the predicted blood flow rate value, HR S is the simulated heart rate data of the user, PR S is the simulated respiratory rate data of the user, CP S is the simulated central venous pressure data of the user, HR0 is the reference heart rate data, PR0 is the reference respiratory rate data, and CP0 is the reference central venous pressure data.

[0025] Furthermore, the formula for calculating the dynamic blood exchange pressure correction value is:

[0026]

[0027] In the formula, Pa' Z is the dynamic blood exchange pressure correction value, Pa for is the predicted blood flow pressure value.

[0028] Further, according to the obtained blood quality parameters, the gas flow rate of the membrane oxygenator is corrected by combining the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value to obtain the gas flow rate of the adaptive membrane oxygenator. The blood quality parameters include partial pressure of oxygen, partial pressure of carbon dioxide, and blood oxygen saturation. The specific formula for calculating the gas flow rate of the adaptive membrane oxygenator is as follows:

[0029]

[0030] In the formula, VA' z is the gas flow rate of the adaptive membrane oxygenator, VA0 is the initial gas flow rate of the membrane oxygenator, PCO2 is the partial pressure of carbon dioxide, is the reference value of the partial pressure of carbon dioxide, ZH P is the blood oxygen content characterization coefficient, ω1 is the weight coefficient of the sum of the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value, ω3 is the weight coefficient of the difference in the partial pressure of carbon dioxide, ω2 is the weight coefficient of the blood oxygen content characterization coefficient, where ω1>ω2>ω3 and ω1, ω2, and ω3 are all greater than 0, and ω1 + ω2 + ω3 = 1;

[0031] Among them, the blood oxygen content characterization coefficient ZH P is characterized by the partial pressure of oxygen and blood oxygen saturation. The specific formula for calculating the blood oxygen content characterization coefficient ZH P is as follows:

[0032]

[0033] In the formula, PO2 is the partial pressure of blood oxygen, SPO2 is the blood oxygen saturation, is the reference value of the partial pressure of blood oxygen, is the reference value of the blood oxygen saturation.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] By collecting the body characteristic parameters of the patient and using the neural network model for training, the present invention can generate high-precision predicted values of blood flow characteristic parameters, set the reference values for ECMO simulation, ensure the optimal blood flow rate and pressure, thereby improving the oxygenation effect and reducing the phenomenon of carbon dioxide retention.

[0036] Secondly, the adaptive adjustment ability of the dynamic blood exchange flow rate and pressure greatly improves the response speed and flexibility of the ECMO system. The system can simulate and collect physiological state information and automatically adjust the blood exchange parameters to test and enhance the adaptability of the ECMO system under changing conditions.

[0037] Finally, the system also demonstrates excellent performance in the optimized control of gas flow. By detecting blood quality parameters such as partial pressure of oxygen, partial pressure of carbon dioxide, and blood oxygen saturation, the system can simulate the quality of blood during circulation and make real-time corrections to the gas flow of the membrane oxygenator to ensure that the oxygen demand and carbon dioxide excretion demand can be maximally met. This improves the oxygenation efficiency, which can reflect the usage effect of the ECMO system. Description of the Drawings

[0038] Figure 1 It is a schematic diagram of the overall system structure of the present invention. Detailed Embodiments

[0039] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with specific embodiments.

[0040] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second", and similar terms used in the present invention do not denote any order, quantity, or importance, but are only used to distinguish different components. The terms such as "comprising" or "including" mean that the elements or items appearing before this term cover the elements or items listed after this term and their equivalents, without excluding other elements or items. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0041] Embodiment:

[0042] Please refer to Figure 1 , the present invention provides a technical solution:

[0043] A simulation circulation test system for extracorporeal membrane oxygenation (ECMO), specifically including:

[0044] A training data acquisition module, which is used to collect blood flow characteristic parameters of several users with different body characteristic parameters as training sample parameters, and map the training sample parameters to the corresponding body characteristic parameters of the users one by one to generate a training sample data set.

[0045] Wherein the blood flow characteristic parameters include blood flow rate and blood flow pressure, and the body characteristic parameters include age, weight, basal metabolic rate, and exercise level.

[0046] The specific method for obtaining blood flow rate data is as follows: Using ultrasonic Doppler technology, the blood flow velocity is measured by transmitting ultrasonic signals and receiving reflected waves. The ultrasonic probe is placed on the skin surface, and a suitable blood vessel is selected. After the ultrasonic signal passes through the blood, the frequency of the returned signal changes, and the blood flow rate can be calculated through the frequency change.

[0047] The specific method for obtaining blood flow pressure data is as follows: An electronic sphygmomanometer or a cuff-type blood pressure measuring instrument is used. By tying the cuff on the upper arm, the systolic and diastolic blood pressures are measured through the process of inflation and deflation.

[0048] Generally, as people age, their basal metabolic rate decreases, and the efficiency of the cardiovascular system may decline. In the elderly, problems such as hardening of the blood vessels and decreased myocardial contractility may occur, resulting in a decrease in blood flow rate; Aging is often accompanied by weakened blood vessel elasticity and arteriosclerosis, leading to an increase in diastolic and systolic blood pressures. Hypertension often occurs in the elderly, and the blood flow pressure is relatively higher than that of the young.

[0049] Weight gain usually leads to an increased burden on the heart. Especially in obese people, the heart needs to pump blood more vigorously to meet the body's needs. Therefore, the blood flow rate throughout the body may increase. However, obesity may also lead to an increase in vascular resistance, thus affecting the stability of the flow rate. Excessive weight is often associated with hypertension. Weight gain leads to an increase in peripheral vascular resistance, thus increasing the blood flow pressure, especially in a static state.

[0050] People with a higher basal metabolic rate usually have a larger cardiac output, and the blood flow rate will increase accordingly because the body's demand for oxygen and nutrients increases. In people with a high basal metabolic rate, during exercise or activity, the blood flow rate will increase significantly to meet the body's needs; A high basal metabolic rate may be related to a higher heart rate. If the heart pumps blood efficiently, the blood flow pressure may be maintained within a relatively ideal range.

[0051] In people who exercise regularly, the heart function will be enhanced, the ventricular wall will thicken, and the amount of blood pumped by the heart per unit time will increase, thus increasing the blood flow rate at rest. During exercise, the blood flow rate will increase significantly to meet the muscles' demand for oxygen and nutrients; Different exercise levels also have different effects on blood flow pressure. Moderate exercise can improve cardiovascular health and reduce resting blood pressure.

[0052] Preprocess the collected blood flow characteristic parameters. The preprocessing is specifically normalization preprocessing. The formula specifically used for normalization preprocessing calculation is:

[0053]

[0054] In the formula, Qi ' is the normalized data of the blood flow rate of the i-th user, Q i is the blood flow rate data of the i-th user, Q min and Q max respectively represent the minimum and maximum flow rate data among the blood flow rate data of all users collected, where i is the index of the user from whom information is collected, i ∈ [1, M], and M is the total number of users from whom information is collected.

[0055] Among them, the method for preprocessing the normalization of blood flow pressure is the same. The specific formula is:

[0056]

[0057] In the formula, U i ' is the normalized data of the blood flow pressure of the i-th user, U i is the blood flow pressure data of the i-th user, U min and U max respectively represent the minimum and maximum flow rate data among the blood flow pressure data of all users collected, where i is the index of the user from whom information is collected, i ∈ [1, M], and M is the total number of users from whom information is collected.

[0058] The method for generating the training sample data set is: one-to-one mapping of the training sample parameters and the corresponding user's body feature parameters to form a corresponding grid, and the formed grid is denoted as the training sample data set.

[0059] The prediction model training module is used to establish a neural network model, use the body feature parameter data of the training sample data set as the input of the neural network model, and the corresponding training sample parameter data as the label to train the neural network model to obtain a blood flow feature extraction model.

[0060] Based on the data of the training sample data set, a neural network model is established. Specifically, a neural network model is established through a long short-term memory network model (LSTM model). For the long short-term memory network model (LSTM model), an activation function and an optimization algorithm are selected. Among them, the Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm for the LSTM model; the formula of the Tanh function is:

[0061]

[0062] In the formula, f(r) represents the Tanh function, and the independent variable r represents the weighted sum of the inputs of the neuron, that is, the result after the inputs received by the neuron from the previous layer are weighted and summed;

[0063] Meanwhile, set the hyperparameters of the LSTM model. The hyperparameters of the LSTM model include: the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of neurons in the hidden layer;

[0064] Among them, the number of network layers is set to a 4-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32;

[0065] The input of the trained blood flow feature extraction model is the body feature parameter data of the user, including age, weight, basal metabolic rate, and exercise level, and the output is the predicted value of the corresponding blood flow feature parameter, including the predicted values of blood flow rate and blood flow pressure. Among them, the exercise level specifically refers to the average daily exercise time of the user.

[0066] Blood flow characteristics often involve time series data, such as changes in heart rate and blood pressure. LSTM is designed specifically for processing and predicting sequence data and can effectively capture the time dependence and dynamic changes in the data. When traditional recurrent neural networks (RNNs) process long sequences, they are prone to problems such as gradient disappearance or explosion, making it difficult for the model to learn features with long-term dependencies. LSTM effectively alleviates this problem through its unique structure, making training on long time series data more stable. Blood flow characteristics are usually affected by multiple body feature parameters. LSTM can process multiple input features simultaneously, adapt to the physiological differences of different individuals, and improve the generalization ability of the model. Since LSTM can process input sequences of different lengths, it can better capture individual differences and diversity when adapting to training samples under different pathological conditions, enhancing the applicability of the model.

[0067] The physiological feature analysis module is used to collect the body feature parameters of the simulated user, input them into the blood flow feature extraction model, obtain the predicted values of the blood flow feature parameters of the simulated user, and set the blood exchange flow rate and blood exchange pressure of the extracorporeal membrane oxygenation (ECMO) device according to the predicted values of the blood flow feature parameters. Use the simulated sample blood to perform simulations in the set extracorporeal membrane oxygenation (ECMO) device to obtain simulated physiological feature parameters.

[0068] The physiological feature parameters include simulated heart rate, simulated respiratory rate, and simulated central venous pressure.

[0069] Among them, the physiological characteristic parameters of the simulated user include simulated heart rate, simulated respiratory rate, and simulated central venous pressure. The specific simulation methods are as follows: Use a sine wave function to simulate the periodic fluctuations of the heart rate, introduce random noise to simulate the natural fluctuations of the heart rate, set different baseline heart rates according to different physiological states, such as rest, exercise, stress, etc., and use conditional statements to adjust the heart rate in the model. For example, the heart rate can be set to a higher value during exercise and a lower value during rest. The process of breathing can be simulated by generating a square wave or a triangular wave to reflect the periodic changes of inhalation and exhalation. The parameters used to simulate the respiratory rate can be adjusted according to different physiological states. For example, the respiratory rate increases during exercise and decreases in a relaxed state. The central venous pressure can be obtained by simulating the blood volume and blood return in the body. A baseline value can be set, and the central venous pressure can be adjusted when there are changes in blood return, such as venous dilation or constriction. According to the above mathematical model and method, values of heart rate, respiratory rate, and central venous pressure are generated over time, and corresponding time series data are generated.

[0070] The exchange parameter adjustment module is used to adaptively adjust the set blood exchange flow rate and blood exchange pressure based on the obtained physiological characteristic parameters to obtain corrected values of the dynamic blood exchange flow rate and the dynamic blood exchange pressure, and acquire the blood quality parameters when operating according to the corrected values.

[0071] Based on the collected physiological characteristic parameters, the set blood exchange flow rate and blood exchange pressure are adaptively adjusted to obtain the corrected value of the dynamic blood exchange flow rate and the corrected value of the dynamic blood exchange pressure. The formula based on which the corrected value of the dynamic blood exchange flow rate is calculated is:

[0072]

[0073] In the formula, V' Z is the corrected value of the dynamic blood exchange flow rate, V for is the predicted value of the blood flow rate, HR S is the heart rate data of the simulated user, PR S is the respiratory rate data of the simulated user, CP S is the central venous pressure data of the simulated user, HR0 is the reference heart rate data, PR0 is the reference respiratory rate data, and CP0 is the reference central venous pressure data.

[0074] It should be noted that generally, the higher the heart rate, the stronger the pumping ability of the heart and the faster the blood circulation speed. In this case, to ensure effective oxygenation and carbon dioxide excretion, the blood flow rate in the ECMO system can be appropriately increased. By increasing the blood flow rate, it can be ensured that when the heart rate increases, the oxygenation membrane can receive more blood flowing through, thereby improving the exchange efficiency of oxygen and carbon dioxide. Therefore, the heart rate data HR of the simulated user SIs directly proportional to the dynamic blood exchange flow rate correction value V', and is represented by the logarithmic function ln[1 + (HR Z - HR0 / HR0)], indicating that as the heart rate data HR of the simulated user S increases, the influence on the dynamic blood exchange flow rate correction value V' S gradually decreases. Z

[0075] The higher the respiratory rate indicates an increased respiratory demand of the patient, especially in cases of hypoxia or dyspnea. A high respiratory rate means an increased demand for oxygen by the body, and at the same time, it may also increase the demand for carbon dioxide excretion. In this case, the blood flow rate should be appropriately increased so that more blood flows through the oxygenation membrane, thereby providing sufficient oxygen supply and carbon dioxide excretion. Therefore, the respiratory rate data PR of the simulated user S is directly proportional to the dynamic blood exchange flow rate correction value V' Z and is represented by an exponential function indicating its significant influence on the dynamic blood exchange flow rate correction value.

[0076] The greater the central venous pressure usually indicates an increased right heart load, which may be caused by fluid overload, heart failure, or other factors. In this case, venous return is affected. At this time, the blood flow rate needs to be appropriately reduced to prevent an additional burden on the heart and ensure the normal operation of the oxygenation membrane. Therefore, the central venous pressure data CP of the simulated user S is inversely proportional to the dynamic blood exchange flow rate correction value V' Z and is in the form of a square (CP S - CP0) 2 indicating that when the central venous pressure data of the simulated user is significantly lower than the reference central venous pressure data, it significantly affects the dynamic blood exchange flow rate correction value.

[0077] The formula for calculating the dynamic blood exchange pressure correction value is as follows:

[0078]

[0079] In the formula, Pa' Z is the dynamic blood exchange pressure correction value, and Pa for is the predicted value of the blood flow pressure.

[0080] Among them, the reason for correcting the dynamic blood exchange pressure correction value Pa' Z is the same as that of the dynamic blood exchange flow rate correction value and will not be elaborated here. Since the adjustment range of the blood exchange pressure is small, it is represented by indicating the influence of the respiratory rate data of the simulated user on the dynamic blood exchange pressure correction value Pa' Z .

[0081] Among them, HR0 is the reference heart rate data, PR0 is the reference respiratory rate data, and CP0 is the reference central venous pressure data. These reference data are specifically set through the respective reference data ranges publicly available in medicine, and the minimum value of each reference data range is uniformly used as the reference value.

[0082] The blood quality adjustment module is used to correct the gas flow rate of the membrane oxygenator by combining the blood quality parameters with the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value to obtain the gas flow rate of the adaptive membrane oxygenator, and perform control according to the real-time gas flow rate of the adaptive membrane oxygenator, the dynamic blood exchange flow rate correction value, and the dynamic blood exchange pressure correction value to complete the simulation circulation test of extracorporeal membrane oxygenation (ECMO).

[0083] The blood quality parameters include partial pressure of oxygen, partial pressure of carbon dioxide, and blood oxygen saturation.

[0084] A pulse oximeter such as a finger clip oximeter can be used. The probe is placed on the patient's finger or earlobe, and the instrument monitors the ratio of oxyhemoglobin to deoxyhemoglobin in the blood through an optical sensor to calculate the partial pressure of oxygen, the partial pressure of carbon dioxide, and the blood oxygen saturation.

[0085] According to the obtained blood quality parameters, the gas flow rate of the membrane oxygenator is corrected by combining the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value to obtain the gas flow rate of the adaptive membrane oxygenator. The specific formula for calculating the gas flow rate of the adaptive membrane oxygenator is as follows:

[0086]

[0087] In the formula, VA' Z is the gas flow rate of the adaptive membrane oxygenator, VA0 is the initial gas flow rate of the membrane oxygenator, PCO2 is the partial pressure of carbon dioxide, is the reference value of the partial pressure of carbon dioxide, ZH P is the blood oxygen content characterization coefficient, ω1 is the weight coefficient of the sum of the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value, ω3 is the weight coefficient of the difference in the partial pressure of carbon dioxide, ω2 is the weight coefficient of the blood oxygen content characterization coefficient, where ω1>ω2>ω3 and ω1, ω2, and ω3 are all greater than 0, and ω1 + ω2 + ω3 = 1;

[0088] It should be noted that when the partial pressure of oxygen in the blood is relatively high, it usually indicates good oxygenation effect and sufficient oxygen supply. In this case, the gas flow rate of the membrane oxygenator should be appropriately reduced because the partial pressure of oxygen in the blood is already high enough and there is no need to continue to supply excessive oxygen. Reducing gas consumption and the pressure of oxygen on the membrane. If the blood oxygen saturation is relatively high, it indicates good oxygen delivery effect and sufficient oxygen supply in the body. In this case, the gas flow rate of the membrane oxygenator can also be considered to be reduced because the ideal blood oxygen saturation has been achieved and excessive oxygen may not be necessary, and oxygen waste can be reduced. Therefore, the blood oxygen content characterization coefficient ZH P is directly proportional to the gas flow rate of the adaptive membrane oxygenator.

[0089] The greater the dynamic blood exchange pressure and the correction value of the dynamic blood exchange flow rate, the gas flow rate of the membrane oxygenator should be increased to match the flow rate and ensure the quality of blood exchange. Therefore, both the dynamic blood exchange pressure and the correction value of the dynamic blood exchange flow rate are directly proportional to the gas flow rate of the adaptive membrane oxygenator, and are represented in the form of a square root to represent their comprehensive influence and make the correction smoother to avoid overcorrection.

[0090] If the partial pressure of carbon dioxide in the blood is relatively high, it means that the carbon dioxide excretion is insufficient, which may be due to insufficient gas flow rate of the membrane oxygenator. It is necessary to increase the gas flow rate of the membrane oxygenator to improve the carbon dioxide excretion efficiency. A high gas flow rate helps to increase the gas exchange efficiency on both sides of the oxygenation membrane, thus more effectively discharging carbon dioxide from the blood. Therefore, the partial pressure of carbon dioxide PCO2 is directly proportional to the gas flow rate VA' of the adaptive membrane oxygenator Z is directly proportional.

[0091] Since the gas flow rate of the membrane oxygenator should match the dynamic blood exchange pressure and the correction value of the dynamic blood exchange flow rate, and the value of the blood oxygen content characterization coefficient represents the quality of blood exchange, which is directly related to the effect of extracorporeal membrane oxygenation ECMO, while the partial pressure of carbon dioxide only has an indirect influence. Therefore, ω1>ω2>ω3 and ω1, ω2, and ω3 are all greater than 0, and ω1 + ω2 + ω3 = 1.

[0092] Among them, the blood oxygen content characterization coefficient ZH P is characterized by the partial pressure of oxygen and the blood oxygen saturation. The blood oxygen content characterization coefficient ZH P The specific calculation formula is as follows:

[0093]

[0094] In the formula, PO2 is the partial pressure of oxygen in the blood, SPO2 is the blood oxygen saturation, is the reference value of the partial pressure of oxygen in the blood, is the reference value of the blood oxygen saturation.

[0095] It should be noted that when the value of the blood oxygen saturation SPO2 in the exchanged blood is less than the reference value of the blood oxygen partial pressure, the blood oxygen content characterization coefficient ZH P is a positive value, and the gas flow rate of the membrane oxygenator should be increased to increase the oxygen content in the blood. Similarly, for the blood oxygen partial pressure, when the PO2 is less than the reference value of the blood oxygen partial pressure the gas flow rate of the membrane oxygenator needs to be increased.

[0096] Among them, the reference value of the blood oxygen partial pressure and the reference value of the blood oxygen saturation are set according to the common ranges specified by medical regulations. The reference value of the blood oxygen saturation is generally taken between 92% and 95%, and the reference value of the blood oxygen partial pressure is usually between 75 mmHg and 100 mmHg.

[0097] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0098] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or by the combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0099] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0100] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered within the protection scope of this application.

Claims

1. An in vitro membrane lung oxygenation (ECMO) simulation circulation test system, characterized in that Specifically, it includes: A training data acquisition module, which is used to collect the blood flow characteristic parameters of users with several different physical characteristic parameters as training sample parameters, map the training sample parameters one by one with the corresponding users' physical characteristic parameters, and generate a training sample data set; A prediction model training module, which is used to establish a neural network model, use the physical characteristic parameter data of the training sample data set as the input of the neural network model, and the corresponding training sample parameter data as labels to train the neural network model to obtain a blood flow characteristic extraction model; A physiological characteristic analysis module, which is used to collect the physical characteristic parameters of a simulated user, input them into the blood flow characteristic extraction model to obtain the predicted value of the blood flow characteristic parameters of the simulated user, set the blood exchange flow rate and blood exchange pressure of the extracorporeal membrane oxygenation (ECMO) device according to the predicted value of the blood flow characteristic parameters, and use the simulated sample blood to perform simulation in the set extracorporeal membrane oxygenation (ECMO) device to obtain simulated physiological characteristic parameters; An exchange parameter adjustment module, which is used to adaptively adjust the set blood exchange flow rate and blood exchange pressure based on the obtained physiological characteristic parameters to obtain the corrected values of the dynamic blood exchange flow rate and the dynamic blood exchange pressure, and obtain the blood quality parameters when operating according to the corrected values; A blood quality adjustment module, which is used to correct the gas flow rate of the membrane oxygenator by combining the blood quality parameters with the corrected value of the dynamic blood exchange flow rate and the corrected value of the dynamic blood exchange pressure to obtain the gas flow rate of the adaptive membrane oxygenator, and perform control according to the real-time gas flow rate of the adaptive membrane oxygenator, the corrected value of the dynamic blood exchange flow rate and the corrected value of the dynamic blood exchange pressure to complete the simulation cycle test of the extracorporeal membrane oxygenation (ECMO); 2. The simulated circulation test system for extracorporeal membrane oxygenation (ECMO) according to claim 1, wherein: Preprocess the collected blood flow characteristic parameters. The specific preprocessing is normalization preprocessing. The method of normalizing the blood flow pressure is the same. The generation method of the training sample data set is: map the training sample parameters one by one with the corresponding users' physical characteristic parameters to form a corresponding grid, and record the formed grid as the training sample data set; Among them, the blood flow characteristic parameters include blood flow rate and blood flow pressure, and the physical characteristic parameters include age, weight, basal metabolic rate and exercise level.

3. The simulation cycle test system for extracorporeal membrane oxygenation (ECMO) according to claim 2, characterized in that: Establish a neural network model through a long short-term memory network model (LSTM model), select the Tanh function as the activation function, and select Adam as the optimization algorithm for the LSTM model; Set the hyperparameters of the LSTM model. The hyperparameters of the LSTM model include: number of network layers, number of iterations, learning rate, batch size, number of training times, batch processing quantity and number of neurons in the hidden layer; Among them, the number of network layers is set to a 4-layer network structure, the number of iterations is set to 200, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of neurons in the hidden layer is 32; The input of the trained blood flow feature extraction model is the user's body feature parameter data, including age, weight, basal metabolic rate, and exercise level, and the output is the predicted values of the corresponding blood flow feature parameters, including blood flow rate and blood flow pressure prediction values, where the exercise level specifically refers to the user's average daily exercise time.

4. The simulated circulation test system for extracorporeal membrane oxygenation (ECMO) according to claim 1, wherein: Based on the collected physiological feature parameters, the set blood exchange flow rate and blood exchange pressure are adaptively adjusted to obtain the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value. The physiological feature parameters include simulated heart rate, simulated respiratory rate, and simulated central venous pressure. The formula based on which the dynamic blood exchange flow rate correction value is calculated is: where, V' Z is the dynamic blood exchange flow rate correction value, V for is the predicted blood flow rate, HR S is the heart rate data of the simulated user, PR S is the respiratory rate data of the simulated user, CP S is the central venous pressure data of the simulated user, HR0 is the reference heart rate data, PR0 is the reference respiratory rate data, and CP0 is the reference central venous pressure data.

5. The simulation circulation test system for extracorporeal membrane oxygenation (ECMO) according to claim 4, characterized in that: The formula based on which the dynamic blood exchange pressure correction value is calculated is: where Pa' Z is the dynamic blood exchange pressure correction value, in Pa for is the predicted value of blood flow pressure.

6. The simulated circulation test system for extracorporeal membrane oxygenation (ECMO) according to claim 5, wherein: According to the obtained blood quality parameters, the gas flow rate of the membrane oxygenator is corrected by combining the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value to obtain the gas flow rate of the adaptive membrane oxygenator. The blood quality parameters include partial pressure of oxygen, partial pressure of carbon dioxide, and blood oxygen saturation. The formula based on which the gas flow rate of the adaptive membrane oxygenator is specifically calculated is: where VA' Z is the gas flow rate of the adaptive membrane oxygenator, VA0 is the initial gas flow rate of the membrane oxygenator, PCO2 is the partial pressure of carbon dioxide, is the reference value of the partial pressure of carbon dioxide, ZH P is the blood oxygen content characterization coefficient, ω1 is the weight coefficient of the sum of the dynamic blood exchange flow rate correction value and the dynamic blood exchange pressure correction value, ω3 is the weight coefficient of the partial pressure difference of carbon dioxide, ω2 is the weight coefficient of the blood oxygen content characterization coefficient, where ω1>ω2>ω3 and ω1, ω2 and ω3 are all greater than 0, and ω1 + ω2 + ω3 = 1; Among them, the blood oxygen content characterization coefficient ZH P is characterized by partial pressure of oxygen and blood oxygen saturation. The blood oxygen content characterization coefficient ZH P The specific calculation formula is as follows: Wherein, PO2 is the partial pressure of oxygen in blood, and SPO2 is the oxygen saturation of blood, is the reference value of the partial pressure of oxygen in blood, is the reference value of the oxygen saturation of blood.

Images