CN116999689A - Traffic determination method, traffic detection model training method, equipment and media - Google Patents

Abstract

This application relates to a flow determination method, a flow detection model training method, equipment and media, and belongs to the technical field of medical devices. The method includes: obtaining perfusion data and blood pressure data corresponding to a transcatheter ventricular assist device and corresponding transcatheter ventricular assist device. The motor operating data is collected; flow detection and processing is performed based on the perfusion data, blood pressure data and motor operating data to obtain the flow data corresponding to the transcatheter ventricular assist device. The technical solution provided by this application greatly reduces the dependence of flow detection on the flow sensor, effectively solves the technical problem of difficulty in detecting the pumping flow of the transcatheter ventricular assist device inside the heart, reduces the complexity of flow detection, and improves improve the efficiency of traffic detection.

Description

Traffic determination method, traffic detection model training method, equipment and media

Technical field

This application relates to a flow determination method, a flow detection model training method, equipment and media, and belongs to the technical field of medical devices.

Background technique

Transcatheter ventricular assist devices are used to provide mechanical circulatory assistance to patients. When the transcatheter ventricular assist device works, it pumps blood from the heart's ventricles to the arteries to achieve ventricular assist function. During the operation of the transcatheter ventricular assist device, it is usually necessary to measure the current pumped blood flow to determine the operating status of the transcatheter ventricular assist device.

However, the pump head of the interventional pump in a transcatheter ventricular assist device needs to be percutaneously inserted into the heart through peripheral blood vessels. Therefore, the pump head is small in size and does not have enough space to install a flow sensor, making it difficult to measure blood flow.

Contents of the invention

This application provides a flow determination method, a flow detection model training method, equipment and media, which solves the technical problem of difficulty in monitoring blood flow pumped by a transcatheter ventricular assist device. This application provides the following technical solutions:

In one aspect, a flow determination method is provided, the method is applied to a transcatheter ventricular assist device, the method includes:

Obtaining perfusion data, blood pressure data corresponding to the transcatheter ventricular assist device, and motor operation data corresponding to the transcatheter ventricular assist device;

Perform flow detection processing according to the perfusion data, the blood pressure data and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device.

Optionally, the perfusion data includes at least one perfusion parameter data, the at least one perfusion parameter data includes at least one of perfusion fluid pressure data, perfusion fluid flow data and perfusion pump rotational speed data, and the blood pressure data includes at least A kind of blood pressure parameter data, the at least one blood pressure parameter data includes at least one of arterial pressure data, ventricular pressure data, and pumping pressure difference data;

Optionally, performing flow detection processing based on the perfusion data, the blood pressure data and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device includes:

The flow rate data is obtained by performing flow detection processing according to the perfusion fluid pressure data, the arterial pressure data and the motor operation data.

Optionally, the motor operating data includes motor speed data and motor current data, and the flow data is obtained by performing flow detection processing according to the perfusion fluid pressure data, the arterial pressure data and the motor operating data, include:

The flow rate data is obtained by performing flow detection processing based on the perfusion fluid pressure data, the arterial pressure data, the motor speed data and the motor current data.

Optionally, the transcatheter ventricular assist device includes:

An interventional catheter pump, which includes a catheter, a pump head connected to the distal end of the catheter, and a coupling assembly connected to the proximal end of the catheter. The pump head can be delivered to a desired location in the heart by the catheter. position to perform blood pumping operations;

A perfusion channel, the perfusion channel at least runs through the conduit, the inlet of the perfusion channel is provided on the coupling component, and the outlet of the perfusion channel is provided at the pump head; the perfusion liquid in the perfusion channel passes through The outlet enters a desired location in the cardiovascular system; the desired location in the cardiovascular system includes at least one of the aorta, pulmonary artery, left ventricle, right ventricle, left atrium, and right atrium;

The perfusion fluid pressure data represents the sum of the perfusion fluid pressures generated by the perfusion fluid from the pressure collection position to the outlet.

Optionally, performing flow detection processing based on the perfusion data, the blood pressure data and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device includes:

Input the perfusion data, the blood pressure data and the motor operation data into a preset flow detection model for flow detection processing, and output the flow data;

Wherein, the flow detection model is a machine learning model obtained after training based on multiple sets of sample data. Each set of sample data includes corresponding perfusion sample data, blood pressure sample data, motor operation sample data and sample flow data.

Optionally, inputting the perfusion data, blood pressure data and motor operation data into a preset flow detection model for flow detection processing and outputting the flow data includes:

Input the perfusate pressure data, arterial pressure data, motor speed data and motor current data into the preset Gaussian model for flow detection processing, and output the flow data;

Wherein, the perfusion data includes the perfusion fluid pressure data, the blood pressure data includes the arterial pressure data, the motor operation data includes the motor speed data and the motor current data, and the preset flow detection The model includes the preset Gaussian model, which is a Gaussian model obtained after training based on multiple sets of sample data.

Optionally, the step of inputting the perfusion data, the blood pressure data and the motor operation data into a preset flow detection model for flow detection processing, and before outputting the flow data, further includes:

The first flow detection model corresponding to the transcatheter ventricular assist device is determined based on the data type of the perfusion data and/or the data type of the blood pressure data; the data type of the perfusion data is used to indicate the type of perfusion parameter data included in the perfusion data. ; The data type of the blood pressure data is used to indicate the type of blood pressure parameter data included in the blood pressure data.

Correspondingly, inputting the perfusion data, the blood pressure data and the motor operation data into a preset flow detection model to perform flow detection processing and outputting the flow data includes:

The perfusion data, the blood pressure data and the motor operation data are input into the first flow detection model, and the flow data is output.

The preset flow detection model includes flow detection models corresponding to different data types. The data type of the perfusion sample data in each group of sample data corresponding to the first flow detection model is consistent with the data type of the currently acquired perfusion data. The data type of the blood pressure sample data is consistent with the data type of the currently acquired perfusion data.

Optionally, the step of inputting the perfusion data, the blood pressure data and the motor operation data into a preset flow detection model for flow detection processing, and before outputting the flow data, further includes:

Obtain the first collection position of the perfusion data and the second collection position of the blood pressure data;

Determine a second flow detection model based on the first collection position and the second collection position;

Among them, the preset flow detection model includes flow detection models corresponding to different collection locations. Each set of sample data corresponding to the second flow detection model includes: perfusion sample data collected based on the first collection location, perfusion sample data collected based on the second collection location blood pressure sample data and motor operation sample data.

Correspondingly, the perfusion data, the blood pressure data and the motor operation data are input into a preset flow detection model for flow detection processing, and the flow data is output, including:

The perfusion data, the blood pressure data and the motor operation data are input into the second flow detection model to perform flow detection processing, and the flow data is output.

On the other hand, a method for training a traffic detection model is provided, and the method includes:

Obtain a sample data set collected during the operation of the transcatheter ventricular assist device. Each set of sample data in the sample data set includes perfusion sample data, blood pressure sample data, motor operation sample data and corresponding perfusion sample data corresponding to the transcatheter ventricular assist device. Flow measurement data;

Perform model training on a preset machine learning model according to the perfusion sample data, the blood pressure sample data, the motor operation sample data and the flow measurement data to obtain a flow detection model;

Wherein, the flow detection model is used to detect the flow data corresponding to the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data generated during the application of the transcatheter ventricular assist device.

Optionally, the sample data set includes multiple sets of training data and multiple sets of test data, and the preset data is preset based on the perfusion sample data, the blood pressure sample data, the motor operation sample data and the flow measurement data. Carry out model training on the machine learning model to obtain the traffic detection model, including:

Perform modeling processing based on the perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data in the multiple sets of training data to generate a first machine learning model;

Input the perfusion sample data, blood pressure sample data, and motor operation sample data in each set of test data into the first machine learning model for flow detection processing, and output the flow detection data corresponding to each set of test data;

Compare the flow measurement data in each set of test data with the flow detection data corresponding to each set of test data, and determine the loss information corresponding to the first machine learning model;

The model parameters of the first machine learning model are adjusted based on the loss information to obtain the traffic detection model.

Optionally, the first machine learning model includes a Gaussian model, and the flow measurement data in each set of test data is compared with the flow detection data corresponding to each set of test data to determine whether the first machine The loss information corresponding to the learning model includes:

Determine the average flow data based on the flow measurement data in each group of sample data;

Compare the average flow data, the flow detection data corresponding to each set of test data, and the flow measurement data in each set of test data to obtain the variance data and root mean square error data corresponding to the Gaussian model. The loss information includes The variance data and the root mean square error data.

Optionally, the flow measurement data is determined by the change in liquid weight within a unit time period.

Optionally, the perfusion sample data in the sample data set includes perfusion fluid pressure sample data corresponding to the transcatheter ventricular assist device, and the blood pressure sample data in the sample data set includes arterial pressure corresponding to the transcatheter ventricular assist device. Sample data, the motor operation sample data in the sample data set includes motor speed sample data and motor current sample data corresponding to the transcatheter ventricular assist device;

Correspondingly, the flow detection model is specifically used to detect the flow data based on the perfusate pressure data, arterial pressure data, motor speed data and motor current data generated during the application of the transcatheter ventricular assist device.

On the other hand, a flow determining device is provided, the flow determining device is applied to a transcatheter ventricular assist device, and the device includes:

A data acquisition module, configured to acquire perfusion data, blood pressure data corresponding to the transcatheter ventricular assist device, and motor operation data corresponding to the transcatheter ventricular assist device;

A flow determination module, configured to perform flow detection processing based on the perfusion data, the blood pressure data and the motor operation data, to obtain flow data corresponding to the transcatheter ventricular assist device.

On the other hand, a training device for a traffic detection model is provided, and the device includes:

A data acquisition module is used to obtain a sample data set collected during the operation of the transcatheter ventricular assist device. Each set of sample data in the sample data set includes perfusion sample data, blood pressure sample data, and motor corresponding to the transcatheter ventricular assist device. Run sample data and corresponding flow measurement data;

A model training module, configured to perform model training on a preset machine learning model based on the perfusion sample data, the blood pressure sample data, the motor operation sample data, and the flow measurement data to obtain a flow detection model;

Wherein, the flow detection model is used to detect the flow data corresponding to the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data generated during the application of the transcatheter ventricular assist device.

On the other hand, a transcatheter ventricular assist device is provided. The transcatheter ventricular assist device includes: a computer device, the computer device includes a processor and a memory; a program is stored in the memory, and the program is processed by the processor. The server is loaded and executed to implement the traffic determination method provided by the above aspect; or, to implement the training method of the traffic detection model provided by the above aspect.

On the other hand, a computer device is provided, the device includes a processor and a memory; a program is stored in the memory, and the program is loaded and executed by the processor to implement the flow determination method provided by the above aspect; or, Implement the training method of the traffic detection model provided in the above aspects.

On the other hand, a computer-readable storage medium is provided, and a program is stored in the storage medium. When the program is executed by a processor, it is used to implement the flow determination method provided by the above aspect; or, implement the flow detection provided by the above aspect. Model training method.

In another aspect, a computer program product includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes to implement the traffic determination method provided by the above aspect; or, implement the flow determination method provided by the above aspect. Training method for traffic detection model.

The flow rate determination method provided in one embodiment of the present application can detect the pumping flow rate of the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data corresponding to the transcatheter ventricular assist device, which greatly reduces the flow detection requirements for the flow rate. The dependence of the sensor effectively solves the technical problem of difficulty in detecting the pumping flow of the transcatheter ventricular assist device inside the heart, reduces the complexity of flow detection, and improves the efficiency of flow detection.

In addition, since the perfusion state of the perfusion fluid, the pressure state of the blood, and the operating state of the motor are all related to the flow rate pumped by the transcatheter ventricular assist device, therefore, based on the perfusion fluid pressure data characterizing the perfusion state of the perfusion fluid, characterizing the blood The arterial pressure data of the pressure state and the motor operating data representing the operating state of the motor are used to determine the flow data, which can ensure the accuracy of flow detection.

In addition, by training a machine learning model as a traffic detection model, the traffic detection model can learn the data correlation between various parameters from the distribution of multiple sets of sample data. By detecting traffic data through the traffic detection model, traffic detection can be ensured. accuracy of results.

The above description is only an overview of the technical solutions of the present application. In order to have a clearer understanding of the technical means of the present application and implement them according to the contents of the specification, the preferred embodiments of the present application are described in detail below with reference to the accompanying drawings.

Description of the drawings

Figure 1 is a block diagram of a transcatheter ventricular assist device provided by an embodiment of the present application;

Figure 2 is a schematic structural diagram of the connection between the perfusion component and the working component provided by an embodiment of the present application;

Figure 3 is a flow chart of a flow determination method provided by an embodiment of the present application;

Figure 4 is a flow chart of a training method for a traffic detection model provided by an embodiment of the present application;

Figure 5 is a block diagram of a flow determination device provided by an embodiment of the present application;

Figure 6 is a block diagram of a training device for a traffic detection model provided by an embodiment of the present application;

Figure 7 is a block diagram of a computer device provided by an embodiment of the present application.

Detailed ways

Specific implementations of the present application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present application but are not intended to limit the scope of the present application.

In the traditional flow determination method, flow data is generally detected by installing a flow sensor in the blood pump. However, for transcatheter ventricular assist devices, since the blood pump in the device is an interventional catheter pump, and the pump head in the interventional catheter pump needs to be implanted into the body during use, the volume of the pump head is usually relatively large. Small. In this case, it is difficult to install a flow sensor in the pump head to detect flow data.

Based on this, this application provides a flow rate determination method, which can perform flow detection processing based on the perfusion data, blood pressure data and motor operation data corresponding to the transcatheter ventricular assist device, thereby detecting the flow data pumped by the transcatheter ventricular assist device. , which can detect flow without the need for a flow sensor in the transcatheter ventricular assist device.

In this application, before introducing the traffic determination method, the application scenarios of the traffic determination method are first introduced.

The flow determination method provided in this application is applied to a transcatheter ventricular assist device, which is used to provide mechanical circulatory assistance to patients. In one illustrative scenario, a transcatheter ventricular assist device can be used as a left ventricular assist, operating to pump blood from the left ventricle into the ascending aorta. In other scenarios that are feasible and cannot be definitely ruled out, a transcatheter ventricular assist device can also be used as a right ventricular assist, pumping blood from the veins into the right ventricle. Alternatively, the transcatheter ventricular assist device may also be used to pump blood from the vena cava and/or right atrium into the right ventricle, from the vena cava and/or right atrium into the pulmonary artery, and/or from the renal vein into the vena cava, etc. This application does not limit the application scenarios of transcatheter ventricular assist devices. In the following embodiments, a transcatheter ventricular assist device used to pump blood in the ventricle to the aorta is used as an example for description.

Referring to FIG. 1 , a block diagram of a transcatheter ventricular assist device 100 is provided according to an embodiment of the present application. The device 100 at least includes an interventional catheter pump 110 .

The pump head of the interventional catheter pump 110 can be percutaneously inserted into the heart through peripheral blood vessels. For example, the pump head will be inserted between the left ventricle and the ascending aorta, and the blood inlet of the pump head will be inserted into the left ventricle. The blood outlet will be placed in the ascending aorta to pump blood from the left ventricle into the ascending aorta to achieve ventricular assist function.

Optionally, the interventional catheter pump 110 includes a pump head 111, a coupler and a driver. The driver is detachably connected to the coupler. After the driver is connected to the coupler, the rotation of the impeller in the pump head 111 can be controlled via the coupler and the catheter.

The driver includes a driving motor, and an impeller is provided in the pump head 111. The impeller is drivingly connected to the driving motor, so as to drive the rotation of the impeller by controlling the rotation of the motor, thereby driving the pump head 111 to work. The pump head 111 may also include other components required for operation. For example, the pump head 111 may also include a pump casing to accommodate an impeller. This embodiment does not enumerate the components included in the pump head 111 one by one.

In the interventional catheter pump 110, the pump head 111 is drivingly connected to the drive motor in the driver through the catheter.

The transcatheter ventricular assist device 100 also includes a perfusion component 120. The perfusion component 120 is used to provide perfusion fluid for the interventional catheter pump 110. The perfusion fluid can be a physiological fluid required to maintain human body functions, such as physiological saline, glucose solution, anticoagulation. agent, or any combination of the above.

Optionally, the perfusion solution is used to perfuse the interventional catheter pump through the catheter to eliminate air in the catheter gap and prevent air from entering the blood vessels and heart through the catheter, thereby causing adverse physiological reactions. The perfusate can flow through the perfusion channel and into the desired location in the cardiovascular system through the opening at the pump head. Optionally, the desired location in the cardiovascular system includes at least one of the aorta, pulmonary artery, left ventricle, right ventricle, left atrium, and right atrium.

In one example, as shown in FIG. 2 , the perfusion assembly 120 includes a perfusate injection interface 121 , through which the perfusate is injected into the interventional catheter pump and flows through the catheter 210 into the human body. The above-mentioned perfusate injection interface 121 may be the inlet of the perfusion channel.

Optionally, the transcatheter ventricular assist device 100 also includes a peristaltic pump for pumping the perfusate. Schematically, the perfusion fluid pipeline can be clamped on the peristaltic pump. This embodiment does not limit the installation position of the peristaltic pump.

In the embodiment of the present application, in order to detect the flow data of the interventional catheter pump, the transcatheter ventricular assist device may also include a first detection unit 130, a second detection unit 140, a third detection unit 150, and a first detection unit 150. The unit 130, the second detection unit 140 and the third detection unit 150 are respectively connected to the control unit 160.

The first detection unit 130 is used to detect perfusion data corresponding to the transcatheter ventricular assist device;

The second detection unit 140 is used to detect blood pressure data corresponding to the transcatheter ventricular assist device;

The third detection unit 150 is used to detect motor operating data corresponding to the transcatheter ventricular assist device;

The control unit 160 is used to obtain the perfusion data, blood pressure data corresponding to the transcatheter ventricular assist device, and the motor operating data corresponding to the transcatheter ventricular assist device; perform flow detection processing based on the perfusion data, blood pressure data, and motor operating data to obtain the transcatheter ventricular Flow data corresponding to the auxiliary device.

The technical solution provided by the embodiments of this application can detect the pumping flow of the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data corresponding to the transcatheter ventricular assist device, which greatly reduces the impact of flow detection on the flow sensor. Dependence, it effectively solves the technical problem of difficulty in detecting the pumping flow of the transcatheter ventricular assist device inside the heart, reduces the complexity of flow detection, and improves the efficiency of flow detection.

Optionally, the control unit 160 performs flow detection processing based on the perfusion data, blood pressure data, and motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device, including: inputting the perfusion data, blood pressure data, and motor operation data into a preset flow rate The detection model performs traffic detection processing and outputs traffic data.

Among them, the flow detection model is a machine learning model obtained after training based on multiple sets of sample data. Each set of sample data includes corresponding perfusion sample data, blood pressure sample data, motor operation sample data and sample flow data.

The traffic detection model in the control unit 160 may be trained based on the control unit 160, or may be configured to the control unit 160 after training in other devices. This embodiment does not limit the training environment of the traffic detection model.

Taking the flow detection model trained based on the control unit 160 as an example, the control unit 160 is also used to: obtain a sample data set collected during the operation of the transcatheter ventricular assist device. Each set of sample data in the sample data set includes the corresponding data of the transcatheter ventricular assist device. Perfusion sample data, blood pressure sample data, motor operation sample data and corresponding flow measurement data; perform model training on the preset machine learning model based on perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data to obtain the flow rate Detection model.

Wherein, the flow detection model is used to detect the flow data corresponding to the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data generated during the application of the transcatheter ventricular assist device.

Illustratively, the sample data set may be collected by the first detection unit, the second detection unit, and the third detection unit, or may be sent by other devices. This embodiment does not limit the acquisition method of the sample data set.

In actual implementation, if the training process of the traffic detection model is implemented in other devices, the steps are the same as the above training steps, and will not be described in detail here in this embodiment.

Optionally, the above control unit 160 can also be implemented as a device independent of the transcatheter ventricular assist device. This embodiment does not limit the implementation of the control unit 160 .

Illustratively, the control unit 160 at least includes a processor and a memory, and a program is stored in the memory. The program is loaded and executed by the processor to implement the traffic determination method in this application; or, to implement the traffic detection model in this application. training methods.

In actual implementation, the transcatheter ventricular assist device may also include other components required for operation, such as power supply components, display screens, etc. This embodiment does not introduce the components included in the transcatheter ventricular assist device one by one.

The flow rate determination method provided by an embodiment of the present application is introduced below. This method is applied to the transcatheter ventricular assist device shown in Figure 1. The execution subject of each step of the method can be the control unit in the transcatheter ventricular assist device, or It may also be other devices that are communicatively connected to the transcatheter ventricular assist device, such as computers, tablets, etc. This embodiment does not limit the device types of other devices. Figure 3 is a flow chart of a traffic determination method provided by an embodiment of the present application. The method at least includes the following steps:

Step 301: Obtain perfusion data, blood pressure data corresponding to the transcatheter ventricular assist device, and motor operation data corresponding to the transcatheter ventricular assist device.

The perfusion data is used to characterize the perfusion status of the perfusate in the perfusion component. In a transcatheter ventricular assist device, the perfusion fluid enters the heart through the internal gap of the catheter, and the perfusion channel (formed by the perfusion pipeline and the catheter) communicates with the blood vessel or the interior of the heart, that is, with the blood flow channel, that is, the flow of the perfusion fluid There is a correlation with the flow of blood. Therefore, perfusion data can be obtained to determine flow data.

Blood pressure data is used to characterize the pressure state of blood in the body. There is a strong correlation between blood flow and blood pressure. Therefore, blood pressure data at a preset position can be acquired when determining blood flow.

Motor operating data is used to characterize the corresponding operating efficiency of the motor in a transcatheter ventricular assist device. The type of fluid (such as viscosity) that is pumped through the transcatheter ventricular assist device affects the energy efficiency of the motor at the set speed. At the set speed, the power consumed by the motor for pumping different types of liquids is also different. It can be seen that there is a correlation between the motor operating data and the type of liquid pumped by the blood pump, and there is also a correlation between the flow rate and the liquid type. At the same time, the motor is also the power output source that drives the impeller to rotate, so the above motor operating data can be obtained to determine traffic data.

Optionally, the above-mentioned perfusion data includes at least one perfusion parameter data. Optionally, the above-mentioned at least one perfusion parameter data includes but is not limited to at least one of the following:

1. Perfusion fluid pressure data. Optionally, the first detection unit 130 includes a perfusion fluid pressure sensor. Optionally, the perfusion fluid pressure sensor can be arranged at a designated position on the perfusion channel. This application does not limit the location of the perfusion fluid pressure sensor.

In an exemplary embodiment, a transcatheter ventricular assist device includes:

An interventional catheter pump includes a catheter, a pump head connected to the distal end of the catheter, and a coupling assembly connected to the proximal end of the catheter. The pump head can be transported by the catheter to a desired location in the heart to perform a blood pumping operation.

The perfusion channel runs through at least the conduit, the inlet of the perfusion channel is set on the coupling component, and the outlet of the perfusion channel is set at the pump head; the perfusate in the perfusion channel enters the desired position in the cardiovascular system through the outlet; in the cardiovascular system The desired location includes at least one of the aorta, pulmonary artery, left ventricle, right ventricle, left atrium, and right atrium.

The perfusate pressure data represents the sum of the perfusate pressure generated by the perfusate from the pressure collection position to the outlet. That is, the perfusate pressure data is the sum of all pipeline pressures between the position of the pressure sensor (ie, the pressure collection position) and the outlet and the outlet pressure.

2. Perfusate flow data. Optionally, the first detection unit 130 includes a perfusate flow sensor. Optionally, the perfusate flow sensor can be disposed in the perfusing pipeline of the perfusate. This application does not limit the placement location of the flow sensor. In addition, the perfusate flow rate can also be determined based on the volume change of the perfusate.

3. Perfusion pump speed data. The perfusion pump is used to drive the perfusion fluid flow, and the perfusion pump speed data is used to characterize the speed of the perfusion pump. The perfusion pump can be a peristaltic pump or other types of pumps (such as embolization pumps, etc.). Optionally, the first detection unit 130 includes a rotational speed sensor corresponding to the perfusion pump, and the rotational speed sensor is used to detect rotational speed data of the rotational speed pump (such as a peristaltic pump).

The above blood pressure data can be detected by the blood pressure sensor accessory in the transcatheter ventricular assist device, transmitted by other equipment, or entered manually.

Optionally, the second detection unit 140 includes a blood pressure sensor that collects blood pressure data. Optionally, the blood pressure data includes at least one blood pressure parameter data, and the at least one blood pressure parameter data includes but is not limited to at least one of the following:

1. Arterial pressure data. Optionally, the arterial pressure data includes but is not limited to aortic pressure data and pulmonary artery pressure data. Because the interventional catheter pump in the transcatheter ventricular assist device works by pumping blood from the ventricle to the arteries, such as the left ventricle to the aorta and the right ventricle to the pulmonary artery. Therefore, there is a correlation between aortic pressure data or pulmonary artery pressure data and pumping flow, which can be selected to determine flow.

In a possible implementation, an arterial pressure sensor is disposed on the catheter for detecting arterial blood pressure. The arterial pressure sensor enters the human body with the percutaneous intervention of the pump head. Once it reaches the designated location, it can measure arterial blood pressure, such as aortic pressure data or pulmonary artery pressure data.

In another possible implementation, the arterial pressure sensor is disposed in an extracorporeal pipeline connected to the artery and is also used to detect arterial blood pressure. The embodiment of the present application does not limit the installation position of the arterial pressure sensor.

2. Ventricular pressure data. Optionally, the ventricular pressure data includes, but is not limited to, left ventricular pressure data and right ventricular pressure data. For similar reasons as arterial pressure data, left ventricular pressure data or right ventricular pressure data are correlated with pump flow, and therefore, the ventricular pressure data described above can be selected to determine flow. The ventricular pressure sensor is similar to the arterial pressure sensor, but its installation location is different from that of the arterial pressure sensor, which is not limited in the embodiments of the present application.

3. Pumping pressure difference data between arterial pressure data and ventricular pressure data. The above pumping pressure difference data represents the pressure difference between arterial pressure and ventricular pressure, which is the pressure difference corresponding to the interventional catheter pump.

Optionally, the motor operating data includes but is not limited to: motor speed data and motor current data. The motor speed data represents the motor speed driving the motor in the transcatheter ventricular assist device, and the motor current data represents the motor current driving the motor in the transcatheter ventricular assist device. Optionally, the drive motor is a brushless DC motor. There can be a time series correspondence between the motor current data and the motor speed data. The motor current data can reflect the power consumption of the motor under the corresponding motor speed data, thus reflecting the operating energy efficiency of the motor under the current liquid type.

In the embodiment of the present application, the obtained perfusion data, blood pressure data and motor operation data correspond to each other. Based on this, the perfusion data, blood pressure data and motor operation data are collected synchronously, or there is a gap between each data in a set of data. The maximum collection time interval is less than the threshold.

Step 302: Perform flow detection processing based on the perfusion data, blood pressure data and motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device.

The above-mentioned perfusion data, blood pressure data and motor operation data have a correlation relationship with the flow rate respectively. In a possible implementation, the corresponding flow data is determined based on the data quantification relationship between the perfusion data, blood pressure data and motor operation data. .

In another possible implementation, the correlation between the above-mentioned types of data can be learned based on a machine learning model to determine the traffic.

Optionally, the above-mentioned step 302 includes: inputting the perfusion data, blood pressure data and motor operation data into a preset flow detection model for flow detection processing, and outputting the flow data.

Among them, the flow detection model is a machine learning model obtained after training based on multiple sets of sample data. Each set of sample data includes corresponding perfusion sample data, blood pressure sample data, motor operation sample data and sample flow data. The specific training process of the traffic detection model is detailed in the following embodiments, and will not be described in detail in this embodiment.

Since the above motor operating data can characterize the operating energy efficiency of the motor, and the operating energy efficiency of the motor is affected by the type of liquid being pumped, such as viscosity, it means that the flow rate is directly related to the operating energy efficiency of the motor, and the correlation factors are relatively complicated. In addition to In addition to changes in liquid type (such as viscosity) that can cause changes in flow rate, the operating status of the motor itself can also cause changes in flow rate.

Specifically, for the motor, at the set speed, the motor pumps different liquid types, and the electric energy consumed by the motor may also be different. For example, a motor that pumps a liquid with a higher viscosity consumes more power than a motor that pumps a liquid with a smaller viscosity. Therefore, motor operating data can be used to determine flow data.

Moreover, when the blood flow path is determined and the blood flow path is connected with the perfusion fluid flow path, the liquid type and blood pressure simultaneously affect the flow rate of the blood pump, and the perfusion fluid flow will also be affected by changes in blood flow. Therefore, by simultaneously inputting the motor operating data, perfusion data, and blood pressure data associated with the liquid type into the trained flow detection model, the learned flow rate can be extracted during training based on the flow detection model without using a flow sensor. , motor operating data, perfusion data and blood pressure data are related to determine the current flow data, which greatly reduces the dependence of flow detection on the flow sensor, reduces the complexity of flow detection, and improves the efficiency of flow detection.

In addition, since the flow detection model is trained directly based on perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data, the flow detection model can directly and effectively learn flow, motor operation data, perfusion data and blood pressure data. There is a correlation between them, that is, the flow detection model does not depend on the determination of the liquid type (such as the determination of the liquid viscosity). Compared with the method in the Bernoulli equation that requires calculating the viscosity and then calculating the flow rate, in the technical solution provided by the embodiment of the present application, the method of flow detection through the flow detection model does not depend on the liquid type, such as the liquid viscosity. It can be understood that the specific method is Because there is no need to input liquid type information into the flow detection model, there is no intermediate step to calculate or determine the liquid type in the entire flow detection process, such as an intermediate step to calculate viscosity.

In addition to the above-mentioned motor operation data and blood pressure data, perfusion data unique to transcatheter ventricular assist devices are also introduced to conduct flow detection from more dimensions. This can not only ensure the accuracy of flow detection of catheter pump flow, but also eliminate the need for actual flow detection. Determining specific blood parameters effectively reduces the calculation steps, reduces the computational complexity of detecting flow data, and effectively improves the detection speed. It is also more suitable for control hosts with limited computing power in transcatheter ventricular assist devices.

In some embodiments, the sample data set has no constraints on liquid type. For example, the sample data set does not set viscosity range constraints for liquid viscosity, and the liquid viscosity corresponding to each set of sample data in the sample data set is not restricted by the preset viscosity range. Since there are data samples corresponding to various liquid types in the sample data set, the flow detection model trained based on these data is a universal flow detection model for each liquid type and is an independent and single model.

In a possible implementation, the traffic detection model may be a trained neural network model. When training the neural network model, the flow measurement data in the sample data can be used as supervision information for the neural network model, thereby constraining the accuracy of the output flow of the trained neural network model, thereby allowing the neural network model to learn the input data (perfusion sample data , blood pressure sample data, motor operation sample data) and the output data (flow measurement data), so that when new data is input, the corresponding flow data can be accurately predicted.

In another possible implementation, the traffic detection model may be a trained Gaussian model. During Gaussian model training, the mean vector and covariance matrix corresponding to each group of sample data will be determined based on the perfusion sample data, blood pressure sample data, motor operation sample data and corresponding flow measurement data in each group of sample data. The covariance matrix can effectively represent the correlation between each variable. Therefore, when the model is applied, for the input of new data, the Gaussian model can effectively predict the corresponding set of new input data based on the historical data distribution of the sample data. traffic data. Moreover, compared with the neural network model with a large number of parameters, the Gaussian model has a relatively small amount of calculation.

Schematically, in order to improve the accuracy of flow detection, the data type of the perfusion sample data is consistent with the data type of the perfusion data; the data type of the blood pressure sample data is consistent with the data type of the blood pressure data; the data type of the motor operation sample data is consistent with the data type of the motor operation data. The type is consistent.

To sum up, the flow rate determination method provided in this embodiment can detect the pumping flow rate of the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data corresponding to the transcatheter ventricular assist device, which greatly reduces the flow rate. The detection's dependence on the flow sensor effectively solves the technical problem of difficulty in detecting the pumping flow of the transcatheter ventricular assist device inside the heart, reduces the complexity of flow detection, and improves the efficiency of flow detection.

Moreover, since the perfusion state of the infusion fluid, the pressure state of the blood, and the operating state of the motor can all affect or reflect the pumping flow rate of the interventional catheter pump, therefore, based on the perfusion data that represents the liquid state of the infusion fluid, the pressure state of the blood represents The blood pressure data and the motor operating data representing the operating status of the motor are used to determine the flow data, which can ensure the accuracy of flow detection.

In addition, traffic data is detected through a traffic detection model. The traffic detection model is a machine learning model trained based on sample data. The machine learning model can learn more accurate correlations between data, so the accuracy of the traffic detection results can be guaranteed. accuracy.

Optionally, based on the above embodiment, the perfusion data includes at least one perfusion parameter data, and the blood pressure data includes at least one blood pressure parameter data. Hereinafter, at least one perfusion parameter data includes perfusion fluid pressure data, and at least one blood pressure parameter. The data includes arterial pressure data as an example for illustration.

In this case, in step 302, flow detection processing is performed based on the perfusion data, blood pressure data and motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device, including: based on the perfusion fluid pressure data, arterial pressure data and motor operation data Perform traffic detection and processing to obtain traffic data.

In a possible implementation, there is no pressure sensor installed on the catheter of the transcatheter ventricular assist device, and the ventricular pressure data is difficult to obtain, while the arterial pressure data can be detected and obtained in vitro. In this case, the data can be measured only based on the easily detected arteries. Pressure data, perfusion fluid pressure data and motor operation data are used for flow detection, and the flow rate of the interventional catheter pump can be accurately output, reducing the difficulty of flow detection.

The above-mentioned arterial pressure data may be aortic pressure (when left ventricular assists), pulmonary artery pressure data (when right ventricular assists), or arterial pressure data at other locations. The embodiments of the present application do not limit this.

Among them, the motor operating data includes but is not limited to motor speed data and motor current data. When the motor operating data includes motor speed data and motor current data, flow detection and processing are performed based on the perfusate pressure data, arterial pressure data and motor operating data to obtain the flow data, including: based on the perfusate pressure data, arterial pressure data, The motor speed data and motor current data are subjected to flow detection and processing to obtain flow data.

There can be a time sequence correspondence between the above motor current data and motor speed data, and the motor current data can reflect the power consumption of the motor under the corresponding motor speed data, thus reflecting the operating energy efficiency of the motor under the current liquid type. Therefore, using specific motor current data and motor speed data as motor operating data can effectively characterize the operating energy efficiency of the current drive motor, thereby ensuring the accuracy of flow detection.

Correspondingly, if the flow detection processing method is based on a preset flow detection model, then the perfusion data, blood pressure data and motor operation data are input into the preset flow detection model for flow detection processing, and the flow data is output, including: The perfusate pressure data, arterial pressure data, motor speed data and motor current data are input into the preset Gaussian model for flow detection and processing, and the flow data is output.

Optionally, the preset flow detection model includes a preset Gaussian model. The preset Gaussian model is a Gaussian model obtained after training based on multiple sets of sample data. Each set of sample data includes perfusion fluid pressure sample data, arterial pressure sample data, Motor speed sample data and motor current sample data.

For the Gaussian model, during training, the perfusate pressure sample data, arterial pressure sample data, motor speed sample data, motor current sample data and flow measurement data in each group of sample data will be used to determine the corresponding data of each group of sample data. The mean vector and covariance matrix of . The covariance matrix can effectively represent the correlation between each variable. Therefore, when the model is applied, for the input of new data, the Gaussian model can effectively predict the corresponding set of new input data based on the historical data distribution of the sample data. traffic data. Moreover, compared with neural network models with large number of parameters, Gaussian models require relatively less calculations and are more accurate. Moreover, each parameter data in these training sample sets also conforms to Gaussian distribution, which effectively improves the accuracy of traffic detection. sex.

In other embodiments, the preset traffic detection model may also include other mathematical models such as neural network models. This embodiment does not limit the type of traffic detection model. Since the sample data of the Gaussian model obeys the normal distribution and has statistical significance, this embodiment uses the preset flow detection model including the preset Gaussian model as an example for explanation.

In other embodiments, the motor operating data may also include motor current data, but not motor speed data; in this case, the sample data also includes motor current sample data, but not motor speed sample data. Alternatively, the motor operation data may also include motor speed data, but not motor current data. In this case, the sample data includes motor speed sample data, but not motor current sample data. This embodiment does not limit the implementation of the motor operation data. .

Optionally, in addition to perfusion fluid pressure data, at least one perfusion parameter data may also include perfusion fluid flow data and/or perfusion pump rotational speed data, etc. In this case, the perfusion fluid flow data and perfusion pump rotation speed data need to be combined when performing flow detection processing. /or perfusion pump speed data processing.

Optionally, in addition to arterial pressure data, at least one blood pressure parameter data may also include ventricular pressure data and/or pressure difference data, etc. In this case, the ventricular pressure data and/or pressure difference need to be combined when performing flow detection processing. data processing.

Since perfusate pressure data and arterial pressure data have a greater impact on the pumping flow rate of the interventional catheter pump, in this embodiment, at least one blood pressure is set by setting at least one perfusion parameter data including perfusate pressure data. Parametric data includes arterial pressure data, which ensures the accuracy of flow detection.

In other embodiments, the at least one perfusion parameter data may include perfusate flow data and/or perfusion pump speed data but not perfusate pressure data; and/or the at least one blood pressure parameter data may include ventricular pressure data and/or or pressure difference data, but does not include arterial pressure data. This embodiment does not limit the implementation manner of perfusion parameter data and blood pressure parameter data.

Optionally, based on the above embodiment, since the perfusion data may be a combination of at least two perfusion parameter data, and the flow detection models corresponding to different combinations need to be trained using the sample data of the corresponding combinations, therefore, the flow detection models corresponding to different combinations are There may be differences. Similarly, there may be different flow detection models corresponding to different combinations of at least two blood pressure parameter data of blood pressure data. Based on this, before step 302, it also includes:

The first flow detection model corresponding to the transcatheter ventricular assist device is determined based on the data type of the perfusion data and/or the data type of the blood pressure data; the data type of the perfusion data is used to indicate the type of perfusion parameter data included in the perfusion data. ; The data type of the blood pressure data is used to indicate the type of blood pressure parameter data included in the blood pressure data.

Correspondingly, the perfusion data, the blood pressure data and the motor operation data are input into a preset flow detection model for flow detection processing, and the flow data is output, including:

The perfusion data, the blood pressure data and the motor operation data are input into the first flow detection model, and the flow data is output.

Among them, the preset flow detection model includes flow detection models corresponding to different data types. The data type of the perfusion sample data in each set of sample data corresponding to the first flow detection model is consistent with the data type of the currently acquired perfusion data, blood pressure The data type of the sample data is consistent with the data type of the currently acquired perfusion data.

For example: the data type of the currently acquired perfusion data is: the perfusion data includes perfusion fluid pressure data and perfusion fluid flow data; the data type of the blood pressure data is: the blood pressure data includes arterial pressure data and ventricular pressure data, then the first flow detection model The perfusion sample data in the corresponding sample data includes perfusate pressure sample data and perfusate flow rate sample data, and the blood pressure sample data in the sample data includes arterial pressure sample data and ventricular pressure sample data.

For another example: the data type of the currently obtained perfusion data is: the perfusion data includes perfusion fluid pressure data, and the data type of the blood pressure data is: the blood pressure data includes arterial pressure data, then the perfusion sample in the sample data corresponding to the first flow detection model The data includes perfusion fluid pressure sample data, where the blood pressure sample data includes arterial pressure sample data.

Optionally, the data type may be received through a human-computer interaction interface; or, the data amounts of perfusion data and blood pressure data of different data types are different. In this case, the data type may also be determined based on the data amount. This embodiment does not specify the data type. The acquisition method is limited.

In this embodiment, by setting different flow detection models for perfusion data of different data types and/or blood pressure data of different data types, an adapted first flow detection model can be determined based on the currently collected data type for flow detection processing. , which can improve the accuracy of traffic detection.

Optionally, based on the above embodiment, since there may be multiple collection locations for each type of perfusion parameter data, and there may be multiple collection locations for each type of blood pressure parameter data, and although the data collected at different collection locations can represent the same kind of data, However, the specific values corresponding to the collected values may be different. Based on this, data collected at different collection locations can correspond to different traffic detection models.

At this time, before step 302, the method further includes: obtaining the first collection position of the perfusion data and the second collection position of the blood pressure data; and determining the second flow rate based on the first collection position and the second collection position. Detection model.

Among them, the preset flow detection model includes flow detection models corresponding to different collection locations. Each set of sample data corresponding to the second flow detection model includes: perfusion sample data collected based on the first collection location, perfusion sample data collected based on the second collection location blood pressure sample data and motor operation sample data.

Correspondingly, the perfusion data, the blood pressure data and the motor operation data are input into a preset flow detection model for flow detection processing, and the flow data is output, including:

The perfusion data, the blood pressure data and the motor operation data are input into the second flow detection model to perform flow detection processing, and the flow data is output.

In this embodiment, by setting flow detection models corresponding to different collection locations, an adapted second flow detection model can be determined based on the currently acquired first collection location of perfusion data and the second collection location of blood pressure data for flow detection. Processing can improve the accuracy of traffic detection.

Based on the above embodiments, the preset traffic detection model in this application is obtained by training the preset machine learning model through the sample data set. The specific training process is introduced below.

Figure 4 is a flow chart of a training method for a traffic detection model provided by an embodiment of the present application. The method at least includes the following steps:

Step 401: Obtain a sample data set collected during the operation of the transcatheter ventricular assist device.

Optionally, each set of sample data in the sample data set includes perfusion sample data, blood pressure sample data, motor operation sample data and corresponding flow measurement data corresponding to the transcatheter ventricular assist device.

Perfusion sample data is used to characterize the perfusion status of the perfusate in the perfusion component at the time the sample data set was collected. In a transcatheter ventricular assist device, the perfusion fluid enters the heart through the internal gap of the catheter, and the perfusion channel (formed by the perfusion pipeline and the catheter) communicates with the blood vessel or the interior of the heart, that is, with the blood flow channel, that is, the flow of the perfusion fluid There is a correlation with the flow of blood. Therefore, flow data can be determined based on perfusion data by learning the correlation between perfusion sample data and blood flow.

Optionally, the perfusate sample data includes but is not limited to at least one of: perfusate pressure data, perfusate flow rate data, and perfusion pump rotational speed data. For a detailed description of each type of perfusate sample data, see Perfusate Data The description of this embodiment will not be repeated here.

The blood pressure sample data is used to characterize the pressure state of the blood in the operating environment of the transcatheter ventricular assist device at the time the sample data set was collected. There is a strong correlation between blood flow and blood pressure. Therefore, flow data can be determined based on blood pressure sample data by learning the strong correlation between blood pressure sample data and blood flow.

Optionally, the blood pressure sample data includes, but is not limited to, at least one of arterial pressure sample data, ventricular pressure sample data, and pumping pressure difference sample data between the arterial pressure sample data and the ventricular pressure sample data, each type. For a detailed description of the blood pressure sample data, please refer to the description of the blood pressure data, which will not be described again in this embodiment.

Motor operating sample data is used to characterize the corresponding operating efficiency of the motor in the transcatheter ventricular assist device at the time the sample data set was collected. The type of fluid (such as viscosity) that is pumped through the transcatheter ventricular assist device affects the energy efficiency of the motor at the set speed. At the set speed, the power consumed by the motor for pumping different types of liquids is also different. It can be seen that there is a correlation between the motor operation sample data and the type of liquid pumped by the blood pump, and there is also a correlation between the flow rate and the type of liquid. At the same time, the motor is also the power output source that drives the impeller to rotate, so it can be learned how to operate the motor The correlation between sample data and blood flow enables flow data to be determined based on motor operating sample data.

Optionally, the motor operation sample data includes but is not limited to at least one of motor speed sample data and motor current sample data. For relevant descriptions of each type of motor operation sample data, see the description of the motor operation data. This embodiment is in This will not be described again.

The above-mentioned perfusate sample data may be collected by the first detection unit, the blood pressure sample data may be collected by the second detection unit, and the motor operation sample data may be collected by the third detection unit; or, the perfusate sample data, blood pressure sample data At least one of the motor operation sample data may also be collected by other devices. This embodiment does not limit the data collection method in the sample data set.

When collecting the sample data set, the operating environment of the transcatheter ventricular assist device can be a real ventricle, or it can be a simulated ventricle to obtain a simulated environment. This embodiment does not limit the collection environment of the sample data set.

The perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data in each set of sample data correspond to each other. Schematically, the perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data in each set of sample data are collected simultaneously; or, the maximum collection time interval between each data in a set of sample data is less than the threshold.

In a possible implementation, the above flow measurement data can be detected by a flow sensor.

In another possible implementation, the flow measurement data is determined by the change in liquid weight per unit time. For example, flow measurement data can be obtained by measuring changes in the weight of a liquid using a balance. This can eliminate the impact of the measurement error of the flow sensor on the accuracy of the flow measurement data. The detection accuracy of the flow detection model trained using this flow measurement data can get rid of the dependence on the accuracy of the flow sensor. The model accuracy can exceed the measurement accuracy of the flow sensor, thereby improving Determine the accuracy of traffic data.

In one example, the transcatheter ventricular assist device includes an adjustment device for adjusting the pumping flow rate or rotational speed of the interventional catheter pump. Accordingly, a sample data set collected during operation of the transcatheter ventricular assist device is obtained, including:

Control the regulating device to work in a first state and obtain at least one set of sample data in the first state; adjust the working state of the regulating device from the first state to a second state and obtain at least one set of sample data in the second state; Among them, the pumping speed corresponding to the second state is greater than the pumping speed corresponding to the first state, or the pumping speed corresponding to the second state is less than the pumping speed of the first state; determine whether the number of adjustments of the working state reaches the preset number; When the preset number of times has not been reached, the second state is regarded as the first state, the working state of the adjusting device is adjusted from the first state to the second state, and at least one set of sample data in the second state is obtained. Steps; stop until the number of adjustments reaches the preset number, or when the adjustment reaches the maximum or minimum value, to obtain a sample data set.

At this time, when collecting sample data, by operating the adjustment device to collect sample data at different pumping speeds, the richness of the sample data set can be ensured, thereby improving the performance of the trained model.

In other embodiments, when collecting sample data, the working state of the adjustment device can also be randomly set. This embodiment does not limit the method of collecting sample data.

Optionally, obtaining a sample data set collected during operation of the transcatheter ventricular assist device includes: obtaining sample data of different models of transcatheter ventricular assist devices to obtain a sample data set.

At this time, when collecting sample data, by collecting sample data of different models of transcatheter ventricular assist devices, the sample data of each type of transcatheter ventricular assist device can be used to train the flow detection model of the corresponding model, which can improve flow detection. accuracy.

In some embodiments, the sample data set has no constraints on liquid type. For example, the sample data set does not set viscosity range constraints for liquid viscosity, and the liquid viscosity corresponding to each set of sample data in the sample data set is not restricted by the preset viscosity range. Since there are data samples corresponding to various liquid types in the sample data set, the flow detection model trained based on these data is a universal flow detection model for each liquid type and is an independent and single model. Step 402: Perform model training on a preset machine learning model based on perfusion sample data, blood pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model.

Wherein, the flow detection model is used to detect the flow data corresponding to the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data generated during the application of the transcatheter ventricular assist device.

In one example, the sample data u includes multiple sets of training data and multiple sets of test data. The preset machine learning model is model trained based on perfusion sample data, blood pressure sample data, motor operation sample data, and flow measurement data to obtain flow detection. The model includes: modeling processing based on perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data in multiple sets of training data to generate a first machine learning model; combining the perfusion sample data in each set of test data, Blood pressure sample data and motor operation sample data are input into the first machine learning model for flow detection processing, and the flow detection data corresponding to each set of test data is output; the flow measurement data in each set of test data and the flow detection data corresponding to each set of test data are Compare and determine the loss information corresponding to the first machine learning model; adjust the model parameters of the first machine learning model based on the loss information to obtain a traffic detection model.

Optionally, the first machine learning model includes a Gaussian model. In a possible implementation, the above-mentioned first machine learning model is a Gaussian model modeled based on the first sample data set; correspondingly, the traffic detection model is a parameter-adjusted Gaussian model. Since the sample data of the Gaussian model obeys the normal distribution, it is statistically significant and can learn the correlation between various parameters. Therefore, the Gaussian model can be selected for traffic detection.

Optionally, compare the traffic measurement data in each set of test data with the traffic detection data corresponding to each set of test data, and determine the loss information corresponding to the first machine learning model, including: based on the traffic measurement data in each set of sample data , determine the average flow data; compare the average flow data, the flow detection data corresponding to each set of test data, and the flow measurement data in each set of test data to obtain the variance data and root mean square error data corresponding to the Gaussian model. At this time, the loss information includes variance data and root mean square error data.

The calculation formula of variance data r ² can be expressed by the following formula:

Among them, n represents the total number of sample groups in the test data; i represents the i-th group of test data, and i is an integer from 1 to n. Represents the flow measurement data in the i-th group of test data;/> represents the average traffic data of the traffic detection data corresponding to the n group of test data; _Xi represents the traffic detection data corresponding to the i-th group of test data.

The calculation formula of root mean square error data RMSE can be expressed by the following formula:

Among them, n represents the total number of sample groups in the test data; i represents the i-th group of test data, and i is an integer from 1 to n. represents the flow measurement data in the i-th group of test data; _Xi represents the flow detection data corresponding to the i-th group of test data.

Optionally, when the variance and the root mean square error are used to determine the loss information at the same time, the variance and the root mean square error also have corresponding weight parameters. At this time, the weighted sum of the variance and the root mean square error is determined to obtain the loss information. . Alternatively, the sum of the variance and the root mean square error may also be determined as the loss information. This embodiment does not limit the calculation method of the loss information.

During the training process of the Gaussian model, the Gaussian model corresponds to multiple sets of model parameters, and the loss data corresponding to each set of parameters is determined respectively, so as to determine the target with the highest traffic detection accuracy (that is, the smallest loss data) based on the loss data corresponding to each set of model parameters. Model parameters, configure the Gaussian model according to the target model parameters, and obtain the above traffic detection model.

In another possible implementation, the machine learning model is a neural network model. The above-mentioned adjustment of the model parameters of the first machine learning model based on the loss information to obtain the traffic detection model includes: determining whether the loss value represented by the loss information is greater than the loss threshold, and whether the number of adjustments to the model parameters is less than or equal to the number threshold; if the loss value is greater than If the loss threshold or the number of adjustments is less than the number threshold, the model parameters of the first machine learning model are adjusted, and the execution is triggered to input the pump load sample data and motor operation sample data in each set of test data into the first machine learning model for flow analysis. Detection processing, outputting the flow detection data corresponding to each group of test data; comparing the flow measurement data in each group of test data with the flow detection data corresponding to each group of test data, and determining the loss information corresponding to the first machine learning model; If the loss value is less than or equal to the loss threshold, or the number of adjustments is greater than or equal to the number threshold, the adjusted model parameters are output to obtain the traffic detection model.

In this embodiment, a flow detection model is obtained by pre-training using a sample data set, and the flow detection model is used to determine the flow data of the transcatheter ventricular assist device, which greatly reduces the dependence of flow detection on the flow sensor and effectively solves the problem of The technical problem of difficult to detect the pumping flow of the catheter ventricular assist device inside the heart reduces the complexity of flow detection and improves the efficiency of flow detection.

Moreover, since the perfusion state of the infusion fluid, the pressure state of the blood, and the operating state of the motor can all affect or reflect the pumping flow rate of the interventional catheter pump, therefore, based on the perfusion sample data that represents the liquid state of the infusion fluid, the pressure that represents the blood The blood pressure sample data of the state and the motor operation sample data representing the operating state of the motor are trained to obtain the flow detection model, which can ensure the accuracy of the flow detection model in determining the flow data.

In addition, the traffic data is detected through the traffic detection model. The traffic detection model is a machine learning model trained based on sample data. The machine learning model can learn more accurate correspondences between data, so the accuracy of the traffic detection results can be guaranteed. accuracy.

In addition, by implementing the machine learning model as a Gaussian model, traffic detection results that conform to the normal distribution can be obtained, improving the accuracy of traffic detection; and compared with neural network models with a large number of parameters, the calculation amount of the Gaussian model is relatively small , Therefore, the calculation efficiency of traffic detection can be improved and computing resources can be saved.

Based on the above embodiments, if the perfusion data includes at least one perfusion parameter data, the at least one perfusion parameter data includes perfusion fluid pressure data; the blood pressure data includes at least one blood pressure parameter data, and the at least one blood pressure parameter data includes arterial pressure data. , then the perfusion sample data in the sample data set includes perfusion fluid pressure sample data corresponding to the transcatheter ventricular assist device, and the blood pressure sample data in the sample data set includes arterial pressure sample data corresponding to the transcatheter ventricular assist device.

Accordingly, model training is performed on the preset machine learning model based on perfusion sample data, blood pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model, including:

Perform model training on the preset machine learning model based on perfusion fluid pressure sample data, arterial pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model.

In a possible implementation, there is no pressure sensor installed on the catheter of the transcatheter ventricular assist device, so ventricular pressure data is difficult to obtain, while arterial pressure sample data can be obtained outside the body. In this case, it can be based only on easily detected arteries. Pressure sample data, perfusion fluid pressure sample data and motor operation sample data train the flow detection model, and can also accurately output the flow rate of the interventional catheter pump, reducing the difficulty of flow detection.

The above-mentioned arterial pressure sample data has the same data type as the arterial pressure data. The arterial pressure sample data can be aortic pressure (when left ventricular assist), pulmonary artery pressure data (when right ventricular assist), or arterial pressure at other locations. sample. The embodiments of the present application do not limit this.

In the above embodiment, the motor operating data includes at least one of motor rotation speed data and motor current data. In the case where the motor operation data includes motor speed data and motor current data, the motor operation sample data in the sample data set includes motor speed sample data and motor current sample data corresponding to the transcatheter ventricular assist device;

In this case, perform model training on the preset machine learning model based on perfusion sample data, blood pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model, including: based on perfusion fluid pressure sample data, arterial pressure sample Data, motor speed sample data, motor current sample data, and flow measurement data are used to train the preset machine learning model to obtain a flow detection model.

Accordingly, the flow detection model is specifically used to detect flow data based on perfusate pressure sample data, arterial pressure sample data, motor speed data and motor current data generated during application of the transcatheter ventricular assist device.

There can be a time series correspondence between the above motor current sample data and motor speed sample data. The motor current sample data can reflect the power consumption of the motor under the corresponding motor speed sample data, thus reflecting the operating energy efficiency of the motor under the current liquid type. . Therefore, using specific motor current sample data and motor speed sample data as motor operating data can effectively characterize the operating energy efficiency of the current drive motor, thereby ensuring the accuracy of the flow data determined by the trained flow detection model.

Since there is a correlation between the perfusate pressure data and the arterial pressure data and the pumping flow of the interventional catheter pump, in this embodiment, the flow detection model is obtained by training based on the perfusate pressure sample data and the arterial pressure sample data, It can ensure the accuracy of traffic data detected by the traffic detection model.

Optionally, in addition to perfusate pressure data, at least one perfusion parameter data may also include perfusate flow data and/or perfusion pump rotational speed data, etc. In this case, the perfusion sample data also includes perfusate flow sample data and/or Perfusion pump speed sample data.

Optionally, in addition to arterial pressure data, at least one blood pressure parameter data may also include ventricular pressure data and/or pressure difference data, etc. In this case, the blood pressure sample data also includes ventricular pressure sample data and/or pressure difference sample data. .

In other embodiments, at least one perfusion parameter data may include perfusate flow rate data and/or perfusion pump rotational speed data, but not perfusion fluid pressure data; in this case, the perfusion sample data includes perfusate flow rate sample data and/or perfusion pump speed data. Pump speed sample data, but not perfusion fluid pressure sample data. And/or, at least one blood pressure parameter data may include ventricular pressure data and/or pressure difference data without including arterial pressure data; in this case, the perfusion sample data includes ventricular pressure sample data and/or pressure difference sample data without Including arterial pressure sample data, this embodiment does not limit the implementation of perfusion sample data and blood pressure sample data.

Optionally, based on the above embodiment, since the perfusion data may be a combination of at least two perfusion parameter data, the blood pressure data may be a combination of at least two blood pressure parameter data. Different combinations constitute different data types, and different data types correspond to flow rates. Detection models may differ. Therefore, it is necessary to train traffic detection models corresponding to different data types.

At this time, obtaining the sample data set collected during the operation of the transcatheter ventricular assist device includes: obtaining perfusion sample data of different data types and blood pressure sample data of different data types.

Accordingly, model training is performed on the preset machine learning model based on perfusion sample data, blood pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model, including: perfusion sample data based on the same data type, the same data type Perform model training on the preset machine learning model using the blood pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model corresponding to the data type of the perfusion sample data and the data type of the blood pressure sample data.

In this embodiment, by training the corresponding flow detection model based on the perfusion sample data of each data type and/or the blood pressure sample data of each data type, the transcatheter ventricular assist device can determine the adaptation based on the currently collected data type. The first traffic detection model is used for traffic detection processing, which can improve the accuracy of traffic detection.

Optionally, based on the above embodiment, data collected at different collection locations may correspond to different traffic detection models. Therefore, it is necessary to train traffic detection models corresponding to different collection locations.

At this time, obtaining the sample data set collected during the operation of the transcatheter ventricular assist device includes: obtaining perfusion sample data collected at different first collection positions, and blood pressure sample data collected at different second collection positions.

Accordingly, model training is performed on the preset machine learning model based on perfusion sample data, blood pressure sample data, motor operation sample data, and flow measurement data to obtain a flow detection model, including:

Carry out model training on the preset machine learning model based on the perfusion sample data collected at the same first collection position, the blood pressure sample data collected at the same second collection position, motor operation sample data and flow measurement data, and obtain the first collection position and The traffic detection model corresponding to the second collection location.

In this embodiment, the corresponding flow detection model is obtained by training based on the perfusion sample data collected at each first collection position and the blood pressure sample data collected at each second collection position, so that the transcatheter ventricular assist device can be based on the currently acquired data. The first acquisition position of the perfusion data and the second acquisition position of the blood pressure data determine an adapted second flow detection model for flow detection processing to improve the accuracy of flow detection.

Figure 5 is a block diagram of a flow determination device provided by an embodiment of the present application, and the flow determination device is applied to a transcatheter ventricular assist device. The device includes at least the following modules: data acquisition module 510 and flow determination module 520.

The data acquisition module 510 is used to acquire perfusion data, blood pressure data corresponding to the transcatheter ventricular assist device, and motor operation data corresponding to the transcatheter ventricular assist device;

The flow determination module 520 is configured to perform flow detection processing based on the perfusion data, the blood pressure data, and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device.

Optionally, the perfusion data includes at least one perfusion parameter data, the at least one perfusion parameter data includes at least one of perfusion fluid pressure data, perfusion fluid flow data and perfusion pump rotational speed data, and the blood pressure data includes at least A kind of blood pressure parameter data, the at least one kind of blood pressure parameter data includes at least one of arterial pressure data, ventricular pressure data, and pumping pressure difference data.

Optionally, the traffic determination module 520 is used to:

The flow rate data is obtained by performing flow detection processing according to the perfusion fluid pressure data, the arterial pressure data and the motor operation data.

Optionally, the motor operating data includes motor speed data and motor current data, and the flow determination module 520 is used to:

The flow rate data is obtained by performing flow detection processing based on the perfusion fluid pressure data, the arterial pressure data, the motor speed data and the motor current data.

Optionally, the traffic determination module 520 includes a determination sub-module 521.

The determination sub-module 521 is used to input the perfusion data, the blood pressure data and the motor operation data into a preset flow detection model for flow detection processing, and output the flow data;

Wherein, the flow detection model is a machine learning model obtained after training based on multiple sets of sample data. Each set of sample data includes corresponding perfusion sample data, blood pressure sample data, motor operation sample data and sample flow data.

Optionally, the determination sub-module 521 is used to:

Input the perfusate pressure data, arterial pressure data, motor speed data and motor current data into the preset Gaussian model for flow detection processing, and output the flow data;

Wherein, the perfusion data includes the perfusion fluid pressure data, the blood pressure data includes the arterial pressure data, the motor operation data includes the motor speed data and the motor current data, and the preset flow detection The model includes the preset Gaussian model, which is a Gaussian model obtained after training based on multiple sets of sample data.

For relevant details, refer to the above method embodiments.

It should be noted that when the flow rate determination device provided in the above embodiment determines the flow rate, the division of the above functional modules is only used as an example. In actual applications, the above function allocation can be completed by different functional modules as needed. , that is, dividing the internal structure of the flow determination device into different functional modules to complete all or part of the functions described above. In addition, the flow rate determination device provided in the above embodiments and the flow rate determination method embodiments belong to the same concept. Please refer to the method embodiments for the specific implementation process, which will not be described again here.

Figure 6 is a block diagram of a training device for a traffic detection model provided by an embodiment of the present application. The device includes at least the following modules: data acquisition module 610 and model training module 620.

The data acquisition module 610 is used to acquire a sample data set collected during the operation of the transcatheter ventricular assist device. Each set of sample data in the sample data set includes perfusion sample data, blood pressure sample data corresponding to the transcatheter ventricular assist device, Motor operating sample data and corresponding flow measurement data;

The model training module 620 is used to perform model training on a preset machine learning model based on the perfusion sample data, the blood pressure sample data, the motor operation sample data, and the flow measurement data to obtain a flow detection model;

Wherein, the flow detection model is used to detect the flow data corresponding to the transcatheter ventricular assist device based on the perfusion data, blood pressure data and motor operation data generated during the application of the transcatheter ventricular assist device.

Optionally, the sample data set includes multiple sets of training data and multiple sets of test data. The model training module 620 includes: a model generation sub-module 621, a traffic detection sub-module 622, a loss calculation sub-module 623 and a parameter adjustment sub-module. 624.

The model generation sub-module 621 is used to perform modeling processing based on the perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data in the multiple sets of training data, and generate a first machine learning model;

The flow detection sub-module 622 is used to input the perfusion sample data, blood pressure sample data, and motor operation sample data in each set of test data into the first machine learning model for flow detection processing, and output the flow rate corresponding to each set of test data. Test data;

The loss calculation sub-module 623 is used to compare the flow measurement data in each set of test data with the flow detection data corresponding to each set of test data, and determine the loss information corresponding to the first machine learning model;

Parameter adjustment sub-module 624 is used to adjust the model parameters of the first machine learning model based on the loss information to obtain the traffic detection model.

Optionally, the first machine learning model includes a Gaussian model, and the loss calculation submodule 623 is used for:

Determine the average flow data based on the flow measurement data in each group of sample data;

Compare the average flow data, the flow detection data corresponding to each set of test data, and the flow measurement data in each set of test data to obtain the variance data and root mean square error data corresponding to the Gaussian model. The loss information includes The variance data and the root mean square error data.

Optionally, the flow measurement data is determined by the change in liquid weight within a unit time period.

Optionally, the perfusion sample data in the sample data set includes perfusion fluid pressure sample data corresponding to the transcatheter ventricular assist device, and the blood pressure sample data in the sample data set includes arterial pressure corresponding to the transcatheter ventricular assist device. Sample data, the motor operation sample data in the sample data set includes motor speed sample data and motor current sample data corresponding to the transcatheter ventricular assist device;

Correspondingly, the flow detection model is specifically used to detect the flow data based on the perfusate pressure data, arterial pressure data, motor speed data and motor current data generated during the application of the transcatheter ventricular assist device.

For relevant details, refer to the above method embodiments.

It should be noted that when the training device for the traffic detection model provided in the above embodiments trains the traffic detection model, only the division of the above functional modules is used as an example. In actual applications, the above functions can be allocated as needed. It is completed by different functional modules, that is, the internal structure of the training device of the traffic detection model is divided into different functional modules to complete all or part of the functions described above. In addition, the training device of the traffic detection model provided in the above embodiments and the training method embodiment of the traffic detection model belong to the same concept. Please refer to the method embodiments for the specific implementation process, which will not be described again here.

Figure 7 is a block diagram of a computer device provided by an embodiment of the present application. The computer device at least includes the control unit in the transcatheter ventricular assist device shown in Figure 1. In other embodiments, it may also be other devices that are communicatively connected to the transcatheter ventricular assist device, such as computers, tablets, and mobile phones. and other computer equipment with processing capabilities. The computer device includes at least a processor 701 and a memory 702 .

The processor 701 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 701 can adopt at least one hardware form among DSP (Digital Signal Processing, digital signal processing), FPGA (Field-Programmable Gate Array, field programmable gate array), and PLA (Programmable Logic Array, programmable logic array). accomplish. The processor 701 may also include a main processor and a co-processor. The main processor is a processor used to process data in the wake-up state, also called CPU (Central Processing Unit, central processing unit); the co-processor is A low-power processor used to process data in standby mode. In some embodiments, the processor 701 may be integrated with a GPU (Graphics Processing Unit, image processor), and the GPU is responsible for rendering and drawing content to be displayed on the display screen. In some embodiments, the processor 701 may also include an AI (Artificial Intelligence, artificial intelligence) processor, which is used to process computing operations related to machine learning.

Memory 702 may include one or more computer-readable storage media, which may be non-transitory. Memory 702 may also include high-speed random access memory, and non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 702 is used to store at least one instruction, and the at least one instruction is used to be executed by the processor 701 to implement the traffic determination provided by the method embodiments in this application. Method or training method for the traffic detection model.

In some embodiments, the computer device optionally further includes: a peripheral device interface and at least one peripheral device. The processor 701, the memory 702 and the peripheral device interface may be connected through a bus or a signal line. Each peripheral device can be connected to the peripheral device interface through a bus, a signal line or a circuit board. Illustratively, peripheral devices include but are not limited to: radio frequency circuits, touch display screens, audio circuits, power supplies, etc.

Of course, the computer device may also include fewer or more components, which is not limited in this embodiment.

Optionally, this application also provides a computer-readable storage medium in which a program is stored, and the program is loaded and executed by a processor to implement the flow determination method or flow detection of the above method embodiment. Model training method.

Optionally, this application also provides a computer product. The computer product includes a computer-readable storage medium. A program is stored in the computer-readable storage medium. The program is loaded and executed by a processor to implement the above method embodiments. The flow determination method or the training method of the flow detection model.

The technical features of the above-described embodiments can be combined in any way. To simplify the description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, All should be considered to be within the scope of this manual.

The above-described embodiments only express several implementation modes of the present application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all fall within the protection scope of the present application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims (16)

1. A method of flow determination, the method being applied to a transcatheter ventricular assist device, the method comprising:

acquiring perfusion data and blood pressure data corresponding to the transcatheter ventricular assist device and motor operation data corresponding to the transcatheter ventricular assist device;

and carrying out flow detection processing according to the perfusion data, the blood pressure data and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device.

2. The method of claim 1, wherein the perfusion data comprises at least one perfusion parameter data comprising at least one of perfusion fluid pressure data, perfusion fluid flow data, and perfusion pump rotational speed data, the blood pressure data comprising at least one blood pressure parameter data comprising at least one of arterial pressure data, ventricular pressure data, and pumping differential pressure data.

3. The method according to claim 2, wherein the performing flow detection processing according to the perfusion data, the blood pressure data, and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device includes:

And carrying out flow detection processing according to the perfusion pressure data, the arterial pressure data and the motor operation data to obtain the flow data.

4. The method of claim 2, wherein the motor operation data includes motor rotation speed data and motor current data, and wherein the performing a flow detection process according to the perfusion pressure data, the arterial pressure data, and the motor operation data to obtain the flow data includes:

and carrying out flow detection processing according to the perfusion pressure data, the arterial pressure data, the motor rotating speed data and the motor current data to obtain the flow data.

5. The method of any one of claims 2 to 4, wherein the transcatheter ventricular assist device comprises:

an interventional catheter pump comprising a catheter, a pump head connected to a distal end of the catheter, a coupling assembly connected to a proximal end of the catheter, the pump head being deliverable by the catheter to a desired location of the heart for a pumping operation;

a perfusion channel penetrating at least the catheter, an inlet of the perfusion channel being disposed on the coupling assembly, an outlet of the perfusion channel being disposed at the pump head; the perfusion fluid in the perfusion channel enters a desired position in the cardiovascular system through the outlet; the desired location in the cardiovascular system includes at least one of the aorta, pulmonary artery, left ventricle, right ventricle, left atrium, right atrium;

The perfusate pressure data characterizes a sum of perfusate pressures generated by the perfusate at a pressure acquisition location to the outlet.

6. The method of claim 1, wherein the performing flow detection processing according to the perfusion data, the blood pressure data, and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device comprises:

inputting the perfusion data, the blood pressure data and the motor operation data into a preset flow detection model to perform flow detection processing, and outputting the flow data;

the flow detection model is a machine learning model obtained after training based on a plurality of groups of sample data, and each group of sample data comprises perfusion sample data, blood pressure sample data, motor operation sample data and sample flow data which correspond to each other.

7. The method according to claim 6, wherein inputting the perfusion data, the blood pressure data, and the motor operation data into a preset flow detection model for flow detection processing, and outputting the flow data, comprises:

inputting perfusion pressure data, arterial pressure data, motor rotating speed data and motor current data into a preset Gaussian model for flow detection processing, and outputting the flow data;

The perfusion data comprise perfusion pressure data, the blood pressure data comprise arterial pressure data, the motor operation data comprise motor rotating speed data and motor current data, the preset flow detection model comprises the preset Gaussian model, and the preset Gaussian model is a Gaussian model obtained after training based on multiple groups of sample data.

8. A method of training a flow detection model, the method comprising:

acquiring a sample data set acquired in the operation process of a transcatheter ventricular assist device, wherein each group of sample data in the sample data set comprises perfusion sample data, blood pressure sample data, motor operation sample data and corresponding flow measurement data corresponding to the transcatheter ventricular assist device;

model training is carried out on a preset machine learning model according to the perfusion sample data, the blood pressure sample data, the motor operation sample data and the flow measurement data to obtain a flow detection model;

the flow detection model is used for detecting flow data corresponding to the transcatheter ventricular assist device according to perfusion data, blood pressure data and motor operation data generated in the application process of the transcatheter ventricular assist device.

9. The method according to claim 8, wherein the sample data u includes a plurality of sets of training data and a plurality of sets of test data, and the model training the preset machine learning model according to the perfusion sample data, the blood pressure sample data, the motor operation sample data and the flow measurement data to obtain a flow detection model includes:

modeling processing is carried out based on perfusion sample data, blood pressure sample data, motor operation sample data and flow measurement data in the plurality of groups of training data, so as to generate a first machine learning model;

inputting perfusion sample data, blood pressure sample data and motor operation sample data in each set of test data into the first machine learning model for flow detection processing, and outputting flow detection data corresponding to each set of test data;

comparing the flow measurement data in each group of test data with the flow detection data corresponding to each group of test data, and determining loss information corresponding to the first machine learning model;

and adjusting model parameters of the first machine learning model based on the loss information to obtain the flow detection model.

10. The method of claim 9, wherein the first machine learning model comprises a gaussian model, wherein comparing the flow measurement data in each set of test data with the flow detection data corresponding to each set of test data determines loss information corresponding to the first machine learning model, comprising:

determining average flow data according to the flow measurement data in each set of sample data;

and comparing the average flow data, the flow detection data corresponding to each group of test data and the flow measurement data in each group of test data to obtain variance data and root mean square error data corresponding to the Gaussian model, wherein the loss information comprises the variance data and the root mean square error data.

11. A method according to any one of claims 8 to 10, wherein the flow measurement data is determined by the amount of change in weight of liquid per unit time.

12. A flow determination device for use with a transcatheter ventricular assist device, the device comprising:

the data acquisition module is used for acquiring perfusion data and blood pressure data corresponding to the transcatheter ventricular assist device and motor operation data corresponding to the transcatheter ventricular assist device;

And the flow determining module is used for carrying out flow detection processing according to the perfusion data, the blood pressure data and the motor operation data to obtain flow data corresponding to the transcatheter ventricular assist device.

13. A training device for a flow detection model, the device comprising:

the data acquisition module is used for acquiring sample data u acquired in the operation process of the catheter ventricular assist device, and each group of sample data in the sample data set comprises perfusion sample data, blood pressure sample data, motor operation sample data and corresponding flow measurement data corresponding to the catheter ventricular assist device;

the model training module is used for carrying out model training on a preset machine learning model according to the perfusion sample data, the blood pressure sample data, the motor operation sample data and the flow measurement data to obtain a flow detection model;

the flow detection model is used for detecting flow data corresponding to the transcatheter ventricular assist device according to perfusion data, blood pressure data and motor operation data generated in the application process of the transcatheter ventricular assist device.

14. A transcatheter ventricular assist device, the transcatheter ventricular assist device comprising a computer device comprising a processor and a memory; the memory has stored therein a program that is loaded and executed by the processor to implement the flow rate determination method according to any one of claims 1 to 6; alternatively, a training method of the flow detection model according to any one of claims 7 to 10 is implemented.

15. A computer device, the device comprising a processor and a memory; the memory has stored therein a program that is loaded and executed by the processor to implement the flow rate determination method according to any one of claims 1 to 6; alternatively, a training method of the flow detection model according to any one of claims 7 to 10 is implemented.

16. A computer-readable storage medium, characterized in that the storage medium has stored therein a program for realizing the flow rate determination method according to any one of claims 1 to 6 when loaded and executed by a processor; alternatively, a training method of the flow detection model according to any one of claims 7 to 10 is implemented.