CN116671978A - Digital twinning-based conformal ultrasonic living body visual monitoring method and system - Google Patents

Abstract

The present disclosure provides a digital twinning-based conformal ultrasonic living body visual monitoring method and a system, which acquire a blood vessel ultrasonic image in real time, and acquire a blood flow velocity spectrum and a piezoelectric pulse wave signal; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals; establishing a standard blood circulation simulation model, establishing an individualized digital twin model, determining individualized parameters and influence factors, establishing a functional relation between measurable and individualized parameters, and realizing real-time simulation of the individualized model by adjusting the parameters in real time; setting an objective function, optimizing parameters to obtain an optimal parameter combination, and determining real-time pressure-volume loop quantized heart contractility and heart beat function end to end; extracting the obtained local characteristic region of the hemodynamic parameters, expanding single-point time to a time axis according to data characteristics by combining a clinical queue, realizing continuous visualization of the circulation state parameters and analyzing the variation trend of the parameters.

Description

Digital twinning-based conformal ultrasonic living body visual monitoring method and system

Technical Field

The disclosure relates to the technical field of ultrasonic sensing and monitoring, in particular to a digital twinning-based conformal ultrasonic living body visual sensing and monitoring method and system.

Background

The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.

Acute coronary syndrome and acute cerebrovascular disease are caused by the sudden and large reduction of blood flow of supply heart and brain, and at present, although detection means of biomarker assay, electrocardiogram, CT or MRI exist, a certain time window is needed for the abnormal occurrence of the detection results, and more importantly, the means are needed to be completed in medical institutions, so that diagnosis delay and the missing of the golden time are often caused.

On the other hand, chronic stable cardiovascular and cerebrovascular diseases are difficult to find in routine examinations due to the hidden symptoms or the symptoms only occurring in special cases of physical exertion, etc., while angiography can find vascular atherosclerosis and lumen stenosis, but has radiation, traumata and high medical and health costs. More importantly, the development of these stable lesions into acute events is not a linear process, and predicting the occurrence of an acute event has been a clinical problem.

Among them, in the clinical management of contemporary cardiovascular diseases, there are two aspects of features that are significantly different from the previous clinical medical environment. First, as emergency levels rise, acute event mortality decreases, but end-stage heart failure and disability status patients increase; secondly, the value and status of rehabilitation therapy are prominent in the final state of cardiovascular and cerebrovascular diseases. However, in rehabilitation therapy, the risk of recurrent acute cardiovascular and cerebrovascular events is high, and tight monitoring is required, which depends on professional medical environment, and the contradiction between the number of patients and the number of professional medical environments/devices is increasing.

There are reported a technique of performing blood flow velocity and cardiac output by ultrasound (patent CN 105708494B) and a technique of performing measurement by optical techniques such as photoplethysmography (cn20110329775. X, CN 114271805A), and in these patents, pulse wave waveforms are generally obtained, heart rate and diameter of a detected blood vessel are extracted to calculate blood pressure and blood pressure change rate, and then converted into stroke volume and cardiac output. Other technical solutions have also been proposed by scholars, such as: CN 114287967A proposes to measure pressure by using ultrasound and microbubbles, CN 215839164U synthesizes the blood vessel wall elasticity information, and CN 208910307U adopts two groups of ultrasound transducers to obtain blood flow impedance index, ankle brachial index and other blood vessel elasticity information. To make the detection more convenient, a series of wearable devices have been developed, such as CN 211094256U, CN 112842392A, CN 113440165A, CN 108697349B.

It is well known that a complete assessment of circulatory and blood flow conditions involves dimensional information of morphology, structure, velocity, frequency, etc., which is also information that a professional ultrasound examination needs to provide. However, the techniques described above compromise the ability to provide multi-dimensional information while achieving continuous monitoring and wear requirements. These techniques can only acquire pulse wave, or can only acquire blood flow velocity by doppler principle, or can only acquire pressure at a certain position of blood vessel, or can only acquire pulse wave conduction velocity and blood vessel wall elasticity, and in summary, the evaluation of the whole circulation state and the heart output state is based on a single dimension or simple linear regression model, and can not completely acquire blood vessel and blood flow dynamic images, blood flow real-time velocity, pulse real-time waveform and heart rate real-time information at the same time. It is well known that the cross-sectional area of an artery is square with radius, which is particularly important in medium and small caliber vascular measurements, and that a single method of estimating blood flow and blood pressure by means of the cross-sectional area of the vessel results in a large deviation of the result when there is a small error in the measurement of the radius (e.g. 6mm for the internal diameter of the common carotid artery, and 30% for the final result if there is a 1mm error in the measurement of the radius). The method for calculating the blood flow velocity by means of Doppler has strict requirements on the relative position of the probe and the blood vessel to be detected and the incidence angle of the sound beam.

In the reported technology, the wearing experience and monitoring duration (such as Kenny research) are sacrificed by increasing the volume of the wearable device to achieve more dimensional information, and the increase of the size and weight of the device undoubtedly brings additional pressure to superficial arteries, so that not only is the blood pressure measured value inaccurate, but also a geometric spectrum broadening effect (geometric spectral broadening) is generated, and the blood flow velocity measurement is wrong. The invention discloses a low-intensity focused ultrasonic acupoint therapeutic apparatus and a method for treating peripheral vascular diseases, which are disclosed in the patent (CN202210533454. X)', wherein a wearable product structure is provided with a plurality of ultrasonic transducers, so that the ultrasonic transducers are always positioned at the acupoints in the treatment process, the ultrasonic transducers are completely attached to the acupoints of a human body, ultrasonic loss is avoided, and the ultrasonic transducers cannot be used for imaging and information detection. The invention patent (CN202080067767. X) "non-invasive, real-time, beat-to-beat, dynamic blood pressure monitoring", provides a flow system for determining cardiac parameters at a fixed location within the cardiovascular system of a subject, the principle of which is that the visualization of vital signs cannot be achieved by detecting pressure waves passing through the fixed location with an ultrasound transducer. The reason for this is that there are drawbacks in the pathophysiological parameter model of transducer architecture, mode of operation, and cardiac circulatory function assessment.

Disclosure of Invention

In order to solve the problems, the disclosure provides a digital twin-based conformal ultrasound vital body visual monitoring method and a digital twin-based cardiac cycle functional hemodynamic parameter calculation, and the digital twin-based conformal ultrasound visual sensor and the digital twin-based cardiac cycle functional hemodynamic parameter calculation are utilized to input real-time hemodynamic data acquired by the digital twin-based conformal ultrasound visual body visual monitoring system into the digital twin-based conformal ultrasound visual body visual monitoring system to perform parameter calculation, so that visual sensing and visual monitoring of the conformal ultrasound vital body sign are realized.

According to some embodiments, the present disclosure employs the following technical solutions:

a digital twinning-based conformal ultrasonic living body visual monitoring method comprises the following steps:

acquiring a blood vessel ultrasonic image in real time, and acquiring a blood flow velocity spectrum and a piezoelectric pulse wave signal; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals;

establishing a standard blood circulation simulation model, establishing an individualized digital twin model, determining individualized parameters and influence factors, establishing a functional relation between measurable and individualized parameters, and realizing real-time simulation of the individualized model by adjusting the parameters in real time; setting an objective function, optimizing parameters to obtain an optimal parameter combination, and determining real-time pressure-volume loop quantized heart contractility and heart beat function end to end;

extracting the obtained local characteristic region of the hemodynamic parameters, expanding single-point time to a time axis according to data characteristics by combining a clinical queue, realizing continuous visualization of the circulation state parameters and analyzing the variation trend of the parameters.

According to some embodiments, the present disclosure employs the following technical solutions:

digital twinning-based conformal ultrasound living body visual monitoring system, comprising:

the transducer data acquisition module is used for acquiring blood vessel ultrasonic images in real time and acquiring blood flow velocity frequency spectrums and piezoelectric pulse wave signals; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals;

the data transmission and processing module is used for establishing a standard blood circulation simulation model and an individualized digital twin model, determining individualized parameters and influence factors, establishing a functional relation between measurable and individualized parameters, and realizing real-time simulation of the individualized model by adjusting the parameters in real time; setting an objective function, optimizing parameters to obtain an optimal parameter combination, and determining real-time pressure-volume loop quantized heart contractility and heart beat function end to end;

extracting the obtained local characteristic region of the hemodynamic parameters, expanding single-point time to a time axis according to data characteristics by combining a clinical queue, realizing continuous visualization of the circulation state parameters and analyzing the variation trend of the parameters.

According to some embodiments, the present disclosure employs the following technical solutions:

a computer readable storage medium having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform the digital twin based conformal ultrasound vital signs visualization monitoring method.

According to some embodiments, the present disclosure employs the following technical solutions:

a terminal device comprising a processor and a computer readable storage medium, the processor configured to implement instructions; the computer readable storage medium is for storing a plurality of instructions adapted to be loaded by a processor and to perform the digital twin based conformal ultrasound living body visualization monitoring method.

Compared with the prior art, the beneficial effects of the present disclosure are:

the present disclosure proposes a conformal ultrasound visual sensor and digital twin-based cardiac cycle functional hemodynamic parameter calculation, and utilizes real-time hemodynamic data obtained by the former to input the latter for parameter calculation, and finally realizes conformal ultrasound vital sign visual sensing and monitoring. The invention can be applied to superficial arteries such as carotid artery, radial artery, brachial artery, femoral artery and the like, and can also be applied to deep arteries such as aorta and the like; the method is not only suitable for cardiac circulatory state evaluation, but also can be used for penile artery and erectile function evaluation. The method is suitable for the categories of mobile medical equipment and individual medical equipment under the mHealth architecture.

Drawings

The accompanying drawings, which are included to provide a further understanding of the disclosure, illustrate and explain the exemplary embodiments of the disclosure and together with the description serve to explain the disclosure, and do not constitute an undue limitation on the disclosure.

FIG. 1 is a flow chart of a digital twinning-based conformal ultrasound living body visualization monitoring method in accordance with an embodiment of the present disclosure;

FIG. 2 is a graph of a hemodynamic parameter calculation model in accordance with an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of a heart model according to an embodiment of the present disclosure;

FIG. 4 is an RCR model of an embodiment of the disclosure;

FIG. 5 is a computational model of an embodiment of the present disclosure;

FIG. 6 is a pressure volume ring schematic diagram of an embodiment of the present disclosure;

FIG. 7 is a schematic diagram of a channel attention mechanism algorithm according to an embodiment of the present disclosure;

FIG. 8 is a schematic diagram of a spatial attention mechanism algorithm of the disclosed embodiments;

fig. 9 is a schematic diagram of a transducer structure according to an embodiment of the present disclosure.

The specific embodiment is as follows:

the disclosure is further described below with reference to the drawings and examples.

It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments in accordance with the present disclosure. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.

Example 1

In one embodiment of the present disclosure, a digital twinning-based conformal ultrasound living body visualization method is provided, including:

step 1: acquiring a blood vessel ultrasonic image in real time, and acquiring a blood flow velocity spectrum and a piezoelectric pulse wave signal; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals;

step 2: establishing a standard blood circulation simulation model, establishing an individualized digital twin model, determining individualized parameters and influence factors, establishing a functional relation between measurable and individualized parameters, and realizing real-time simulation of the individualized model by adjusting the parameters in real time; setting an objective function, optimizing parameters to obtain an optimal parameter combination, and determining real-time pressure-volume loop quantized heart contractility and heart beat function end to end;

step 3: extracting the obtained local characteristic region of the hemodynamic parameters, expanding single-point time to a time axis according to data characteristics by combining a clinical queue, realizing continuous visualization of the circulation state parameters and analyzing the variation trend of the parameters.

As an embodiment, the specific implementation process of the digital twin-based conformal ultrasound vital sign visualization sensing and monitoring method provided by the present disclosure includes, as shown in fig. 1:

1. acquisition parameter part

Acquiring a blood vessel ultrasonic image in real time, and acquiring a blood flow velocity spectrum and a piezoelectric pulse wave signal; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals;

specifically, the blood vessel ultrasonic image is displayed in real time, the blood vessel ultrasonic image comprises a cross section and a long axis, color blood flow Doppler is displayed in real time, a blood flow velocity spectrum is obtained in real time, a piezoelectric pulse wave signal is obtained, and the heart rate, the blood pressure and the pulse wave propagation velocity are obtained according to the piezoelectric pulse wave signal.

(1) Acquiring pulse wave signals

A pulse wave signal is acquired using a piezoelectric transducer. The measurement position first needs to be determined. Common measurement sites include the brachial artery, radial artery, carotid artery, etc. The flexible hydrogel is arranged on the side close to the skin, so that wearing comfort is increased, and the couplant function is exerted to improve signal transmission quality. The piezoelectric transducer is connected to the signal collector. The signal collector needs to be able to convert the electrical signal output by the piezoelectric transducer into a digital signal, and perform signal processing and storage.

(2) Signal processing

The acquired pulse wave signal needs to be subjected to signal processing such as filtering and amplifying. The common filtering methods include low-pass filtering and band-pass filtering, which can remove high-frequency noise and clutter. Amplifying the signal can improve the signal-to-noise ratio, making the signal clearer.

(3) Heart rate calculation

Heart rate refers to the number of beats per unit time and is an important indicator for measuring the state of cardiovascular function. From the pulse wave signal, the heart rate can be calculated. The following method can be adopted:

(a) The baseline drift method is a simple and practical heart rate calculation method and is suitable for pulse wave signals with single peaks. The method comprises the following specific steps: removing baseline drift from the pulse wave signal; selecting peak points in the signals; calculating the time interval between adjacent peak points; the time interval is converted to heart rate per minute.

(b) The slope method is a heart rate calculation method based on signal slope change and is suitable for multimodal pulse wave signals. The method comprises the following specific steps: performing first-order differential operation on the pulse wave signals to obtain slope signals; selecting a peak point in the slope signal; calculating the time interval between adjacent peak points; the time interval is converted to heart rate per minute.

(c) The fast Fourier transform is a heart rate calculation method based on signal frequency domain analysis, and is suitable for complex pulse wave signals. The method comprises the following specific steps: performing FFT (fast Fourier transform) on the pulse wave signals to obtain spectrograms of the signals; selecting peak points in the spectrogram; the frequency difference between adjacent peak points, i.e. heart rate, is calculated.

(4) Blood pressure calculation

Calculation of blood pressure requires measurement of systolic and diastolic blood pressure. Systolic pressure is the highest pressure of the vessel wall as blood is pumped from the heart and diastolic pressure is the lowest pressure of the vessel wall during relaxation of the heart. The blood pressure can be calculated by measuring the waveform and amplitude of the pulse wave. The method which can be adopted is as follows:

(a) A method based on pulse wave morphology. This method is calculated based on the relationship between the pulse wave morphology and the blood pressure. The method comprises the following specific steps: detecting a systolic start point and a diastolic end point of the pulse wave signal; calculating systolic and diastolic pulse pressure differences, i.e. the difference between systolic and diastolic pressures; the blood pressure is calculated from the known relationship between the pulse wave pattern and the blood pressure.

(b) A method based on pulse wave velocity. This method is calculated based on the relationship between the pulse wave velocity and the blood pressure. The method comprises the following specific steps: measuring pulse wave propagation velocity; the blood pressure is calculated from the known relationship between the pulse wave velocity and the blood pressure.

(c) A method based on a machine learning algorithm. The method is to build a model between pulse wave signals and blood pressure by using a machine learning algorithm, and predict the blood pressure value through the model. The method comprises the following specific steps: collecting a large amount of pulse wave signals and blood pressure value data, and preprocessing and extracting features; training and optimizing the data by using a machine learning algorithm (such as a support vector machine, a neural network and the like); and predicting the new pulse wave signal by using the trained model to obtain a blood pressure value.

(d) Pulse wave velocity calculation.

Pulse wave velocity refers to the velocity of pulse waves propagating in blood vessels, and is an important parameter in pulse wave signals, and can be used for evaluating physiological states such as elasticity and hardness of blood vessels. The following is a commonly used method for calculating pulse wave velocity:

(a) A method for calculating pulse wave velocity using pulse wave travel time and distance between two measurement points of an arterial vessel covered by a sensor. The method comprises the following specific steps: fixing sensors, such as piezoelectric sensors or photoelectric sensors, on two measuring points respectively; simultaneously recording the time of pulse wave signals received by the two sensors, and calculating the pulse wave propagation time between the two measuring points; measuring the distance between two measurement points, typically using body surface distance or ultrasonic ranging; from the known distance and pulse wave travel time, the pulse wave velocity is calculated.

(b) A method for estimating pulse wave velocity by using parameters such as morphological information of pulse wave signals and blood pressure values. The method comprises the following specific steps: measuring pulse wave signals by a piezoelectric sensor and the like, and recording morphological information of the pulse waves; calculating characteristic parameters of the blood vessel, such as the elastic modulus of the blood vessel and the like, according to the known relation between the pulse wave form and the blood pressure; pulse wave velocity is calculated by using blood vessel characteristic parameters, pulse waveform state information and the like, and is usually calculated by adopting a mathematical model or a machine learning algorithm.

As shown in fig. 2, the cross-sectional area and the area change rate of the blood vessel and the area change rate per unit time are calculated from the blood vessel ultrasonic image. From the blood flow velocity spectrum, heart rate, velocity Time Integral (VTI), maximum/average blood flow velocity are obtained, doppler shock index (Doppler shock index =hr/VTI), blood flow, stroke volume=cross-sectional area×systolic flow rate integral, cardiac output (cardioac output=sv×hr) is calculated.

2. Data processing part

2.12.1 establishing a standard blood circulation simulation model comprises the following steps: and obtaining a geometric model, boundary conditions and calculation parameters of a standard blood circulation system by using physiological statistical information, wherein the calculation model adopts a 1DCFD and 0D circuit coupling model, the 1D CFD simulates arteries, and the circuit coupling model simulates downstream arterioles and microcirculation.

Because of the certain deviation between the obtained hemodynamic parameters, especially heart rate, blood pressure and stroke volume, and the real parameters of the user, there are reports of using a statistical method to establish a functional relation calibration method between the real values and the measured values, using an artificial intelligence method to train a depth network between the measured values and the real values, establishing a computational fluid dynamics simulation model and the like, the accuracy and the calculation speed are poor, and the direct depth network calibration method has the requirement on the training data amount as long as hundreds of thousands of cases because physiological elements between excessive real values and the measured values are ignored. The present disclosure proposes a fast calibration method that utilizes digital twinning.

2.1.1 establishing a simulation model of standard blood circulation. First, physiological statistical information (including cardiac output, vessel length, vessel size, vessel branching structure, and arterial downstream microcirculation resistance) is used to derive a standard blood circulation system geometric model, boundary conditions and calculation parameters, wherein the boundary model includes a heart model as shown in fig. 3, and an RCR model as shown in fig. 4. And secondly, coupling 1D CFD of the simulated artery with a 0D circuit model for modeling downstream arterioles and microcirculation to construct a calculation model, wherein a control equation of the 1D CFD model is as follows, as shown in fig. 5:

and obtaining the relationship between the real-time blood flow velocity and the blood pressure distribution of the artery, the cross-sectional area of the arterial vessel and the size and the position distribution of the pulse wave through a standard simulation model.

2.1.2 establishing an individualized digital twin model includes: and determining the individuation parameters and influence factors thereof, establishing a functional relation between measurable and individuation parameters, simulating the individuation model in real time, selecting one or more optimization targets, and adjusting the parameters to minimize the difference between the model calculation value and the actual measurement value of the optimization targets to obtain the optimal parameter combination.

(1) Individualization parameters and influencing factors thereof are determined. The distinctions between the individualization model and the standard model are mainly reflected in the geometric model and the calculation parameters. These parameters can be divided into two broad categories, one being intrinsic and one being real-time. Intrinsic parameters include: 1) Geometric model parameters such as vessel length, vessel mean cross-sectional area (a_0), which are generally related to height, weight, age, gender. 2) Parameters such as Young's modulus of the vessel wall (E in the control equation), vessel wall thickness (h), blood flow viscosity (v), whether the vessel is stenotic, whether the vessel wall is smooth (delta), which are generally related to cardiovascular health. The real-time parameters include: 1) In the heart model, voltage (E in the circuit model), capacitance (C), inductance (L). 2) Resistance to microcirculation downstream of the artery (RCR model). These real-time parameters are related to the real-time physical state in addition to the above-mentioned influencing factors.

(2) A functional relationship between the measurable quantity and the individualizing parameter is established. The function is mainly completed by adopting the existing deep learning network. The measurable quantities are related to the influencing factors of the individualization parameters, mainly comprising the height, weight, age, sex and cardiovascular condition index (such as heart rate, blood pressure, heart rhythm, cardiac output, stroke volume, ventricular ejection fraction and the like) of the individual, and the quantities form the characteristic vector of the deep learning network. In the training data, the standard of the individualization parameters is obtained by a reverse engineering method. Therefore, when the system is tested and used, the difference value between the individuation parameter and the standard parameter can be obtained through the measurable data and the deep learning network, and the simulation of the individuation model on the individual in the normal state is realized through adjusting the parameter. Input is a measurable parameter, input to the CNN network, and output is the difference.

(3) Real-time simulation of the personalized model, measuring the measured local hemodynamic parameters of the carotid artery, the radial artery, the brachial artery and the femoral artery, and matching the calculation result of the personalized model with the measured parameters by adjusting the real-time parameters, wherein the real-time parameters are adjusted by establishing a functional relation between the measured parameters and the real-time parameter changes through a deep learning network.

Specifically, the local hemodynamic parameters including heart rate, blood pressure, blood flow and stroke volume are measured by using the wearable device, and the calculation result of the individuation model is consistent with the measurement parameters of the wearable device by adjusting the real-time parameters. The method for adjusting the real-time parameters also establishes a functional relation between the measurement data of the wearable equipment and the real-time parameter change through the deep learning network. The functional relationship is: input is a measurable parameter, and Input CNN, output is a value of the change in the real-time parameter. CNN is this functional relationship. The functional relationship of CNN is obtained through training. The feature vector of the deep learning network consists of differences between one or more wearable device measurement parameters and a normal average value of the wearable device measurement parameters. In the training data, the gold standard of the variation of the real-time parameters is obtained by a reverse engineering method. Therefore, when the real-time simulation method is used for testing, the difference value between the real-time parameter and the normal state of the real-time parameter is obtained according to the change of the measurement value of the wearable equipment, and the real-time simulation of the personalized model is realized by adjusting the real-time parameter.

(4) And (5) reversely engineering to obtain individuation parameter standards. First, one or more optimization targets are selected, that is, parameters are adjusted, so that the difference between the model calculation value and the actual measurement value of the optimization target is minimized. Secondly, setting optimization constraint, wherein the constraint comprises that the fluctuation range of parameters is required to be within a physiologically reasonable range, and the change rule of the same parameter difference value among different individuals accords with the physiological rule. And finally, optimizing the parameters by selecting a proper optimization algorithm to obtain an optimal parameter combination. Optimization objective selection must meet the following conditions: 1) Measurable. 2) Has a functional relation with the optimization parameters and has enough sensitivity to the variation of the optimization parameters.

2.1.3 determination of real-time pressure-volume Ring

The heart performs work to produce blood ejection, the metrics of which include SV and CO, while the contractile state of the heart is related not only to the heart muscle itself, but also to preload, afterload, which can be represented by the Frank-starling mechanism and the left ventricular pressure-volume loop. Methods have been reported for acquiring this pressure volume loop using cardiac ultrasound, but the end-diastole left ventricular volume and end-systole volume calculated by simultaneous acquisition of blood pressure and cardiac ultrasound images must be obtained. From 2.1.1 and 2.1.2, the relationship between the pressure and ejection volume in the heart chamber of the left ventricle and the hemodynamic parameters such as arterial blood pressure, blood flow, etc. can be obtained, and a large number of virtual data sets can be generated. To accommodate the light weight requirement, as shown in fig. 6, the present disclosure uses hemodynamic parameters such as diastolic pressure, systolic pressure, blood flow, heart rate, etc. further acquired by the transducer as inputs, and outputs left ventricular end-diastolic pressure (i.e., volume ring P1), left ventricular end-systolic pressure (i.e., volume ring P2), and end-systolic and end-diastolic blood flow differences (i.e., volume ring P3), trains the depth network, and the training dataset includes data acquired by clinical reality and virtual dataset generated as described above. In real-time monitoring, the real-time pressure-volume ring can be determined end-to-end. The heart contractility and heart beat function can be quantified by the real-time pressure-volume loop slope.

The heart rate, the blood pressure, the stroke volume and the change of the cardiac output which are monitored in real time can realize the monitoring of the heart contractility and the stroke work through the digital twin. The clinical queue, divided by the above time series data, also comprises established myocardial contractility monitoring data with hemodynamic parameters and cardiac ultrasound, and a patient's queue of whether an acute cardiac event occurs or not. The model will employ an encoder+classifier architecture and perform end-to-end training. For time series data we can use encoders based on one-dimensional convolutional neural networks, for cardiac ultrasound data we will use encoders based on feedforward neural networks. The abstract representation of each modality will be put into the classifier to implement the predictive task. In addition, a method of channel and spatial attention mechanism can be added, as shown in fig. 7 and 8, to extract the data local feature region and enhance the model performance. In addition, the model can also send out abnormal warning signals when the monitored hemodynamic parameters and the vital sign index change rate exceed set thresholds according to clinical diagnosis and treatment standards, such as the conditions of sudden drop of blood pressure, sudden drop or increase of heart rate and sudden drop of contractile force.

Example 2

In one embodiment of the disclosure, a digital twinning-based conformal ultrasound living body visual monitoring system is provided, as shown in fig. 9, and includes a transducer portion, a conformal, light and thin patch-type wireless ultrasound diagnostic device with a certain extensibility, which has a tail wing-like structure, the tail wing is flexible, the transducer is arranged on the tail wing, the transducer can be a single array element or multiple array elements, and multiple main frequencies can be selected according to the use scene array elements, such as 7-12MHz for carotid artery imaging, 5-10MHz for pulse detection, and the like. The tail wing is divided into two parts, and the data transmission and processing modules are integrated in the middle. And part of tail surfaces have larger area, the sound beam range can cover 3-4 times of the diameter width of the common carotid artery, the Doppler angle is fully covered, and the simultaneous detection of the long axis and the short axis of the common carotid artery is realized. The tail wing area of the other part is smaller, and the sound beam range can cover 2 times of the diameter width of the carotid artery, so that the pulse is detected. The invented ultrasonic diagnostic apparatus may be worn in the form of a binding band or adhered to the skin surface with a gel sound guiding material having tackiness.

The data transmission and processing module comprises an array element excitation and signal receiving unit, a signal processing intelligent chip unit, a signal wireless transmission unit and an energy supply unit. The intelligent chip can be any general chip or special chip. In view of power consumption, in order to keep the mobile diagnosis and treatment requirement of the diagnosis equipment for a long time, the two tail wings can work simultaneously or in a time-sharing manner. Typical time-sharing operation forms are: pulse detection empennage with low power consumption works continuously, and imaging empennage with high power consumption works regularly; the pulse detection tail fin with low power consumption continuously works, and the imaging tail fin with high power consumption starts working according to a user demand instruction; the pulse detection tail fin with low power consumption continuously works, and when hemodynamic anomalies are found through calculation of the algorithm module, the imaging tail fin with high power consumption is automatically started to work, so that more accurate hemodynamic parameters are obtained.

Example 3

In one embodiment of the present disclosure, a computer readable storage medium is provided in which are stored a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform the steps of the digital twin based conformal ultrasound vital signs visualization monitoring method.

Example 4

In one embodiment of the disclosure, a terminal device is provided, including a processor and a computer readable storage medium, where the processor is configured to implement instructions; the computer readable storage medium is for storing a plurality of instructions adapted to be loaded by a processor and to perform the digital twin based conformal ultrasound living body visualization monitoring method steps.

The present disclosure is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While the specific embodiments of the present disclosure have been described above with reference to the drawings, it should be understood that the present disclosure is not limited to the embodiments, and that various modifications and changes can be made by one skilled in the art without inventive effort on the basis of the technical solutions of the present disclosure while remaining within the scope of the present disclosure.

Claims (10)

1. The digital twinning-based conformal ultrasonic living body visual monitoring method is characterized by comprising the following steps of:

acquiring a blood vessel ultrasonic image in real time, and acquiring a blood flow velocity spectrum and a piezoelectric pulse wave signal; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals;

establishing a standard blood circulation simulation model, establishing an individualized digital twin model, determining individualized parameters and influence factors, establishing a functional relation between measurable and individualized parameters, and realizing real-time simulation of the individualized model by adjusting the parameters in real time; setting an objective function, optimizing parameters to obtain an optimal parameter combination, and determining real-time pressure-volume loop quantized heart contractility and heart beat function end to end;

extracting the obtained local characteristic region of the hemodynamic parameters, expanding single-point time to a time axis according to data characteristics by combining a clinical queue, realizing continuous visualization of the circulation state parameters and analyzing the variation trend of the parameters.

2. The digital twin based conformal ultrasound living body visualization monitoring method according to claim 1, wherein the blood vessel cross-sectional area and area change rate and unit time area change rate are calculated according to blood vessel ultrasound images, and heart rate, velocity time integral and maximum or average blood flow velocity are obtained according to blood flow velocity spectrum, and doppler shock index, blood flow, stroke volume and cardiac output are calculated at the same time.

3. The digital twin based conformal ultrasound living body visualization monitoring method of claim 1, wherein establishing a standard blood circulation simulation model comprises: and obtaining a geometric model, boundary conditions and calculation parameters of a standard blood circulation system by using physiological statistical information, wherein the calculation model adopts a 1D CFD and 0D circuit coupling model, the 1D CFD simulates arteries, and the circuit coupling model simulates downstream arterioles and microcirculation.

4. The digital twin-based conformal ultrasound living body visual monitoring method according to claim 3, wherein the physiological statistical information comprises cardiac output, blood vessel length, blood vessel size, blood vessel branch structure and arterial downstream microcirculation resistance, and the relationship between the real-time blood flow velocity and blood pressure distribution of the artery, the arterial blood vessel cross-sectional area and the size and position distribution of pulse wave is obtained through a standard simulation model.

5. The digital twinning-based conformal ultrasound living body visualization method of claim 1, wherein the establishing an individualized digital twinning model comprises: and determining the individuation parameters and influence factors thereof, establishing a functional relation between measurable and individuation parameters, simulating the individuation model in real time, selecting one or more optimization targets, and adjusting the parameters to minimize the difference between the model calculation value and the actual measurement value of the optimization targets to obtain the optimal parameter combination.

6. The digital twinning-based conformal ultrasound living body visualization method according to claim 5, wherein said simulating the personalized model in real time comprises: the measured local hemodynamic parameters of the carotid artery, the radial artery, the brachial artery and the femoral artery are measured, the calculation result of the individuation model is matched with the measured parameters by adjusting real-time parameters, and the function relation between the measured parameters and the real-time parameter changes is established through a deep learning network.

7. The digital twin based conformal ultrasound living body visualization monitoring method according to claim 1, wherein the setting objective function, optimizing parameters to obtain an optimal parameter combination, and determining the real-time pressure-volume loop quantized heart contractility and heart beat function end-to-end comprises: selecting one or more optimization targets, and adjusting parameters to enable the difference value between the model calculated value and the actual measured value of the optimization targets to be minimum; secondly, setting optimization constraints, wherein the constraints comprise parameter fluctuation ranges and change rules of the same parameter difference among different individuals; and optimizing the parameters by selecting a proper optimization algorithm to obtain an optimal parameter combination.

8. Digital twinning-based conformal ultrasonic living body visual monitoring system is characterized by comprising:

the transducer data acquisition module is used for acquiring blood vessel ultrasonic images in real time and acquiring blood flow velocity frequency spectrums and piezoelectric pulse wave signals; acquiring hemodynamic parameters according to the piezoelectric pulse wave signals;

the data transmission and processing module is used for establishing a standard blood circulation simulation model and an individualized digital twin model, determining individualized parameters and influence factors, establishing a functional relation between measurable and individualized parameters, and realizing real-time simulation of the individualized model by adjusting the parameters in real time; setting an objective function, optimizing parameters to obtain an optimal parameter combination, and determining real-time pressure-volume loop quantized heart contractility and heart beat function end to end;

extracting the obtained local characteristic region of the hemodynamic parameters, expanding single-point time to a time axis according to data characteristics by combining a clinical queue, realizing continuous visualization of the circulation state parameters and analyzing the variation trend of the parameters.

9. A computer readable storage medium having stored therein a plurality of instructions adapted to be loaded by a processor of a terminal device and to perform the digital twinning-based conformal ultrasound vital sign visualization monitoring method of any of claims 1-7.

10. A terminal device comprising a processor and a computer readable storage medium, the processor configured to implement instructions; a computer readable storage medium for storing a plurality of instructions adapted to be loaded by a processor and to perform the digital twin based conformal ultrasound vital sign visualization monitoring method of any of claims 1-7.