A fistula function monitoring system and terminal device for hemodialysis patients

Abstract

The present invention belongs to the field of hemodialysis monitoring equipment, and provides a hemodialysis patient arteriovenous fistula function monitoring system and terminal equipment. The system comprises: an acquisition module, which is used to obtain the intensity of reflected light returned after detection light is irradiated to blood vessels under the skin of a human forearm, generate a current signal according to the reflected light, and convert the current signal into a measurement signal; a feature extraction module, which is used to obtain an intensity change sequence of the reflected light based on the measurement signal, separate a DC component and an AC component from the intensity change sequence, and obtain a blood perfusion index, a heart rate, and blood oxygen saturation, systolic pressure and diastolic pressure; a function evaluation module, which uses the maximum and minimum values ​​of the blood perfusion index, blood oxygen saturation, systolic pressure and diastolic pressure as first features, and adds heart rate as a second feature, and obtains the stenosis degree and blood flow of the blood vessel based on a machine learning classifier evaluation; the present invention can solve the problem in the prior art that the function of the arteriovenous fistula of a hemodialysis patient is difficult to intuitively monitor.

Description

Technical Field

[0001] The present invention relates to the field of hemodialysis monitoring equipment, and in particular to a hemodialysis patient's fistula function monitoring system and terminal equipment. Background Art

[0002] Arteriovenous fistula (AVF) refers to a vascular fistula that connects the forearm artery and vein through surgery. Arteriovenous fistula is a key lifeline for hemodialysis patients, and its health status directly affects the effect of dialysis treatment and the patient's quality of life. After the fistula is established and during dialysis, patients need to monitor the function of the fistula on a daily basis, such as blood flow and whether it is narrowed.

[0003] However, existing methods for monitoring the function of arteriovenous fistulas mainly rely on ultrasound Doppler devices, which are large, expensive, and require professional operation, limiting their application in home care and daily monitoring. In addition, accurate indicators for evaluating arteriovenous fistulas usually require complex instruments and operating procedures, which is inconvenient for patients to monitor at home during hemodialysis, resulting in failure to detect the decline in fistula function in time, delaying the timing of intervention, causing fistula failure and affecting patient dialysis, etc., resulting in re-surgical intervention, increasing the patient's disease pain and medical expenses. Summary of the invention

[0004] In view of the above technical problems, the present invention provides a hemodialysis patient fistula function monitoring system and terminal equipment to solve the problem in the prior art that the function of the arteriovenous fistula during the hemodialysis interval is difficult to monitor and manage.

[0005] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0006] According to one aspect of the present invention, a system for monitoring the function of internal fistula in hemodialysis patients is disclosed, the system comprising: An acquisition module, the acquisition module is used to acquire the intensity of reflected light returned after the detection light irradiates the blood vessels under the skin of the human forearm, the detection light has a preset wavelength, and when the intensity of the reflected light is acquired, a current signal is generated according to the reflected light, the current signal is converted into a voltage signal, and after removing environmental noise and dynamically adjusting the signal gain, the voltage signal is converted into a measurement signal; a feature extraction module, the feature extraction module being used to obtain an intensity variation sequence of the reflected light based on the measurement signal, separate a DC component and an AC component from the intensity variation sequence, obtain a blood perfusion index based on a ratio of the DC component to the AC component, and obtain a heart rate based on a pulsation frequency of the measurement signal; the feature extraction module is also used to extract blood oxygen saturation, systolic pressure, and diastolic pressure; A function evaluation module, which takes the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, and the diastolic pressure as the first feature, and takes the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, the diastolic pressure, and the heart rate as the second feature, and based on a machine learning classifier, obtains the degree of stenosis of the blood vessel according to the first feature, obtains the blood flow in the blood vessel according to the second feature, and obtains the function of the fistula according to the degree of stenosis and the blood flow.

[0007] Furthermore, the acquisition module includes a photodetector, a transimpedance amplifier, a bandpass filter and a programmable gain amplifier, the photodetector is used to receive the reflected light and generate a current signal, the transimpedance amplifier converts the current signal into the voltage signal, the bandpass filter is used to filter out the environmental noise signal in the voltage signal, the programmable gain amplifier dynamically adjusts the signal gain according to the amplitude of the voltage signal, and when filtering out the environmental noise signal, determines the cutoff frequency according to the heart rate.

[0008] Furthermore, the detection light has multiple wavelengths, and the feature extraction module calculates the blood oxygen saturation based on light absorption characteristics.

[0009] Furthermore, the blood oxygen saturation is obtained based on a blood oximeter, and the systolic pressure and the diastolic pressure are obtained based on an electronic sphygmomanometer.

[0010] Furthermore, the feature extraction module is also used to perform time domain and frequency domain analysis on the intensity variation sequence of the reflected light to extract information containing pulse waveform features to obtain the DC component and the AC component.

[0011] Furthermore, the machine learning classifier includes one of a k-nearest neighbor, a naive Bayes classifier, and a support vector machine.

[0012] Furthermore, the support vector machine, when training, includes: The first feature and the second feature are used as training sets, and labels are defined, and the labels are set to be positive or negative according to the degree of stenosis and the size of the blood flow; Standardizing the training set; Setting a classification goal and establishing an optimization objective function of the support vector machine to maximize the interval of the classification boundary while minimizing the classification error; Based on a linear kernel function or a radial basis kernel function, the training set is mapped to a high-dimensional feature space, and a weight vector and a bias are determined by optimizing and solving a decision hyperplane of a support vector machine; A cross-validation method is used to tune the regularization parameter of the classifier and the linear kernel function or the radial basis kernel function to improve the generalization performance of the support vector machine.

[0013] Furthermore, the function evaluation module generates a quality control report according to the evaluation results of the stenosis degree and the blood flow, and transmits the report to the user terminal via the wireless communication module.

[0014] Furthermore, the acquisition module is used to compensate for artifacts and noise on the acquired measurement signal to improve signal quality and evaluation accuracy.

[0015] According to another aspect of the present invention, a system for monitoring fistula function in hemodialysis patients is disclosed, wherein the terminal device comprises a measuring head, an analog circuit, and a controller, wherein the measuring head comprises a light emitting diode and a photodiode, the analog circuit comprises a transimpedance amplifier, a bandpass filter, a programmable gain amplifier, and an analog-to-digital converter, the controller controls the light emitting diode to emit measuring light through the analog circuit, and generates a current signal after receiving reflected light of the measuring light through the photodiode, the transimpedance amplifier is used to convert the current signal into a voltage signal and then amplify it, the bandpass filter is used to remove environmental noise signals from the voltage signal, the programmable gain amplifier is used to adjust the gain of the voltage signal so that its amplitude adapts to the dynamic range of the analog-to-digital converter, the analog-to-digital converter is used to convert the voltage signal into a measuring signal, and the controller is also used to: Obtaining an intensity variation sequence of the reflected light based on the measurement signal, separating a DC component and an AC component from the intensity variation sequence, obtaining a blood perfusion index based on a ratio of the DC component to the AC component, and obtaining a heart rate based on a pulsation frequency of the measurement signal; Extract blood oxygen saturation, systolic and diastolic blood pressure; The maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, and the diastolic pressure are used as the first feature, and the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, the diastolic pressure, and the heart rate are used as the second feature. Based on a machine learning classifier, the degree of stenosis of the blood vessel is evaluated according to the first feature, the blood flow in the blood vessel is evaluated according to the second feature, and the function of the fistula is obtained according to the degree of stenosis and the blood flow.

[0016] The technical solution disclosed in this disclosure has the following beneficial effects: The present invention discloses an integrated solution, which realizes the complete process from signal acquisition, feature extraction to blood vessel quality assessment. The system uses optical principles to obtain blood vessel characteristics; calculates the blood perfusion index based on the separated DC component and AC component, combines the characteristics such as heart rate, blood oxygen saturation and blood pressure, and uses a machine learning classifier to accurately assess the degree of vascular stenosis and blood flow, providing hemodialysis patients with a convenient, low-cost solution for monitoring their fistula function anytime and anywhere. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a structural block diagram of a system for monitoring the fistula function of a hemodialysis patient in an embodiment of this specification; Figure 2 A waveform diagram of a measurement signal in an embodiment of this specification; Figure 3 is a structural diagram of a blood vessel in an embodiment of this specification; Figure 4 This is a structural block diagram of a terminal device for monitoring fistula function of hemodialysis patients in an embodiment of this specification. DETAILED DESCRIPTION

[0018] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as being limited to the examples set forth herein; on the contrary, these embodiments are provided so that the present disclosure will be more comprehensive and complete, and the concepts of the example embodiments are fully conveyed to those skilled in the art. The described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0019] In addition, the accompanying drawings are only schematic illustrations of the present disclosure. The same reference numerals in the figures represent the same or similar parts, and thus their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities, which do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0020] In one embodiment, if Figure 1-3 As shown, this specification provides a system for monitoring the fistula function of a hemodialysis patient, the system comprising: The acquisition module 101 is used to obtain the intensity of the reflected light returned after the detection light is irradiated to the blood vessels under the skin of the human forearm, the detection light has a preset wavelength, and when the intensity of the reflected light is obtained, a current signal is generated according to the reflected light, the current signal is converted into a voltage signal, and after removing the environmental noise and dynamically adjusting the signal gain, the voltage signal is converted into a measurement signal.

[0021] Specifically, the acquisition module 101 includes a photodetector, a transimpedance amplifier, a bandpass filter and a programmable gain amplifier. The photodetector is used to receive the reflected light and generate a current signal. The transimpedance amplifier converts the current signal into the voltage signal. The bandpass filter is used to filter out the environmental noise signal in the voltage signal. The programmable gain amplifier dynamically adjusts the signal gain according to the amplitude of the voltage signal. When filtering out the environmental noise signal, the cutoff frequency is determined according to the heart rate.

[0022] Among them, the arteriovenous fistula is generally located in the forearm of the human body, and the light beam emitted from the photodetector penetrates, refracts and reflects back from the blood vessels in the arteriovenous fistula to be measured. Then, the photodetector detects the change in the intensity of the reflected light caused by the pulsation of the blood vessels at the arteriovenous fistula and converts it into a current signal. Next, based on the transimpedance amplifier circuit, the current signal from the photodetector is converted into a voltage signal. Here, in order to deal with the interference of motion artifacts and ambient light (generally 50 or 60Hz), a 4th-order bandpass filter with a cutoff frequency of about 0.2 and 10Hz can be used to remove unwanted frequency components. The cutoff frequency of the noise can be determined in real time by the feature extraction module below based on the heart rate analyzed from the measured voltage signal. Since each person's heart rate varies greatly (i.e., about 50-110Hz), the adaptability of the cutoff frequency can be set, that is, a programmable gain amplifier is set to improve the signal-to-noise ratio by controlling the voltage signal to fill the entire dynamic range of the analog-to-digital conversion. Due to the design of the fourth-order bandpass filter and the programmable gain amplifier, it is possible to present a signal waveform with almost no DC and full dynamic range. Finally, the measurement signal from the analog-to-digital conversion is transmitted to the client of the smartphone or laptop wirelessly after feature extraction.

[0023] A feature extraction module 102, the feature extraction module 102 is used to obtain an intensity variation sequence of the reflected light based on the measurement signal, separate a DC component and an AC component from the intensity variation sequence, obtain a blood perfusion index based on a ratio of the DC component to the AC component, and obtain a heart rate based on a pulsation frequency of the measurement signal; the feature extraction module is also used to extract blood oxygen saturation, systolic pressure and diastolic pressure.

[0024] The function evaluation module 103 uses the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, and the diastolic pressure as the first feature, and uses the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, the diastolic pressure, and the heart rate as the second feature, and based on a machine learning classifier, obtains the degree of stenosis of the blood vessel according to the first feature evaluation, obtains the blood flow in the blood vessel according to the second feature evaluation, and obtains the function of the fistula according to the degree of stenosis and the blood flow.

[0025] In the above-mentioned feature extraction module 102 and function evaluation module 103, the feature extraction module 102 mainly analyzes the measurement signal to obtain a number of features for quality evaluation, and the function evaluation module 103 analyzes the fistula function according to these features. Figure 2 As shown in the figure, the DC component comes from static tissues, such as veins, skin, and bones, while the AC component comes from the pulsation of the measured blood vessels. Figure 2 It consists of a large DC component (>90%) and a small AC component (<10%), and the pulse frequency is the same as the heart rate (about 50-110 times / minute). The measurement signal can be described by the known physical theory Beer-Lambert law as follows: ; in, Indicates the light intensity received from the PD; Indicates the intensity of incident light; and represent the light absorption coefficients of blood and tissue, respectively; and denote the molar concentrations in blood and tissues, respectively; and Represent the optical transmission paths of blood and tissue respectively. Since the diameter of the measured blood vessel changes with the heart beat, is time-varying. In addition, by applying Taylor approximation to the exponential function of the above formula, we can obtain: ; In order to derive the length of the optical transmission path to reduce interference from other tissues, the above AC and DC are standardized to obtain: ; Other organizations ( ) is normalized, and AC / DC is defined as the perfusion index. Taking into account the blood oxygen level, Can be expanded to: ; is the blood oxygen level, that is, blood oxygen saturation, , are the light absorption coefficients of oxygenated hemoglobin and hemoglobin, respectively. Combined with the perfusion index, we can get: ; Where PI represents the perfusion index, from which it can be obtained that the light transmission path is proportional to the time-varying diameter of the measured blood vessel pulsation. In this way, the light transmission path can be obtained according to PI, and then the time-varying diameter of the blood vessel pulsation can be obtained to evaluate the degree of vascular stenosis and blood flow in the machine classifier.

[0026] In this embodiment, before evaluating the quality of the blood vessel, a dynamic model of the blood vessel can be established, which can be expressed as follows according to the momentum and mass conservation equations:

[0027] in and Respectively indicate the location and time Instantaneous blood flow and blood pressure at the site; It also indicates the axis of the blood vessel, such as Figure 3 shown.

[0028] ; Where R represents vascular resistance; L represents vascular inertia; G represents the influence of vascular branches, which is assumed to be zero; C represents vascular compliance; Indicates the dynamic viscosity of blood; Indicates the density of blood (approximately 1.056 g / cm³); represents the Poisson's ratio of the blood vessel; represents the diameter of the blood vessel at the measurement point; represents the thickness of the blood vessel wall; E represents the Young's elastic modulus of the blood vessel; and is a parameter that describes the effect of heart rate pulsation and can be approximately expressed as: ; Where W represents the Womersley number, which is defined as: ; in represents the heart rate. Assume that R, L, and C are independent of t and z, and that the blood vessels are of finite length. Therefore, the momentum and mass conservation equations can be solved to obtain:

[0029] ; in represents the positive direction of the blood pressure wave, and Indicates the blood flow in the positive direction. The hemodynamic model is the basic equation that determines the first and second characteristics of the evaluation of the stenosis degree and blood flow of the blood vessel.

[0030] Specifically, the degree of stenosis is the ratio of the cross-sectional area between the normal blood vessel and the stenotic blood vessel, which can be expressed as: ; Where d and D represent the diameters of stenotic vessels and normal vessels, respectively. The above formula can also be re-expressed as: ; in represents the basic thickness of the blood vessel, usually assumed to be a constant; represents the thickness of endothelial cell proliferation. In order to obtain the vascular thickness h, the average vascular compliance is introduced, that is: ; in and Respectively represent the maximum and minimum cross-sectional areas of the blood vessels; SBP and DBP represent systolic and diastolic blood pressures, respectively. Combining the definition of C in the above formula, the blood vessel thickness can be obtained as: ; in and Represent the maximum and minimum diameters of blood vessels respectively. and Optical transmission path To express, and then summarize it: ; in, and Respectively represent the corresponding The maximum and minimum perfusion indexes can be measured because AC and DC are and can be measured directly. Similarly, due to the maximum and minimum Both SpO2 and α are important factors affecting DOS, and both are mathematically and physiologically independent of each other. Therefore, these five independent first features can affect DOS, that is, the degree of stenosis, and the degree of stenosis of the blood vessels can be evaluated by the machine classifier algorithm.

[0031] In addition, regarding the blood flow of blood vessels, the characteristic impedance of blood vessels can be referenced. The characteristic impedance can be defined as: ; in Representing the characteristic impedance of the blood vessel, the above equation can be rearranged as: ; Assume that the measurement point is at z=0, such as Figure 3 As shown, the above formula can be further transformed into: ; To obtain the average blood flow, This can be approximated using mean blood pressure (MBP): ; Then, by combining the above formula, the average blood flow of the measured blood vessel can be obtained, that is: ; in represents the mean blood flow at the vessel, and the absolute value sign is added to obtain the amplitude of the complex signal. In summary, all dependencies of the mean blood flow can be specified by the following formula: ; Therefore, the above six indicators can be used as the second feature to evaluate blood flow.

[0032] After obtaining the first feature and the second feature, the converted digital detection signal is transmitted to a smartphone client or a laptop computer via wireless or wired transmission for the classifier to evaluate the quality of the blood vessel. The machine classifier can be any one of the k nearest neighbor, naive Bayes classifier and support vector machine. When using the K nearest neighbor, the distance matrix is ​​calculated using the Euclidean distance, and the optimized k value is obtained by using the grid search technique. When using the naive Bayes classifier, the first feature and the second feature distribution are both assumed to be Gaussian distributions, so the mathematical equation of the probability density function of the test data y being classified into the category Ck can be expressed as: ; Where x represents a continuous value; y represents test data; Ck represents the kth category; and represent the mean and variance of the training data set respectively.

[0033] When using a support vector machine, it is necessary to train the support vector machine. During the training, the steps include: taking the first feature and the second feature as training sets, defining labels, and setting the labels to positive or negative according to the degree of stenosis and the size of the blood flow; standardizing the training set; setting a classification target, and establishing an optimization objective function of the support vector machine to maximize the interval of the classification boundary while minimizing the classification error; mapping the training set to a high-dimensional feature space based on a linear kernel function or a radial basis kernel function, and determining a weight vector and a bias by optimizing and solving the decision hyperplane of the support vector machine; and using a cross-validation method to tune the regularization parameter of the classifier, and the linear kernel function or the radial basis kernel function, so as to improve the generalization performance of the support vector machine.

[0034] In this embodiment, the support vector machine classifier calculation process can be described as a dual quadratic form of a Lagrangian optimization problem: ; in is the Lagrangian objective function; Is a directional quantity and the separating hyperplane of the bias vector b; and is the Lagrange multiplier; is a feature vector with class labels; is a slack variable representing misclassified topics; is the penalty for misclassification. Furthermore, the solution formula leads to the dual form: ; The constraints are: ; In order to deal with nonlinear separable classification problems, the kernel function technology is introduced to convert the dot product Transform to another feature space. In this embodiment, the radial basis function is used as the kernel function: ; in Represents the dot product The radial basis kernel function of is an adjustable parameter representing the scaling factor. It is determined by using a grid search technique and The optimized value of .

[0035] Since the first feature and the second feature are dependent on each other, the support vector machine is used for analysis. The dependency is handled by establishing a hyperplane between the features. In order to avoid weighting errors, all features are standardized to between -1 and 1 before training. Then, the category is calculated by the decision function during classification. If the result is greater than 0, it is classified as a positive class; if the result is less than or equal to 0, it is classified as a negative class. Then, the classification result is combined with a threshold (such as DOS = 30% or BFV = 600mL / min) to determine whether the classification meets the requirements.

[0036] In one embodiment, the detection light has multiple wavelengths, and the feature extraction module calculates the blood oxygen saturation based on light absorption characteristics.

[0037] The detection light may have multiple wavelengths, such as red light and infrared light, wherein the wavelength of red light is used to enhance the absorption of reduced hemoglobin, and the wavelength of infrared light is used to enhance the absorption of oxygenated hemoglobin; the feature extraction module calculates the blood oxygen saturation based on the light absorption characteristics by analyzing the absorption changes of detection lights of different wavelengths, specifically including the following steps: The AC component and the DC component of the red light and the infrared light are collected respectively, and the wavelengths of the infrared light and the red light may be 660nm and 940nm; Calculate the ratio of red light to infrared light, the expression is: ; in, and They are the AC and DC components of red light respectively; and They are the AC component and DC component of infrared light respectively; Based on the ratio and the preset calibration curve, the blood oxygen saturation is calculated by the empirical formula: ; Where A and B are empirical constants determined by system calibration.

[0038] In one embodiment, the blood oxygen saturation is obtained based on a blood oximeter, and the systolic pressure and the diastolic pressure are obtained based on an electronic sphygmomanometer. Systolic pressure refers to the maximum pressure exerted by blood on the blood vessel wall when the heart contracts. This is the pressure exerted when the heart pumps blood into the vascular system, and usually reaches its highest value in the middle of the heart contraction. Diastolic pressure refers to the minimum pressure exerted by blood on the blood vessel wall when the heart relaxes. This is the minimum pressure maintained in the blood vessels during the period when blood flows back to the heart between two heart beats.

[0039] In one embodiment, the feature extraction module is further used to perform time domain and frequency domain analysis on the intensity variation sequence of the reflected light to extract information containing pulse waveform features to obtain the DC component and the AC component. Figure 2 As shown in the figure, during the extraction, the intensity change sequence of the reflected light is denoised and normalized to remove noise such as ambient light interference and motion artifacts, and the signal amplitude is standardized to enhance signal stability. Then, the overall trend of the signal is extracted using a smoothing filter algorithm, and the low-frequency component is separated as a DC component. The characteristics of the pulse waveform, such as the pulse wave amplitude, pulse interval time (reflecting the heart rate), etc., are extracted by detecting the local extreme points (peaks and valleys) of the signal. Then, the preprocessed signal is fast Fourier transformed to convert the signal from the time domain to the frequency domain; by analyzing the spectrum amplitude, the main frequency is extracted in the frequency range of 0.5Hz to 5Hz.

[0040] In one embodiment, the function evaluation module generates a quality control report based on the evaluation results of the stenosis degree and the blood flow, and transmits the report to the user terminal via the wireless communication module.

[0041] In one embodiment, the acquisition module is used to compensate for artifacts and noise on the acquired measurement signal to improve signal quality and evaluation accuracy.

[0042] Based on the same idea, Figure 2 As shown, an exemplary embodiment of the present disclosure also provides a terminal device for monitoring fistula function of hemodialysis patients, the terminal device comprising a measuring head 210, an analog circuit 220, and a controller 230, the measuring head 210 comprising a light emitting diode 211 and a photodiode 212, the analog circuit 220 comprising a transimpedance amplifier 221, a bandpass filter 222, a programmable gain amplifier 223, and an analog-to-digital converter 224, the controller 230 controls the light emitting diode 211 to emit a measuring light through the analog circuit 220, and generates a current signal after receiving the reflected light of the measuring light through the photodiode 212, the transimpedance amplifier 221 is used to convert the current signal into a voltage signal and then amplify it, the bandpass filter 222 is used to remove the environmental noise signal from the voltage signal, the programmable gain amplifier 223 is used to adjust the gain of the voltage signal so that its amplitude adapts to the dynamic range of the analog-to-digital converter 224, the analog-to-digital converter 224 is used to convert the voltage signal into a measuring signal, and the controller 230 is also used to: Obtaining an intensity variation sequence of the reflected light based on the measurement signal, separating a DC component and an AC component from the intensity variation sequence, obtaining a blood perfusion index based on a ratio of the DC component to the AC component, and obtaining a heart rate based on a pulsation frequency of the measurement signal; Extract blood oxygen saturation, systolic and diastolic blood pressure; The maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, and the diastolic pressure are used as the first feature, and the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, the diastolic pressure, and the heart rate are used as the second feature. Based on a machine learning classifier, the degree of stenosis of the blood vessel is evaluated according to the first feature, the blood flow in the blood vessel is evaluated according to the second feature, and the function of the fistula is obtained according to the degree of stenosis and the blood flow.

[0043] This terminal device is an integrated solution that realizes the complete process from signal acquisition, feature extraction to vascular quality assessment. The system uses optical principles to obtain vascular characteristics; calculates the blood perfusion index based on separated DC components and AC components, combines heart rate, blood oxygen saturation and blood pressure and other characteristics, and uses machine learning classifiers to accurately assess the degree of vascular stenosis and blood flow, providing hemodialysis patients with a convenient, low-cost solution for monitoring their vascular quality anytime and anywhere.

[0044] The specific physical form of the above terminal device may be a measuring pen, which emits measuring light and receives reflected light. The undisclosed parts of the above terminal device may refer to the system embodiment.

[0045] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the exemplary implementation of the present disclosure.

[0046] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to the exemplary embodiment of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0047] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

Claims

1. A system for monitoring the function of fistula in hemodialysis patients, characterized in that: The system comprises: An acquisition module, the acquisition module is used to acquire the intensity of reflected light returned after the detection light irradiates the blood vessels under the skin of the human forearm, the detection light has a preset wavelength, and when the intensity of the reflected light is acquired, a current signal is generated according to the reflected light, the current signal is converted into a voltage signal, and after removing environmental noise and dynamically adjusting the signal gain, the voltage signal is converted into a measurement signal by analog-to-digital conversion; a feature extraction module, the feature extraction module being used to obtain an intensity variation sequence of the reflected light based on the measurement signal, separate a DC component and an AC component from the intensity variation sequence, obtain a blood perfusion index based on a ratio of the DC component to the AC component, and obtain a heart rate based on a pulsation frequency of the measurement signal; the feature extraction module is also used to extract blood oxygen saturation, systolic pressure, and diastolic pressure; A function evaluation module, which takes the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, and the diastolic pressure as the first feature, and takes the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, the diastolic pressure, and the heart rate as the second feature, and based on a machine learning classifier, obtains the degree of stenosis of the blood vessel according to the first feature, obtains the blood flow in the blood vessel according to the second feature, and obtains the function of the fistula according to the degree of stenosis and the blood flow.

2. A hemodialysis patient fistula function monitoring system according to claim 1, characterized in that: The acquisition module includes a photodetector, a transimpedance amplifier, a bandpass filter and a programmable gain amplifier. The photodetector is used to receive the reflected light and generate a current signal. The transimpedance amplifier converts the current signal into the voltage signal. The bandpass filter is used to filter out the environmental noise signal in the voltage signal. The programmable gain amplifier dynamically adjusts the signal gain according to the amplitude of the voltage signal. When filtering out the environmental noise signal, the cutoff frequency is determined according to the heart rate.

3. A hemodialysis patient fistula function monitoring system according to claim 1, characterized in that: The detection light has multiple wavelengths, and the feature extraction module calculates the blood oxygen saturation based on light absorption characteristics.

4. A hemodialysis patient fistula function monitoring system according to claim 1, characterized in that: The blood oxygen saturation is obtained based on a blood oximeter, and the systolic pressure and the diastolic pressure are obtained based on an electronic sphygmomanometer.

5. The internal fistula function monitoring system for hemodialysis patients according to claim 1, characterized in that: The feature extraction module is also used to perform time domain and frequency domain analysis on the intensity variation sequence of the reflected light to extract information containing pulse waveform features and obtain the DC component and the AC component.

6. The internal fistula function monitoring system for hemodialysis patients according to claim 1, characterized in that: The machine learning classifier includes one of a k-nearest neighbor, a naive Bayes classifier, and a support vector machine.

7. A system for monitoring the internal fistula function of hemodialysis patients according to claim 6, characterized in that: The support vector machine, when training, includes: The first feature and the second feature are used as training sets, and labels are defined, and the labels are set to be positive or negative according to the degree of stenosis and the size of the blood flow; Standardizing the training set; Setting a classification goal and establishing an optimization objective function of the support vector machine to maximize the interval of the classification boundary while minimizing the classification error; Based on a linear kernel function or a radial basis kernel function, the training set is mapped to a high-dimensional feature space, and a weight vector and a bias are determined by optimizing and solving a decision hyperplane of a support vector machine; A cross-validation method is used to tune the regularization parameter of the classifier and the linear kernel function or the radial basis kernel function to improve the generalization performance of the support vector machine.

8. The internal fistula function monitoring system for hemodialysis patients according to claim 1, characterized in that: The function evaluation module generates a quality control report according to the evaluation results of the stenosis degree and the blood flow, and transmits the report to the user terminal through the wireless communication module.

9. The internal fistula function monitoring system for hemodialysis patients according to claim 1, characterized in that: The acquisition module is used to compensate for artifacts and noise on the acquired measurement signal to improve signal quality and evaluation accuracy.

10. A terminal device for monitoring fistula function of hemodialysis patients, characterized in that: The terminal device includes a measuring head, an analog circuit, and a controller. The measuring head includes a light emitting diode and a photodiode. The analog circuit includes a transimpedance amplifier, a bandpass filter, a programmable gain amplifier, and an analog-to-digital converter. The controller controls the light emitting diode to emit measuring light through the analog circuit, and generates a current signal after receiving the reflected light of the measuring light through the photodiode. The transimpedance amplifier is used to convert the current signal into a voltage signal and then amplify it. The bandpass filter is used to remove the environmental noise signal from the voltage signal. The programmable gain amplifier is used to adjust the gain of the voltage signal so that its amplitude adapts to the dynamic range of the analog-to-digital converter. The analog-to-digital converter is used to convert the voltage signal into a measuring signal. The controller is also used to: Obtaining an intensity variation sequence of the reflected light based on the measurement signal, separating a DC component and an AC component from the intensity variation sequence, obtaining a blood perfusion index based on a ratio of the DC component to the AC component, and obtaining a heart rate based on a pulsation frequency of the measurement signal; Extract blood oxygen saturation, systolic and diastolic blood pressure; The maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, and the diastolic pressure are used as the first feature, and the maximum and minimum values ​​of the blood perfusion index, the blood oxygen saturation, the systolic pressure, the diastolic pressure, and the heart rate are used as the second feature. Based on a machine learning classifier, the degree of stenosis of the blood vessel is evaluated according to the first feature, the blood flow in the blood vessel is evaluated according to the second feature, and the function of the fistula is obtained according to the degree of stenosis and the blood flow.

Images