Intelligent hemostasis system and method

Abstract

The invention relates to the technical field of clinical medical treatment, in particular to an intelligent hemostasis system and method. Blood pressure, heart rate, hemostasis pressure, patient body data and patient wound data are collected through the hemostasis device, a hemostasis time-pressure prediction model is constructed through a differential evolution algorithm and a WaveNet model, prediction of a relation curve of the hemostasis pressure and the hemostasis time is achieved, and the hemostasis pressure and the hemostasis time are predicted through the relation curve. The hemostasis device can adjust the pressure applied by the hemostasis device according to needs, and the problems that in the war wound treatment process, the hemostasis pressure of a tourniquet is too large, the hemostasis time is too long, consequently, limbs or pressed local parts are necrotic, and the hemostasis effect is poor due to the too small pressure are solved.

Description

Technical Field

[0001] The present invention relates to the field of clinical medical technology, and in particular to an intelligent hemostasis system and method. Background Art

[0002] During the rescue of the wounded, there are cases of misuse of tourniquets, including unnecessary use of tourniquets, tourniquets placed too high and placed for too long, resulting in the use of tourniquets for too long. The longer the tourniquet takes to be fully tightened on the limbs, the greater the risk of compartment syndrome, vascular thrombosis, rhabdomyolysis and irreversible muscle necrosis, which may lead to severe tissue loss and limb amputation, threatening life or causing unnecessary disability.

[0003] Moreover, the hemostatic effect depends on the medical staff to control the hemostatic pressure and time based on their experience, such as adjusting the air pressure in the hemostatic airbag to achieve the regulation of the hemostatic pressure. This subjective judgment method may lead to improper pressure, thus causing excessive bleeding or thrombosis, etc. Existing hemostatic devices lack an accurate pressure monitoring mechanism, especially in the face of individual differences among patients, it is difficult to achieve personalized hemostatic pressure regulation. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention aims to provide an intelligent hemostasis system and method to address the problems of excessive tourniquet hemostasis pressure, long hemostasis time leading to necrosis of limbs or compressed parts, and poor hemostasis due to insufficient pressure during combat wound treatment.

[0005] In order to solve the above problems, the present invention adopts the following technical solutions: In one aspect, the present invention provides an intelligent hemostasis system, comprising a data acquisition module, a data processing module, a model building module, a prediction module and a hemostasis module.

[0006] The data acquisition module is used to collect the patient's blood pressure, heart rate and the pressure between the hemostatic device and the skin through the hemostatic device, and collect the patient's body data and the patient's wound data.

[0007] The data processing module is used to pre-process the collected data, construct the data of the patient's blood pressure, heart rate and the pressure between the hemostasis device and the skin into a time series based on the hemostasis time, and use the patient's body data and the patient's wound data as labels of the time series to form time series data.

[0008] The model building module is used to use time series data of successful hemostasis without excessive pressure and excessive hemostasis time to build a hemostasis time-pressure prediction model through a differential evolution algorithm and a WaveNet model.

[0009] The prediction module is used to input the patient's blood pressure, heart rate, patient body data and patient wound data through a hemostasis time-pressure prediction model to obtain a relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time.

[0010] The hemostasis module is used to control the hemostasis device to adjust the pressure between the hemostasis device and the skin according to a relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time.

[0011] As an implementable embodiment, the patient's body data includes age, gender, BMI, and body fat; the patient's wound data includes part, area, depth, and amount of bleeding per unit time.

[0012] As an implementable embodiment, the WaveNet model is composed of a two-dimensional convolution feature extraction layer, a one-dimensional sequence reorganization layer, two residual modules, a fully connected layer and an output layer in sequence; the residual module obtains output one through two causal dilated convolution layers, a batch normalization layer and a random inactivation layer, obtains output two through 1×1 convolution, and combines output one and output two to obtain the output of the residual module.

[0013] The causal dilated convolution layer uses a dilated convolution with a dilation coefficient to expand the receptive field:

[0014] in, is the input time series data, W is the convolution kernel weight, b is the bias term, is the output of the convolutional layer, is the activation function.

[0015] The WaveNet model introduces gated recurrent units:

[0016] in, They are the update gate and the reset gate, is the Sigmoid function, is element-wise multiplication, In hidden state.

[0017] The output layer performs linear regression:

[0018] in, is the predicted value, and are the weight and bias of the output layer respectively.

[0019] As an implementable method, the method of constructing a hemostasis time-pressure prediction model by using a differential evolution algorithm and a WaveNet model includes: optimizing the weight of the WaveNet model by using a differential evolution algorithm, including: Randomly initialize individuals in the population, each individual represents a set of weights of the WaveNet model; Perform mutation operation on each individual to generate test individuals; Perform crossover operation on the test individuals and the original individuals to generate candidate solutions; A selection operation is performed between the original individuals and the candidate solutions, and individuals with higher fitness are selected to enter the next generation; until the loss function of the WaveNet model meets the requirements.

[0020] As an implementable embodiment, the hemostatic device includes a detection and control body and a tourniquet, and the detection and control body is provided with a blood pressure detection device, a heart rate monitoring device, a skin pressure detection device, a time module, a retractable device and / or a patient body data and patient wound data input device. One end of the tourniquet is inserted into the detection and control body to achieve a detachable connection with the detection and control body, and the other end is embedded in the retractable device. The retractable device adjusts the pressure between the hemostatic device and the skin by tightening or relaxing.

[0021] As an implementable embodiment, the outer side of the tourniquet is serrated, and the retracting and releasing device includes a motor and a rotor fixedly connected to the motor and matching the serrated shape of the tourniquet.

[0022] In another aspect, the present invention provides an intelligent hemostasis method, comprising: The patient's blood pressure, heart rate, and pressure between the hemostasis device and the skin are collected through the hemostasis device, and the patient's body data and the patient's wound data are collected.

[0023] The collected data is preprocessed, and the data of the patient's blood pressure, heart rate, and the pressure between the hemostasis device and the skin are constructed into a time series based on the hemostasis time. The patient's body data and the patient's wound data are used as labels of the time series to form time series data.

[0024] The time series data of successful hemostasis without excessive pressure and prolonged hemostasis time were used to construct a hemostasis time-pressure prediction model using the differential evolution algorithm and the WaveNet model.

[0025] Through the hemostasis time-pressure prediction model, the patient's blood pressure, heart rate, patient body data and patient wound data are input to obtain the relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time.

[0026] According to the relationship curve between the pressure between the hemostatic device and the skin and the hemostatic time, the hemostatic device is controlled to adjust the pressure between the hemostatic device and the skin.

[0027] As an implementable embodiment, the patient's body data includes age, gender, BMI, and body fat; the patient's wound data includes part, area, depth, and amount of bleeding per unit time.

[0028] As an implementation method, the WaveNet model is composed of a two-dimensional convolution feature extraction layer, a one-dimensional sequence reorganization layer, two residual modules, a fully connected layer and an output layer in sequence; the residual module obtains output one by two causal dilation convolution layers, a batch normalization layer and a random dropout layer, obtains output two by 1×1 convolution, and combines output one and output two to obtain the output of the residual module; The causal dilated convolution layer uses a dilated convolution with a dilation coefficient to expand the receptive field:

[0029] in, is the input time series data, W is the convolution kernel weight, b is the bias term, is the output of the convolutional layer, is the activation function; The WaveNet model introduces gated recurrent units:

[0030] in, They are the update gate and the reset gate, is the Sigmoid function, is element-wise multiplication, is a hidden state; The output layer performs linear regression:

[0031] in, is the predicted value, and are the weight and bias of the output layer respectively.

[0032] As an implementable method, the method of constructing a hemostasis time-pressure prediction model by using a differential evolution algorithm and a WaveNet model includes: optimizing the weight of the WaveNet model by using a differential evolution algorithm, including: Randomly initialize individuals in the population, each individual represents a set of weights of the WaveNet model; Perform mutation operation on each individual to generate test individuals; Perform crossover operation on the test individuals and the original individuals to generate candidate solutions; A selection operation is performed between the original individuals and the candidate solutions, and individuals with higher fitness are selected to enter the next generation; until the loss function of the WaveNet model meets the requirements.

[0033] As an implementable embodiment, the hemostatic device includes a detection and control body and a tourniquet, and the detection and control body is provided with a blood pressure detection device, a heart rate monitoring device, a skin pressure detection device, a time module, a retractable device and / or a patient body data and patient wound data input device. One end of the tourniquet is inserted into the detection and control body to achieve a detachable connection with the detection and control body, and the other end is embedded in the retractable device. The retractable device adjusts the pressure between the hemostatic device and the skin by tightening or relaxing.

[0034] As an implementable embodiment, the outer side of the tourniquet is serrated, and the retracting and releasing device includes a motor and a rotor fixedly connected to the motor and matching the serrated shape of the tourniquet.

[0035] The beneficial effects of the present invention are as follows: the present invention collects blood pressure, heart rate, hemostatic pressure, patient body data and patient wound data through a hemostatic device, and constructs a hemostatic time-pressure prediction model through a differential evolution algorithm and a WaveNet model, thereby realizing the prediction of a relationship curve between hemostatic pressure and hemostatic time. Through the relationship curve, the hemostatic device can adjust the pressure applied by the hemostatic device as needed, thereby avoiding the problems of excessive hemostatic pressure of the tourniquet due to different human factors during the treatment of combat injuries, necrosis of limbs or compressed local areas due to excessive hemostatic time, and poor hemostasis due to insufficient pressure. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a schematic diagram of an intelligent hemostasis system of the present invention.

[0037] Figure 2 Schematic diagram of the hemostatic device of the present invention.

[0038] Figure 3 The present invention is a flow chart of an intelligent hemostasis method. DETAILED DESCRIPTION

[0039] The present invention is further described in detail below in conjunction with specific embodiments.

[0040] It should be noted that these embodiments are only used to illustrate the present invention rather than to limit the present invention. Simple improvements to the method based on the concept of the present invention all fall within the scope of protection claimed by the present invention.

[0041] See also Figure 1 , which is an intelligent hemostasis system, includes a data acquisition module 100, a data processing module 200, a model building module 300, a prediction module 400 and a hemostasis module 500.

[0042] The data acquisition module 100 is used to collect the patient's blood pressure, heart rate, and the pressure between the hemostatic device and the skin through the hemostatic device, and collect the patient's body data and the patient's wound data. The patient's body data includes age, gender, BMI, body fat, etc.; the patient's wound data includes part, area, depth, and amount of bleeding per unit time. The data collection process can be detected by the hemostatic device, or data can be input. Of course, relevant parameters such as the patient's blood flow rate and blood pressure at the wound can also be introduced.

[0043] The data processing module 200 is used to pre-process the collected data, construct the data of the patient's blood pressure, heart rate and the pressure between the hemostasis device and the skin into a time series based on the hemostasis time, and use the patient's body data and patient wound data as labels of the time series to form time series data.

[0044] Preprocessing includes: Each parameter is normalized and the original data is mapped to the interval [0, 1]. The formula is as follows:

[0045] in: Xnorm is the normalized data; X is the original data; Xmin is the minimum value in the original data set; Xmax is the maximum value in the original data set.

[0046] The purpose of normalization is to eliminate the dimensional effects between different parameters so that the contribution of each parameter to model training is at the same level.

[0047] Detect and process outliers. If the original value of a parameter exceeds the known reasonable range, it is corrected using linear interpolation or neighborhood-based methods.

[0048] The model building module 300 is used to construct a hemostasis time-pressure prediction model by using the time series data of successful hemostasis without excessive pressure and excessive hemostasis time through the differential evolution algorithm and the WaveNet model.

[0049] The WaveNet model consists of a two-dimensional convolution feature extraction layer, a one-dimensional sequence reorganization layer, two residual modules, a fully connected layer and an output layer in sequence; the residual module obtains output one through two causal dilated convolution layers, a batch normalization layer and a random inactivation layer, obtains output two through 1×1 convolution, and combines output one and output two to obtain the output of the residual module.

[0050] The causal dilated convolution layer uses a dilated convolution with a dilation coefficient to expand the receptive field:

[0051] in, is the input time series data, W is the convolution kernel weight, b is the bias term, is the output of the convolutional layer, is the activation function.

[0052] The WaveNet model introduces gated recurrent units:

[0053] in, They are the update gate and the reset gate, is the Sigmoid function, is element-wise multiplication, In hidden state.

[0054] The output layer performs linear regression:

[0055] in, is the predicted value, and are the weight and bias of the output layer respectively.

[0056] The WaveNet model can be used to explore the relationship between time series under different labels, that is, the appropriate hemostasis time and hemostasis pressure under different patient conditions and wound conditions.

[0057] The hemostasis time-pressure prediction model is constructed by using the differential evolution algorithm and the WaveNet model, including: using the differential evolution algorithm to optimize the weight of the WaveNet model, including: Randomly initialize individuals in the population, each individual represents a set of weights of the WaveNet model:

[0058] in, is the population size, i.e. the number of individuals; is the i-th weight parameter of the i-th individual.

[0059] Perform mutation operation on each individual to generate test individuals, cross the test individuals with the original individuals to generate candidate solutions:

[0060] in, is the jth parameter of the i-th candidate solution of the t+1th generation; are three different individuals randomly selected from the current population; F is the scaling factor that controls the difference weight.

[0061] The crossover operation is used to combine the mutated candidate solution with the current solution:

[0062] in, is the jth parameter of the i-th candidate solution after crossover; J is the randomly selected crossover index; CR is the crossover probability; To generate a random number between [0,1].

[0063] A selection operation is performed between the original individuals and the candidate solutions, and individuals with higher fitness are selected to enter the next generation; until the loss function of the WaveNet model meets the requirements or the maximum number of iterations is reached.

[0064] The prediction module 400 is used to obtain the relationship curve between the pressure between the hemostatic device and the skin and the hemostatic time by inputting the patient's blood pressure, heart rate, patient's body data and patient's wound data through the hemostatic time-pressure prediction model. By inputting the patient's blood pressure, heart rate, patient's body data and patient's wound data through the trained hemostatic time-pressure prediction model, the appropriate hemostatic pressure for a period of time in the future, that is, the relationship curve between the pressure between the hemostatic device and the skin and the hemostatic time can be obtained. The hemostasis module 500 is used to control the hemostasis device to adjust the pressure between the hemostasis device and the skin according to the relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time. Through the above process, the pressure applied by the hemostasis device can be adjusted as needed based on the hemostasis time, so as to avoid the problem of excessive hemostasis pressure of the tourniquet during the treatment of combat injuries, necrosis of the limb or the compressed part due to too long hemostasis time, and poor hemostasis due to too little pressure.

[0065] A monitoring module may also be provided to monitor the implemented hemostasis time and hemostasis pressure, and to issue an alarm when the duration and pressure reach a level that may cause a wound or physical injury.

[0066] See also Figure 2 The hemostatic device can be a detection and control body and a tourniquet. The detection and control body is provided with a blood pressure detection device, a heart rate monitoring device, a skin pressure detection device, a time module (for synchronizing data time), a retractable device and / or a patient body data and patient wound data input device. One end of the tourniquet is inserted into the detection and control body to achieve a detachable connection with the detection and control body, and the other end is embedded in the retractable device. The retractable device adjusts the pressure between the hemostatic device and the skin by tightening or relaxing.

[0067] The outer side of the tourniquet is serrated, and the retracting and releasing device includes a motor and a rotor fixedly connected to the motor and matching the serrated shape of the tourniquet. The motor drives the rotor to rotate to achieve tightening or loosening of the tourniquet.

[0068] See also Figure 3, a smart hemostasis method, including: S100, collecting the patient's blood pressure, heart rate, and pressure between the hemostatic device and the skin through the hemostatic device, and collecting the patient's body data and the patient's wound data.

[0069] S200, preprocessing the collected data, constructing the data of the patient's blood flow rate, bleeding volume, blood pressure, heart rate, and the pressure between the hemostasis device and the skin into a time series based on the hemostasis time, and using the patient's body data and the patient's wound data as labels of the time series to form time series data.

[0070] S300, using the time series data of successful hemostasis without excessive pressure and prolonged hemostasis time, a hemostasis time-pressure prediction model was constructed through the differential evolution algorithm and the WaveNet model.

[0071] S400, using a hemostasis time-pressure prediction model, inputting the patient's blood pressure, heart rate, patient body data and patient wound data to obtain a relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time.

[0072] S500, controlling the hemostatic device to adjust the pressure between the hemostatic device and the skin according to a curve of the relationship between the pressure between the hemostatic device and the skin and the hemostatic time.

[0073] Among them, the patient's physical data includes age, gender, BMI, and body fat; the patient's wound data includes part, area, depth, and amount of bleeding per unit time.

[0074] Among them, the WaveNet model is composed of a two-dimensional convolution feature extraction layer, a one-dimensional sequence reorganization layer, two residual modules, a fully connected layer and an output layer in sequence; the residual module obtains output one through two causal dilated convolution layers, a batch normalization layer and a random inactivation layer, obtains output two through a 1×1 convolution, and combines output one and output two to obtain the output of the residual module.

[0075] The causal dilated convolution layer uses a dilated convolution with a dilation factor to expand the receptive field:

[0076] in, is the input time series data, W is the convolution kernel weight, b is the bias term, is the output of the convolutional layer, is the activation function.

[0077] The WaveNet model introduces gated recurrent units:

[0078] in, They are the update gate and the reset gate, is the Sigmoid function, is element-wise multiplication, In hidden state.

[0079] The output layer performs linear regression:

[0080] in, is the predicted value, and are the weight and bias of the output layer respectively.

[0081] As an implementable method, the method of constructing a hemostasis time-pressure prediction model by using a differential evolution algorithm and a WaveNet model includes: optimizing the weight of the WaveNet model by using a differential evolution algorithm, including: Randomly initialize individuals in the population, each individual represents a set of weights of the WaveNet model; Perform mutation operation on each individual to generate test individuals; Perform crossover operation on the test individuals and the original individuals to generate candidate solutions; A selection operation is performed between the original individuals and the candidate solutions, and individuals with higher fitness are selected to enter the next generation; until the loss function of the WaveNet model meets the requirements.

[0082] Among them, the hemostatic device includes a detection and control body and a tourniquet. The detection and control body is provided with a blood pressure detection device, a heart rate monitoring device, a skin pressure detection device, a time module, a retractable device and / or a patient body data and patient wound data input device. One end of the tourniquet is inserted into the detection and control body to achieve a detachable connection with the detection and control body, and the other end is embedded in the retractable device. The retractable device adjusts the pressure between the hemostatic device and the skin by tightening or relaxing.

[0083] The outer side of the tourniquet is serrated, and the retracting and releasing device comprises a motor and a rotor fixedly connected to the motor and matching the serrated shape of the tourniquet.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described with reference to the preferred embodiments of the present invention, it should be understood by those skilled in the art that various changes may be made in form and details without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. An intelligent hemostasis system, characterized in that: It includes a data acquisition module, a data processing module, a model building module, a prediction module and a hemostasis module; The data acquisition module is used to collect the patient's blood pressure, heart rate, and pressure between the hemostatic device and the skin through the hemostatic device, and collect the patient's body data and the patient's wound data; The data processing module is used to pre-process the collected data, construct the data of the patient's blood pressure, heart rate, and the pressure between the hemostasis device and the skin into a time series based on the hemostasis time, and use the patient's body data and the patient's wound data as labels of the time series to form time series data; The model building module is used to use time series data of successful hemostasis without excessive pressure and excessive hemostasis time to build a hemostasis time-pressure prediction model through a differential evolution algorithm and a WaveNet model; The prediction module is used to input the patient's blood pressure, heart rate, patient body data and patient wound data through the hemostasis time-pressure prediction model to obtain a relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time; The hemostasis module is used to control the hemostasis device to adjust the pressure between the hemostasis device and the skin according to a relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time.

2. The intelligent hemostasis system according to claim 1, characterized in that: The patient's physical data includes age, gender, BMI, and body fat; the patient's wound data includes part, area, depth, and amount of bleeding per unit time.

3. The intelligent hemostasis system according to claim 1, characterized in that: The WaveNet model is composed of a two-dimensional convolution feature extraction layer, a one-dimensional sequence reorganization layer, two residual modules, a fully connected layer and an output layer in sequence; the residual module obtains output one by two causal dilation convolution layers, a batch normalization layer and a random dropout layer, obtains output two by 1×1 convolution, and combines output one and output two to obtain the output of the residual module; The causal dilated convolution layer uses a dilated convolution with a dilation coefficient to expand the receptive field: in, is the input time series data, W is the convolution kernel weight, b is the bias term, is the output of the convolutional layer, is the activation function; The WaveNet model introduces gated recurrent units: in, They are the update gate and the reset gate, is the Sigmoid function, is element-wise multiplication, is a hidden state; The output layer performs linear regression: in, is the predicted value, and are the weight and bias of the output layer respectively.

4. The intelligent hemostasis system according to claim 3, characterized in that: The method of constructing a hemostasis time-pressure prediction model by using a differential evolution algorithm and a WaveNet model includes: optimizing the weight of the WaveNet model by using a differential evolution algorithm, including: randomly initializing individuals in a population, each individual representing a set of weights of the WaveNet model; Perform mutation operation on each individual to generate test individuals; Perform crossover operation on the test individuals and the original individuals to generate candidate solutions; A selection operation is performed between the original individuals and the candidate solutions, and individuals with higher fitness are selected to enter the next generation; until the loss function of the WaveNet model meets the requirements.

5. The intelligent hemostasis system according to claim 4, characterized in that: The hemostatic device includes a detection and control body and a tourniquet. The detection and control body is provided with a blood pressure detection device, a heart rate monitoring device, a skin pressure detection device, a time module, a retractable device and / or a patient body data and patient wound data input device. One end of the tourniquet is inserted into the detection and control body to achieve a detachable connection with the detection and control body, and the other end is embedded in the retractable device. The retractable device adjusts the pressure between the hemostatic device and the skin by tightening or relaxing.

6. The intelligent hemostasis system according to claim 5, characterized in that: The outer side of the tourniquet is serrated, and the retracting and releasing device comprises a motor and a rotor fixedly connected to the motor and matching the serrated shape of the tourniquet.

7. An intelligent hemostasis method, characterized in that: include: Collecting the patient's blood pressure, heart rate, and pressure between the hemostatic device and the skin through the hemostatic device, as well as the patient's physical data and the patient's wound data; Preprocessing the collected data, constructing the data of the patient's blood pressure, heart rate, and the pressure between the hemostasis device and the skin into a time series based on the hemostasis time, and using the patient's body data and the patient's wound data as labels of the time series to form time series data; Using the time series data of successful hemostasis without excessive pressure and prolonged hemostasis time, a hemostasis time-pressure prediction model was constructed using the differential evolution algorithm and the WaveNet model. Through the hemostasis time-pressure prediction model, the patient's blood pressure, heart rate, patient body data and patient wound data are input to obtain the relationship curve between the pressure between the hemostasis device and the skin and the hemostasis time; According to the relationship curve between the pressure between the hemostatic device and the skin and the hemostatic time, the hemostatic device is controlled to adjust the pressure between the hemostatic device and the skin.

8. The intelligent hemostasis method according to claim 7, characterized in that: The WaveNet model is composed of a two-dimensional convolution feature extraction layer, a one-dimensional sequence reorganization layer, two residual modules, a fully connected layer and an output layer in sequence; the residual module obtains output one by two causal dilation convolution layers, a batch normalization layer and a random dropout layer, obtains output two by 1×1 convolution, and combines output one and output two to obtain the output of the residual module; The causal dilated convolution layer uses a dilated convolution with a dilation coefficient to expand the receptive field: in, is the input time series data, W is the convolution kernel weight, b is the bias term, is the output of the convolutional layer, is the activation function; The WaveNet model introduces gated recurrent units: in, They are the update gate and the reset gate, is the Sigmoid function, is element-wise multiplication, is a hidden state; The output layer performs linear regression: in, is the predicted value, and are the weight and bias of the output layer respectively.

9. The intelligent hemostasis method according to claim 8, characterized in that: The method of constructing a hemostasis time-pressure prediction model by using a differential evolution algorithm and a WaveNet model includes: optimizing the weight of the WaveNet model by using a differential evolution algorithm, including: randomly initializing individuals in a population, each individual representing a set of weights of the WaveNet model; Perform mutation operation on each individual to generate test individuals; Perform crossover operation on the test individuals and the original individuals to generate candidate solutions; A selection operation is performed between the original individuals and the candidate solutions, and individuals with higher fitness are selected to enter the next generation; until the loss function of the WaveNet model meets the requirements.

10. The intelligent hemostasis system according to claim 7, characterized in that: The hemostatic device includes a detection and control body and a tourniquet. The detection and control body is provided with a blood pressure detection device, a heart rate monitoring device, a skin pressure detection device, a time module, a retractable device and / or a patient body data and patient wound data input device. One end of the tourniquet is inserted into the detection and control body to achieve a detachable connection with the detection and control body, and the other end is embedded in the retractable device. The retractable device adjusts the pressure between the hemostatic device and the skin by tightening or relaxing.

Images