Wearable gout ultrasound treatment system based on artificial intelligence algorithm

Abstract

The present invention discloses a wearable gout ultrasonic treatment system based on an artificial intelligence algorithm, comprising: a data acquisition module, used to obtain historical gout case samples of a target object; a feature extraction module, used to extract feature data of the target object before and after the use of an ultrasonic prescription from the historical gout case samples, wherein the ultrasonic prescription refers to ultrasonic energy generated by controlling the amplitude and frequency of a wearable ultrasonic treatment device; a prediction training module, used to input the extracted feature data into a fully connected neural network for training, to obtain a trained fully connected neural network, which can predict blood growth rate and pain values ​​under different ultrasonic prescriptions; and an optimization output module, used to select the best target ultrasonic prescription result adapted to each target object according to the predicted blood growth rate and pain value.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a wearable gout ultrasonic treatment system based on an artificial intelligence algorithm. Background Art

[0002] Gout is a disease caused by uric acid metabolism disorder, which is caused by long-term deposition of uric acid crystals in joints and surrounding tissues. When these crystals move or compress surrounding nerves and soft tissues, they can cause severe pain. The location, area and depth of uric acid crystals in joints vary from patient to patient, and are irregular and uncertain.

[0003] The current treatment methods for gout are as follows: 1. Drug treatment and infusion therapy, which are quick and effective, but cannot fundamentally treat the disease and are prone to side effects; 2. Traditional Chinese medicine treatment of gout, mainly including oral Chinese medicine, external application, acupoint application, acupuncture, etc., but its disadvantage is that it is slow to take effect. 3. Surgical treatment, which is prone to postoperative complications.

[0004] Current ultrasound technology is very sensitive to gout crystals. Traditional ultrasound therapeutic devices only have control but no real-time effect monitoring, and are not intelligent enough to give the best ultrasound prescription suitable for an individual. At the same time, they are single, have poor adaptability, and poor treatment effect. Therefore, there is an urgent need for a wearable gout ultrasound treatment system based on artificial intelligence algorithms. Summary of the invention

[0005] The purpose of the present invention is to provide a wearable gout ultrasonic treatment system based on an artificial intelligence algorithm, which improves the treatment efficiency and adaptability of the ultrasonic therapeutic instrument.

[0006] The present invention provides a wearable gout ultrasound treatment system based on an artificial intelligence algorithm, comprising:

[0007] A data collection module is used to obtain historical gout case samples of the target subjects;

[0008] A feature extraction module, used to extract feature data of the target object before and after the use of the ultrasound prescription from the historical gout case samples, wherein the ultrasound prescription refers to the ultrasonic energy generated by controlling the amplitude and frequency of the wearable ultrasonic treatment device;

[0009] A prediction training module is used to input the extracted feature data into a fully connected neural network for training, so as to obtain a trained fully connected neural network that can predict blood velocity growth rate and pain value under different ultrasound prescriptions;

[0010] The optimization output module is used to select the best target ultrasound prescription result suitable for each target object according to the predicted blood growth rate and pain value.

[0011] Preferably, the ultrasound prescription includes ultrasound array energy focusing type, energy focusing depth and energy action surface, wherein the energy focusing type includes preset interval point shape, preset interval point-connected point shape, preset interval continuous shape and random scattered point shape; the energy focusing depth includes the energy focusing depth generated by the superposition of energy generated by one or more ultrasonic vibration sources; the energy action surface includes a plane, a convex surface, a concave surface, a curved surface and a surface adapted to the target position.

[0012] Preferably, the feature extraction module further includes:

[0013] The characteristic data corresponding to the historical gout case samples before and after the use of ultrasonic prescriptions include the target object's region, gender, age, blood flow rate, blood oxygen content, type of ultrasonic prescription, time / number of times the ultrasonic prescription is used, blood growth rate and pain value, among which the types of ultrasonic prescriptions include sine waves, square waves, triangle waves and sound waves related to the range, area and depth of ultrasonic energy.

[0014] Preferably, the feature extraction module further includes:

[0015] The label setting module is used to obtain the initial blood flow rate, blood oxygen content and pain value before the ultrasonic prescription is applied; obtain the blood flow rate, blood oxygen content and pain value at a preset time after the ultrasonic prescription is applied; calculate the difference between the initial blood flow rate, blood oxygen content and pain value and the blood flow rate, blood oxygen content and pain value after the application, and obtain the blood flow rate growth rate / blood growth rate and pain reduction rate / pain value after the application based on the difference information;

[0016] The data preprocessing module normalizes the data set and actual result labels obtained in step 1, and limits the actual input data value of the input layer and the label data of the output layer to between 0 and 1.

[0017] Preferably, the prediction training module further includes:

[0018] Construct a fully connected neural network, X = (x1, ..., x n ) is the actual input vector of the first input layer, H = (h1..., h m ) is used as the output vector of the second hidden layer, Y = (y1, y2) is used as the predicted output result of the third output layer, and A = (a ij ) n×m is the weight matrix from the input layer to the hidden layer, B=(b jk ) m×2 is the weight matrix from the hidden layer to the output layer, C=(c1,...,c m) is the bias vector of the hidden layer, D = (d1, d2) is the bias vector of the output layer; configure the ReLU function as the activation function of the hidden layer; configure the Sigmoid function as the activation function of the output layer; the maximum number of iterative training is max_epochs; the error threshold for stopping training is ε; the label of the network is Y o =(r1, r2);

[0019] Among them, x1…x n represents the extracted feature data; n is the number of neurons in the input layer; m is the number of neurons in the hidden layer; y1 and y2 represent the predicted blood growth rate and pain value respectively, a ij is the weight parameter value from the i-th neuron in the input layer to the j-th neuron in the hidden layer; b jk is the weight parameter value from the jth neuron in the hidden layer to the kth neuron in the output layer. r1 and r2 represent the actual blood growth rate and pain value, respectively.

[0020] Preferably, the prediction training module further includes:

[0021] The network is initialized. The initialization parameters include the weight matrix A from the input layer to the hidden layer; the weight matrix B from the hidden layer to the output layer; the bias vector C of the hidden layer; the bias vector D of the output layer; and the learning rate is set to 1e -4 ;

[0022] Calculate the output of the hidden layer. The calculation method is as follows In the formula, R(·) is the activation function of the hidden layer. The ReLU function is selected as the activation function of the hidden layer. Its function expression is R(t)=max(0, t), where t represents the input of the activation function, R(t) represents the output of the activation function, and a pj is the weight parameter value from the pth neuron in the input layer to the jth neuron in the hidden layer; x p Refers to the characteristic element of the input, c j is the bias vector of the hidden layer;

[0023] Calculate the predicted output of the output layer. The calculation method is as follows In the formula, P(·) is the activation function of the output layer. Sigmoid function is selected as the activation function of the output layer, and its function definition is: In the formula, z represents the input of the activation function, P(z) represents the output of the activation function, and d k is the bias vector of the output layer. The Sigmoid function limits the output to the range of 0 to 1. pk is the weight parameter value from the pth neuron in the hidden layer to the kth neuron in the output layer;

[0024] The mean square error between the predicted result Y and the actual label value Y0 is used as the loss function for training the fully connected neural network. The expression of the loss function is:

[0025] The parameters of the fully connected neural network are updated according to the adaptive moment estimation algorithm, and the learning rate is set to 1e -4 , the maximum number of iterative training times max_epochs is 1000, and back propagation is performed through the adaptive estimation algorithm to update the weight matrix A from the input layer to the hidden layer, the weight matrix B from the hidden layer to the output layer, the bias vector C of the hidden layer, and the bias vector D of the output layer;

[0026] Determine whether the training process is over. If |L(Y, Y o )|>ε, then return to network initialization and continue training and solving; if |L(Y, Y0)|≤ε, then end the training process; if the current training times reaches the maximum training times max_epochs, then end the training process;

[0027] Determine updated network parameters and save the network parameters after training, that is, determine the weight matrix A from the input layer to the hidden layer, the weight matrix B from the hidden layer to the output layer, the bias vector C of the hidden layer, and the bias vector D of the output layer; obtain the trained fully connected neural network according to the above steps, predict the blood growth rate and pain value through the updated network, and select the best ultrasound prescription and usage time according to the prediction results.

[0028] Preferably, the optimization output module comprises:

[0029] Obtaining the blood growth rate and the duration at the blood growth rate according to the prediction result corresponding to the ultrasound prescription, and generating a statistical list, wherein the statistical list includes the use time of the ultrasound prescription, the ultrasound prescription ID, the average blood growth rate, the use duration, and the blood increment that changes with the use duration;

[0030] The ultrasound prescription treatment effect score is obtained based on the blood flow in the statistical list, and after sorting by the score, the data can be tracked throughout the process based on the ultrasound prescription ID;

[0031] The therapeutic effect score of each ultrasound prescription = the blood increment I NT that changes with the duration of use, I NT = average blood growth rate * duration of use.

[0032] Preferably, the optimization output module comprises:

[0033] Detecting the blood flow rate of the target subject before using the ultrasound prescription by using a wearable ultrasound therapeutic device;

[0034] Download and update the corresponding AI ultrasound prescription from the server, where the AI ​​ultrasound prescription is a target ultrasound prescription adapted to the target object after network prediction;

[0035] Determine whether to select AI ultrasound prescription or default ultrasound prescription. If the default ultrasound prescription is selected, the default M sets of ultrasound prescriptions are used, and the first ultrasound prescription list ID = M sets of default ultrasound prescriptions. Alternatively, if AI ultrasound prescription is selected, determine whether N sets of AI ultrasound prescriptions exist. If so, N sets of AI ultrasound prescriptions are used, and the ultrasound prescription list ID = N sets of AI ultrasound prescriptions.

[0036] Select the ultrasound prescription ID corresponding to the scheme in the current ultrasound prescription list, execute the ultrasound operation corresponding to the ID, continuously monitor the blood flow rate and pain value of the target object to generate monitoring results, and determine whether to interrupt the current operation and select the ultrasound prescription ID according to the monitoring results;

[0037] When the current operation is interrupted, the monitored ultrasound prescription ID, blood flow rate, pain value and interruption mark are uploaded to the server, and the plan in the list is reselected, and the blood flow rate and pain value are continuously monitored to continue to determine whether it is interrupted;

[0038] If the current operation is not interrupted, it is determined whether the current plan is ended. If so, the currently monitored data is uploaded to the server and the ultrasound prescription is updated;

[0039] Traverse the ultrasound prescription list ID until the operation is completed, and calculate the treatment effect score under the current ultrasound prescription based on the maximum blood growth rate and duration.

[0040] Preferably, the wearable ultrasound treatment device comprises:

[0041] A flexible circuit substrate, on which a plurality of ultrasonic transducer units are distributed to form ultrasonic arrays of different combinations;

[0042] An ultrasonic transducer unit, wherein the ultrasonic transducer unit is connected to the flexible circuit substrate, and the ultrasonic transducer unit comprises a plurality of ultrasonic transducers and an ultrasonic transducer fixing sleeve adapted to the shape of the ultrasonic transducer, wherein an electrode is arranged on the inner side of the ultrasonic transducer fixing sleeve, and the electrode is connected to the flexible circuit substrate, and at the same time, the electrodes of each ultrasonic transducer are gathered to a bus junction box on the flexible circuit substrate and are connected to an ultrasonic host; the ultrasonic transducer comprises an ultrasonic transducer body, and a through hole is arranged on the upper surface of the ultrasonic transducer body, and when the ultrasonic transducer fixing sleeve is arranged on the ultrasonic transducer and clamped with each other, the ultrasonic transducer body is connected to the electrode, and an anti-fool structure is arranged on the ultrasonic transducer body.

[0043] The present invention also provides a readable storage medium, on which a processing program is stored, and when the processing program is executed by a processor, the following steps are implemented: obtaining historical gout case samples of the target object; extracting characteristic data of the target object before and after the use of an ultrasonic prescription from the historical gout case samples, wherein the ultrasonic prescription refers to ultrasonic energy generated by controlling the amplitude and frequency of a wearable ultrasonic treatment device; inputting the extracted characteristic data into a fully connected neural network for training to obtain a trained fully connected neural network that can predict blood growth rate and pain values ​​under different ultrasonic prescriptions; and selecting the best target ultrasonic prescription result that is suitable for each target object based on the predicted blood growth rate and pain value.

[0044] With respect to the prior art, the present invention has the following beneficial effects:

[0045] The wearable gout ultrasonic treatment system based on artificial intelligence algorithm provided by the present invention, in the face of the irregularity of uric acid crystals and the uncertainty of their position in the joints, uses a variety of changing combinations of ultrasonic prescriptions as the treatment basis of gout ultrasonic treatment instrument. With artificial intelligence models and algorithms as the core, it subverts the shortcomings of traditional ultrasonic treatment instruments that only have control but no real-time effect monitoring, and are not intelligent enough to give the best therapeutic ultrasonic prescription suitable for individuals. At the same time, it overcomes the characteristics of being single, poorly adaptable, and having poor therapeutic effects. It has stronger pertinence and adaptability for uric acid crystals in gout patients, predicts the best treatment plan, and greatly improves the gout treatment effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a principle block diagram of a wearable gout ultrasound treatment system based on an artificial intelligence algorithm according to an embodiment of the present invention;

[0047] Figure 2-3 This is a schematic diagram of the structure of an ultrasonic transducer unit in one embodiment of the present invention;

[0048] Figure 4 A schematic diagram of a combined ultrasonic array in one embodiment of the present invention;

[0049] Figure 5 This is a diagram of a fully connected neural network architecture in one embodiment of the present invention;

[0050] Figure 6 This is a workflow diagram of a wearable gout ultrasound treatment system based on an artificial intelligence algorithm in one embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0053] It should be noted that the concepts such as "once" and "second" mentioned in the present application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0054] It should be noted that the modifications of "one" and "plurality" mentioned in the disclosure of this application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0055] Embodiment 1

[0056] like Figure 1 As shown, the present invention provides a wearable gout ultrasound treatment system based on artificial intelligence algorithm, comprising:

[0057] A data collection module is used to obtain historical gout case samples of the target subjects;

[0058] The feature extraction module is used to extract feature data of the target object before and after the use of the ultrasonic prescription from the historical gout case samples. The ultrasonic prescription refers to the ultrasonic energy generated by controlling the amplitude and frequency of the wearable ultrasonic therapeutic device; the wearable ultrasonic therapeutic device used in this embodiment adopts an ultrasonic multi-array source design, and its internal circuit is integrated and monitored by remote control through a PC or APP, so as to achieve wearability. Under the control of the ultrasonic host, the frequency and amplitude of each array source are controlled and driven to form an array waveform, and a variety of array energy focusing types, focusing depths and energy action surfaces are generated to continuously physically treat the gout affected area. These elements constitute the most basic ultrasonic model elements of the gout ultrasonic therapeutic device, which are referred to as ultrasonic prescriptions below; based on the most basic ultrasonic model elements, the coherence of the wave is fully utilized. Under the premise that the ultrasonic host controls the drive to generate the same frequency and the same vibration direction, the regional wave energy is superimposed and weakened, which can cope with the individuality and irregularity of the affected area, so that the ultrasonic prescription has more variability and adjustability, and constitutes the multi-level strong and weak combination characteristics of the ultrasonic prescription, which is another element of the ultrasonic basic model.

[0059] A prediction training module is used to input the extracted feature data into a fully connected neural network for training, so as to obtain a trained fully connected neural network that can predict blood velocity growth rate and pain value under different ultrasound prescriptions;

[0060] The optimization output module is used to select the best target ultrasound prescription result suitable for each target object according to the predicted blood growth rate and pain value.

[0061] In this embodiment, a method for predicting blood growth rate and pain value of gout-affected parts of the human body based on ultrasonic prescriptions using a fully connected neural network is used. By utilizing a fully connected neural network, a data-driven machine learning method is used to predict blood growth rate and pain value based on historical gout case samples. The sample content includes personal data, such as region, gender, age, current blood flow rate, current blood oxygen content and ultrasonic prescription data, such as prescription, prescription usage time, etc., and the problem is reduced to a regression problem. The main processing process is: in the model training stage, features are extracted from personal information, human body detection and historical data of ultrasonic prescriptions, a fully connected neural network is trained, and a gradient descent algorithm is executed to find the optimal parameters; in the model inference stage, the trained fully connected neural network is used to predict blood growth rate and pain value according to the prescription, and the predicted blood growth rate and pain value are used to select the ultrasonic prescription with the best effect.

[0062] In various embodiments of the present application, the wearable ultrasonic therapeutic device includes: a flexible circuit substrate 3, on which a plurality of ultrasonic transducer units are distributed to form ultrasonic arrays of different combinations; an ultrasonic transducer unit, which is connected to the flexible circuit substrate 3, and includes a plurality of ultrasonic transducers 1 and an ultrasonic transducer fixing sleeve 8 whose shape is adapted to the ultrasonic transducer 1, an electrode 8-1 is provided on the inner side of the ultrasonic transducer fixing sleeve 8, and the electrode 8-1 is connected to the flexible circuit substrate 3. At the same time, the electrode 8-1 of each ultrasonic transducer 1 is converged to the bus junction box 2 on the flexible circuit substrate 3 and is connected to the ultrasonic host; the ultrasonic transducer 1 includes an ultrasonic transducer body, and a through hole 9 is provided on the upper surface of the ultrasonic transducer body. When the ultrasonic transducer fixing sleeve 8 is sleeved on the ultrasonic transducer 1 and clamped with each other, the electrode 12 passes through the through hole 9, the ultrasonic transducer body is connected to the electrode 12, and an anti-fool structure is provided on the ultrasonic transducer body.

[0063] Those skilled in the art will appreciate that the ultrasonic transducer unit used in this embodiment is as follows: Figure 2-3 As shown, the design includes an anti-stuck hole 10 and a transducer electrode 12. When the ultrasonic transducer 1 is clamped with the suction box 13, the ultrasonic transducer unit is connected with the electrode 12 on the transducer fixing sleeve 8. A magnetic ring 11 is also provided on the ultrasonic transducer body. The wearable combined ultrasonic array uses a flexible circuit substrate 3, which can adapt to different target positions and execute corresponding ultrasonic prescriptions. Figure 4As shown, several ultrasonic transducer fixing sleeves 8 are designed on the flexible circuit substrate 3. For example, taking a 3*8 combined ultrasonic array as an example, ultrasonic transducers can be inserted arbitrarily according to the size of the gout surface, or ultrasonic transducers 1 can be partially not inserted to form physical ultrasonic arrays of different combinations; the ultrasonic transducer fixing sleeve 8 adopts an anti-stupefaction snap-on design, and the flexible circuit substrate 3 is connected through the transducer matching electrode 12 on the ultrasonic transducer fixing sleeve 8, and at the same time, the electrode 12 of each ultrasonic transducer unit is gathered to the bus junction box 2 of the substrate to connect the ultrasonic transducer fixing sleeve with the ultrasonic host; each ultrasonic transducer unit is an independent individual, and the ultrasonic array is flexibly combined as needed, which is more convenient to adapt to the irregularity of the affected part of the gout patient and improves the therapeutic effect of the ultrasonic prescription; the ultrasonic host automatically identifies the array combination model of the working ultrasonic array, which is convenient for the ultrasonic host to drive the ultrasonic combination array, and generate the corresponding ultrasonic energy and waveform characteristics according to the ultrasonic prescription, so as to realize the flexible combination of the ultrasonic array change and the matching correspondence with the ultrasonic prescription. If an ultrasonic array that does not match the ultrasonic prescription appears, the ultrasonic host will issue a prompt in time; if the ultrasonic prescription is manually enforced, on the one hand, the host only drives the installed ultrasonic transducer unit, and if the ultrasonic transducer unit is not installed, it will not drive the protective therapeutic device. In addition, the treatment data will be marked and will not be processed as AI data. Hot-swappable ultrasonic transducer unit: The transducer fixing sleeve is designed with an anti-fool mouth 10 and an electrode 12 to match, which realizes the hot-swappable ultrasonic transducer unit and improves flexibility; at the same time, it supports the timely increase, decrease or replacement of the ultrasonic transducer 1. A bus junction box 2 is designed on the substrate to connect a group of wires to the ultrasonic host, which improves the wearability of the therapeutic device and greatly reduces the number of connecting wires caused by the multiple ultrasonic transducers 1 in the ultrasonic array being connected to the host, thereby reducing the occurrence of faults. In the face of the uncertain surface of the gout affected area, the ultrasonic multi-array design of this embodiment adopts a modular design of the ultrasonic array. According to the size of the surface of the patient's gout affected area, an ultrasonic multi-array covering the surface of the gout affected area is combined through ultrasonic array units, which greatly improves the flexibility and adaptability of the ultrasonic array. See. Figure 4 As shown, the first layer is a wearable ultrasonic therapeutic device, and the second and third layers are two forms of focusing formed by an ultrasonic array. There may also be other ultrasonic focusing forms to form different ultrasonic prescriptions; wherein, the wearable ultrasonic therapeutic device of the first layer includes an ultrasonic transducer 1, a bus junction box 2, and a flexible circuit substrate 3, which adopts an ultrasonic transducer fixing flexible patch and a transducer fixing sleeve 8, and multiple ultrasonic transducers 1 and bus junction boxes 2 are arranged on the ultrasonic transducer fixing flexible patch; a layer of ultrasonic focusing focal spot 4 is distributed on a focusing layer 5 formed by the second layer of ultrasonic array; and two layers of ultrasonic focusing focal spots 6 are distributed on a second focusing layer 7 formed by the third layer of ultrasonic array.

[0064] Since gout is a disease caused by uric acid metabolism disorder, its main characteristics are arthritis and the deposition of uric acid crystals in joints and surrounding tissues. Usually, urate crystal stones are formed in joints and move in the joints or compress the surrounding nerves and soft tissues, which will cause severe pain. Since the pain site, pain intensity, number of attacks, etc. of gout patients vary from person to person, even in the same pain site, the site, degree and shape of urate crystals are different. The wearable ultrasonic therapeutic device used in this embodiment drives each array source parameter to produce a specified amplitude and frequency, and forms a specific energy focusing type, energy focusing depth and energy action surface for the gout affected area, so that the ultrasonic energy and coverage can reach the affected area and depth, promote the decomposition of uric acid crystals, and increase blood flow rate to excrete uric acid.

[0065] In various embodiments of the present application, the ultrasound prescription includes the ultrasound array energy focusing type, energy focusing depth and energy action surface, wherein the energy focusing type includes preset interval point shape, preset interval point-connected point shape, preset interval continuous shape and random scattered point shape; the energy focusing depth includes the energy focusing depth generated by the superposition of energy generated by one or more ultrasonic vibration sources; the energy action surface includes a plane, a convex surface, a concave surface, a curved surface and a surface adapted to the target position. It can be understood that the ultrasonic therapeutic apparatus used in this embodiment generates energy by controlling the amplitude and frequency, including various changes in energy focusing type, energy focusing action surface, and energy focusing depth, and forms the following energy focusing types for the gout affected area: energy focusing types are divided into: point-shaped with a certain interval, point-connected with a certain interval, continuous with a certain interval, and randomly scattered points; the energy focusing action surface is divided into: plane, convex surface, concave surface, arc surface, etc.; the energy focusing depth is the energy focusing depth generated by the superposition of energy generated by a single ultrasonic vibration source or multiple ultrasonic vibration sources; by designing this form of characteristic energy, the advantage is that it can not only produce continuous energy effect on uric acid crystals in irregular and uncertain joints and surrounding tissues, but also accelerate the blood flow around the gout affected area and the degree of decomposition and output of uric acid crystals, and achieve the characteristics of wide energy action range, strong continuity and no restrictions on uric acid crystals, which is very suitable for various joints with gout uric acid crystals. With the continuous use of the gout ultrasonic therapeutic apparatus, uric acid crystals and tophi can be controlled within a very small range or completely decomposed, thereby controlling the onset of gout.

[0066] In various embodiments of the present application, the feature extraction module also includes: the feature data before and after the use of the ultrasonic prescription corresponding to the historical gout case sample includes the target object's location, gender, age, blood flow rate, blood oxygen content, ultrasonic prescription type, ultrasonic prescription use time / number of times, blood growth rate and pain value, wherein the ultrasonic prescription type includes sine wave, square wave, triangle wave and sound waves related to the range, area and depth of ultrasonic energy. The pain value collected in this embodiment is the gout pain value obtained by the brain wave regulated by the peripheral nerve, and the blood flow rate is obtained under the ultrasonic prescription on the gout affected part, which is combined into a treatment result model of the artificial intelligence algorithm of the gout ultrasonic therapeutic instrument. The waveform of the ultrasonic prescription used in this embodiment is divided into: sine wave, square wave, triangle wave and other forms of waveforms; sine wave: moderate sharpness; square wave: strong persistence; triangle wave: large sharpness, different waveform designs can better match the different ranges, areas and depths of gout affected parts, improving the treatment efficiency and adaptability of the ultrasonic therapeutic instrument.

[0067] In various embodiments of the present application, the feature extraction module also includes: a label setting module, which is used to obtain the initial blood flow rate, blood oxygen content and pain value before the ultrasonic prescription takes effect; obtain the blood flow rate, blood oxygen content and pain value at a preset time after the ultrasonic prescription takes effect; calculate the difference between the initial blood flow rate, blood oxygen content and pain value and the blood flow rate, blood oxygen content and pain value after the effect, and derive the blood flow rate growth rate / blood growth rate and pain decrease rate / pain value after the effect based on the difference information; a data preprocessing module, which normalizes the data set obtained in the first step and the actual result label, and limits the actual input data value of the input layer and the label data of the output layer to between 0 and 1.

[0068] The method for predicting the blood growth rate and pain value of the gout affected area based on the ultrasonic prescription of the fully connected neural network in this embodiment includes the following steps: obtaining the data set required by the network: first, for each patient using the gout ultrasonic therapeutic device, obtain the patient's region, gender, and age information, and obtain the type of ultrasonic prescription used by the patient and the number of times / time the patient has used the ultrasonic prescription. Before each ultrasonic prescription is taken, measure the patient's initial blood flow rate, blood oxygen content, and pain value, and measure the patient's blood flow rate and pain value at a fixed time after the ultrasonic prescription is taken, calculate the blood growth rate and pain value, and thus obtain a nine-tuple {region, gender, age, blood flow rate, blood oxygen content, prescription type, prescription use time, blood growth rate, pain value}. Take the first seven items as the input data X of the network, and the last two items as the output data labels of the network. Each time the patient uses the ultrasonic prescription, repeatedly sample the above data to obtain a large number of data sets, which are divided into training sets and test sets for network training and testing;

[0069] Construct a fully connected neural network. The fully connected neural network is a three-layer network with full connection between each layer. The specific network construction framework is as follows: X = (x1, ..., x n ) is the actual input vector of the first input layer, H = (h1, ..., h m ) is used as the output vector of the second hidden layer, Y = (y1, y2) is used as the predicted output result of the third output layer, and A = (a ij ) n×m is the weight matrix from the input layer to the hidden layer, B=(b jk ) m×2 is the weight matrix from the hidden layer to the output layer, C = (c1, .., c m ) is the bias vector of the hidden layer, D = (d1, d2) is the bias vector of the output layer; configure the ReLU function as the activation function of the hidden layer; configure the Sigmoid function as the activation function of the output layer; the maximum number of iterative training is max_epochs; the error threshold for stopping training is ε; the label of the network is Y0 = (r1, r2); where x1…x n represents the extracted feature data; n is the number of neurons in the input layer, such as n = 7; m is the number of neurons in the hidden layer, such as m = 128; y1 and y2 represent the predicted blood growth rate and pain value respectively, a ii is the weight parameter value from the i-th neuron in the input layer to the j-th neuron in the hidden layer; b jk is the weight parameter value from the jth neuron in the hidden layer to the kth neuron in the output layer. r1 and r2 represent the actual blood growth rate and pain value, respectively.

[0070] The network is initialized. The initialization parameters include the weight matrix A from the input layer to the hidden layer; the weight matrix B from the hidden layer to the output layer; the bias vector C of the hidden layer; the bias vector D of the output layer; and the learning rate is set to 1e -4 ;

[0071] Calculate the output of the hidden layer. The calculation method is as follows In the formula, R(·) is the activation function of the hidden layer. The ReLU function is selected as the activation function of the hidden layer. Its function expression is R(t)=max(0, t), where t represents the input of the activation function and R(t) represents the output of the activation function.

[0072] Calculate the predicted output of the output layer. The calculation method is as follows In the formula, P(·) is the activation function of the output layer. Sigmoid function is selected as the activation function of the output layer, and its function definition is: In the formula, z represents the input of the activation function, P(z) represents the output of the activation function, and d kis the bias vector of the output layer. The Sigmoid function limits the output to the range of 0 to 1. p is the output feature of the above hidden layer;

[0073] The mean square error between the predicted result Y and the actual label value Y0 is used as the loss function for training the fully connected neural network. The expression of the loss function is:

[0074] The parameters of the fully connected neural network are updated according to the adaptive moment estimation algorithm, and the learning rate is set to 1e -4 , the maximum number of iterative training times max_epochs is 1000, and back propagation is performed through the adaptive estimation algorithm to update the weight matrix A from the input layer to the hidden layer, the weight matrix B from the hidden layer to the output layer, the bias vector C of the hidden layer, and the bias vector D of the output layer;

[0075] Determine whether the training process is over. If L(Y, Y0)|>ε, return to network initialization and continue training and solving; if |L(Y, Y0)|≤ε, end the training process; if the current training times reaches the maximum training times max_epochs, end the training process;

[0076] Determine the updated network parameters and save the network parameters after training, that is, determine the weight matrix A from the input layer to the hidden layer, the weight matrix B from the hidden layer to the output layer, the bias vector C of the hidden layer, and the bias vector D of the output layer; obtain the trained fully connected neural network according to the above steps, predict the blood velocity and pain value (r1, r2) through the updated network, and select the best ultrasound prescription and usage time according to the prediction results, see Figure 5 shown.

[0077] In order to improve the accuracy of the prediction, constraints on the measurement of blood flow velocity are set. When obtaining blood flow velocity data, in order to avoid the influence of the non-static state of the gout affected area on the measurement of blood flow velocity values ​​during treatment, the therapeutic device will mark the measured data when it monitors the height changes of the affected area. The marked data does not participate in the calculation of the model.

[0078] Specifically, the optimization output module includes: obtaining the blood growth rate and the duration at the blood growth rate according to the prediction results corresponding to the ultrasound prescription, generating a statistical list, the statistical list includes the use time of the ultrasound prescription, the ultrasound prescription ID, the average blood growth rate, the use time and the blood increment that changes with the use time; obtaining the ultrasound prescription treatment effect score according to the blood flow in the statistical list, and being able to track the data throughout the process according to the ultrasound prescription ID after sorting by the score; wherein each ultrasound prescription treatment effect score = blood increment INT that changes with the use time, INT = average blood growth rate * use time. It can be understood that this embodiment also adopts a gout ultrasound therapeutic instrument treatment effect scoring mechanism, and forms the following data list 1 based on the blood growth rate obtained from the treatment results of the ultrasound prescription and the continuous treatment time at this growth rate.

[0079]

[0080] The blood flow increment during treatment = blood level. The single ultrasound prescription treatment score = INT, that is, the blood flow increment during treatment. The comprehensive score of gout ultrasound therapy instrument is the sum of each ultrasound prescription treatment score, as shown in the table, which is 186. At the same time, the system sorts the ultrasound prescription treatment scores in reverse order, which is convenient for the system to make further comparative analysis of the same ultrasound prescription treatment score, and is used for data tracking and analysis of the whole process of treatment.

[0081] This embodiment also adopts a mutual pinching algorithm, that is, the relationship between the use time of the ultrasound prescription and the blood growth rate, and eliminates the non-intersection part of the best ultrasound prescription IDlist obtained by the mutual pinching algorithm and the best ultrasound prescription IDlist obtained by the fully connected neural network prediction method, so as to further optimize the accuracy of the artificial intelligence algorithm.

[0082] This embodiment uses the priority of multiple personal best ultrasound prescription lists: each person's used ultrasound prescription ID will have a treatment score, as shown in the following Table 2:

[0083]

[0084]

[0085] In the above list, the data with the same ultrasound prescription ID and the latest usage time are obtained, and the priority is higher when the average blood flow rate is faster, as shown in Table 3 below:

[0086] Serial number Time of use Ultrasound Prescription ID Average blood growth rate Treatment duration Priority 4 2024010609:00:00 5 0.12 600 1 5 2024010609:30:00 8 0.07 600 2 600 600 600

[0087] The advantage of adjusting the priority of multiple personal optimal ultrasound prescription lists is that it helps patients intelligently select ultrasound prescriptions with better treatment effects, which greatly improves the intelligent rate of selecting appropriate optimal ultrasound prescriptions.

[0088] In this embodiment, the workflow of the wearable ultrasound therapy system is as follows: Figure 6 As shown, the specific steps include:

[0089] Detecting the blood flow rate of the target subject before using the ultrasound prescription by using a wearable ultrasound therapeutic device;

[0090] Download and update the corresponding AI ultrasound prescription from the server, where the AI ​​ultrasound prescription is a target ultrasound prescription adapted to the target object after network prediction;

[0091] Determine whether to select AI ultrasound prescription or default ultrasound prescription. If the default ultrasound prescription is selected, the default M sets of ultrasound prescriptions are used, and the first ultrasound prescription list ID = M sets of default ultrasound prescriptions. Alternatively, if AI ultrasound prescription is selected, determine whether there are N sets of AI ultrasound prescriptions. If so, N sets of AI ultrasound prescriptions are used, and the ultrasound prescription list ID = N sets of AI ultrasound prescriptions. The default ultrasound prescription here is an ultrasound prescription ranking list generated according to the preset weights during factory settings.

[0092] Select the ultrasound prescription ID corresponding to the scheme in the current ultrasound prescription list, execute the ultrasound operation corresponding to the ID, continuously monitor the blood flow rate and pain value of the target object to generate monitoring results, and determine whether to interrupt the current operation and select the ultrasound prescription ID according to the monitoring results;

[0093] When the current operation is interrupted, the monitored ultrasound prescription ID, blood flow rate, pain value and interruption mark are uploaded to the server, and the plan in the list is reselected, and the blood flow rate and pain value are continuously monitored to continue to determine whether it is interrupted;

[0094] If the current operation is not interrupted, it is determined whether the current plan is ended. If so, the currently monitored data is uploaded to the server and the AI ​​ultrasound prescription is updated;

[0095] Traverse the ultrasound prescription list ID until the operation is completed, and calculate the treatment effect score under the current ultrasound prescription based on the maximum blood growth rate and duration.

[0096] The wearable gout ultrasound treatment system based on the artificial intelligence algorithm described in the above embodiment can play a variety of roles in the treatment of gout, including the following aspects: 1. Relieve pain: Ultrasound can stimulate tissues and nerve endings by generating mechanical vibrations, thereby reducing pain. It can promote blood circulation, improve tissue oxygenation and nutrient supply, and also has analgesic effects, which can relieve joint pain caused by gout. 2. Eliminate inflammation: Ultrasound can produce local thermal effects, increase tissue temperature, thereby promoting vasodilation and increased blood flow. This helps to accelerate the removal of inflammatory mediators and the discharge of waste, and reduce inflammatory reactions around joints. 3. Promote the decomposition of uric acid crystals: Ultrasound can promote the decomposition and removal of uric acid crystals. It can change the structure of uric acid crystals, making them easier to be removed by the body, thereby reducing the deposition of uric acid crystals in joints and surrounding tissues. 4. Improve joint mobility: Ultrasound can promote joint mobility and flexibility by stimulating soft tissues and structures around joints. This is very beneficial for reducing joint stiffness and improving joint function.

[0097] Through the unique design of the embodiments of the present invention, in the face of the irregularity of uric acid crystals and the uncertainty of their positions in joints, a variety of changing combinations of ultrasound prescriptions are used as the treatment basis of the gout ultrasound therapeutic device. With artificial intelligence models and algorithms as the core, the shortcomings of traditional ultrasonic therapeutic devices that only have control but no real-time effect monitoring, and are not intelligent enough to give the best therapeutic ultrasound prescription suitable for individuals, as well as the characteristics of being single, poorly adaptable, and having poor therapeutic effects, have been overturned. The uric acid crystals of gout patients have stronger pertinence and adaptability, and the best treatment plan is predicted, greatly improving the treatment effect of gout.

[0098] In this embodiment, the peripheral nerve regulation used is a method of controlling and treating diseases related to the peripheral nervous system using neuromodulation technology. It can affect the excitability of peripheral neurons, thereby changing their perception of pain. Obtaining pain values ​​usually requires the use of some assessment tools, such as pain assessment scales, pain diaries, etc., to assess the patient's pain level and frequency. Peripheral nerve regulation can achieve the effect of alleviating pain by affecting nerve conduction and changing the patient's perception and tolerance of pain. When applying peripheral nerve regulation technology to obtain pain values, attention should be paid to selecting appropriate assessment tools and methods to ensure the accuracy and reliability of the assessment results. At the same time, a personalized treatment plan should be formulated according to the patient's specific situation and needs to achieve the best treatment effect.

[0099] Based on the same concept, the present invention provides a wearable gout ultrasound treatment device based on an artificial intelligence algorithm, comprising: a memory, the memory is used to store a processing program; a processor, the processor is used to execute a computer program to implement the following operations: obtaining historical gout case samples of the target object; extracting characteristic data of the target object before and after the use of an ultrasound prescription from the historical gout case samples, wherein the ultrasound prescription refers to the ultrasonic energy generated by controlling the amplitude and frequency of a wearable ultrasound treatment device; inputting the extracted characteristic data into a fully connected neural network for training to obtain a trained fully connected neural network, which can predict blood growth rate and pain values ​​under different ultrasound prescriptions; and selecting the best target ultrasound prescription result adapted to each target object based on the predicted blood growth rate and pain value.

[0100] In some embodiments of the present application, a readable storage medium is also provided, on which a processing program is stored, and when the processing program is executed by a processor, the following steps are implemented: obtaining historical gout case samples of the target object; extracting characteristic data of the target object before and after the use of an ultrasonic prescription from the historical gout case samples, wherein the ultrasonic prescription refers to ultrasonic energy generated by controlling the amplitude and frequency of a wearable ultrasonic treatment device; inputting the extracted characteristic data into a fully connected neural network for training to obtain a trained fully connected neural network that can predict blood growth rate and pain values ​​under different ultrasonic prescriptions; and selecting the best target ultrasonic prescription result that is suitable for each target object based on the predicted blood growth rate and pain value.

[0101] It is understandable that, for the aforementioned wearable gout ultrasound treatment system based on artificial intelligence algorithm, if it is implemented in the form of software function modules and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or part of the contribution to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0102] Computer readable storage media may include data signals propagated in baseband or as part of a carrier wave, wherein readable program codes are carried. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program codes contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0103] The program code involved in executing the technical solution disclosed in the present application can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and also conventional procedural programming languages ​​such as C language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on the remote computing device, or completely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0104] 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 in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wearable gout ultrasound treatment system based on artificial intelligence algorithm, characterized in that: include: A data collection module is used to obtain historical gout case samples of the target subjects; A feature extraction module, used to extract feature data of the target object before and after the use of the ultrasound prescription from the historical gout case samples, wherein the ultrasound prescription refers to the ultrasonic energy generated by controlling the amplitude and frequency of the wearable ultrasonic treatment device; A prediction training module is used to input the extracted feature data into a fully connected neural network for training, so as to obtain a trained fully connected neural network that can predict blood velocity growth rate and pain value under different ultrasound prescriptions; An optimization output module is used to select the best target ultrasound prescription result suitable for each target object according to the predicted blood growth rate and pain value; The optimization output module comprises: Detecting the blood flow rate of the target subject before using ultrasound prescription by using a wearable ultrasound therapeutic device; Download and update the corresponding AI ultrasound prescription from the server, where the AI ​​ultrasound prescription is a target ultrasound prescription adapted to the target object after network prediction; Determine whether to select AI ultrasound prescription or default ultrasound prescription. If the AI ​​ultrasound prescription is not selected and the default ultrasound prescription is not selected, the default M sets of ultrasound prescriptions are used, and the current ultrasound prescription list ID = M sets of default ultrasound prescriptions. If AI ultrasound prescription or default ultrasound prescription is selected, it is determined whether N sets of AI ultrasound prescription exist. If yes, N sets of AI ultrasound prescription are used, and the current ultrasound prescription list ID = N sets of AI ultrasound prescription; Select the ultrasound prescription ID corresponding to the scheme in the current ultrasound prescription list, execute the ultrasound operation corresponding to the ID, continuously monitor the blood flow rate and pain value of the target object to generate monitoring results, and determine whether to interrupt the current operation and select the ultrasound prescription ID according to the monitoring results; When the current operation is interrupted, the monitored ultrasound prescription ID, blood flow rate, pain value and interruption mark are uploaded to the server, and the plan in the list is reselected, and the blood flow rate and pain value are continuously monitored to continue to determine whether it is interrupted; If the current operation is not interrupted, it is determined whether the current plan is ended. If so, the currently monitored data is uploaded to the server and the ultrasound prescription is updated; Traverse the ultrasound prescription list ID until the operation is completed, and calculate the treatment effect score under the current ultrasound prescription based on the maximum blood growth rate and duration.

2. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The ultrasound prescription includes the ultrasound array energy focusing type, energy focusing depth and energy action surface, wherein the energy focusing type includes preset interval point shape, preset interval point-connected point shape, preset interval continuous shape and random scattered point shape; the energy focusing depth includes the energy focusing depth generated by the superposition of energy generated by one or more ultrasound vibration sources; the energy action surface includes a plane, a convex surface, a concave surface, a curved surface and a surface adapted to the target position.

3. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The feature extraction module also includes: The characteristic data corresponding to the historical gout case samples before and after the use of ultrasonic prescriptions include the target object's region, gender, age, blood flow rate, blood oxygen content, type of ultrasonic prescription, time / number of times the ultrasonic prescription is used, blood growth rate and pain value, among which the types of ultrasonic prescriptions include sine waves, square waves, triangle waves and sound waves related to the range, area and depth of ultrasonic energy.

4. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The feature extraction module also includes: The label setting module is used to obtain the initial blood flow rate, blood oxygen content and pain value before the ultrasonic prescription is applied; obtain the blood flow rate, blood oxygen content and pain value at a preset time after the ultrasonic prescription is applied; calculate the difference between the initial blood flow rate, blood oxygen content and pain value and the blood flow rate, blood oxygen content and pain value after the application, and obtain the blood flow rate growth rate / blood growth rate and pain reduction rate / pain value after the application based on the difference information; The data preprocessing module normalizes the data set and actual result labels obtained in step 1, and limits the actual input data value of the input layer and the label data of the output layer to between 0 and 1.

5. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The prediction training module also includes: Construct a fully connected neural network, X = (x1, ..., x n ) is the actual input vector of the first input layer, H = (h1, ..., h m ) is used as the output vector of the second hidden layer, Y = (y1, y2) is used as the predicted output result of the third output layer, and A = (a ij ) n×m is the weight matrix from the input layer to the hidden layer, B=(b jk ) m×2 is the weight matrix from the hidden layer to the output layer, C=(c1,...,c m ) is the bias vector of the hidden layer, D = (d1, d2) is the bias vector of the output layer; configure the ReLU function as the activation function of the hidden layer; configure the Sigmoid function as the activation function of the output layer; the maximum number of iterative training is max_epochs; the error threshold for stopping training is ε; the label of the network is Y0 = (r1, r2); Among them, x1…x n represents the extracted feature data; n is the number of neurons in the input layer; m is the number of neurons in the hidden layer; y1 and y2 represent the predicted blood growth rate and pain value respectively, a ij is the weight parameter value from the i-th neuron in the input layer to the j-th neuron in the hidden layer; b jk is the weight parameter value from the jth neuron in the hidden layer to the kth neuron in the output layer. r1 and r2 represent the actual blood growth rate and pain value, respectively.

6. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The prediction training module also includes: The network is initialized. The initialization parameters include the weight matrix A from the input layer to the hidden layer; the weight matrix B from the hidden layer to the output layer; the bias vector C of the hidden layer; the bias vector D of the output layer; and the learning rate is set to 1e -4 ; Calculate the output of the hidden layer. The calculation method is as follows In the formula, R(·) is the activation function of the hidden layer. The ReLU function is selected as the activation function of the hidden layer. Its function expression is R(t)=max(0, t), where t represents the input of the activation function and R(t) represents the output of the activation function. Calculate the predicted output of the output layer. The calculation method is as follows In the formula, P(·) is the activation function of the output layer. Sigmoid function is selected as the activation function of the output layer, and its function definition is: In the formula, z represents the input of the activation function, P(z) represents the output of the activation function, and d k is the bias vector of the output layer. The Sigmoid function limits the output to the range of 0 to 1. The mean square error between the predicted result Y and the actual label value Y0 is used as the loss function for training the fully connected neural network. The expression of the loss function is: The parameters of the fully connected neural network are updated according to the adaptive moment estimation algorithm, and the learning rate is set to 1e -4 , the maximum number of iterative training times max_epochs is 1000, and back propagation is performed through the adaptive estimation algorithm to update the weight matrix A from the input layer to the hidden layer, the weight matrix B from the hidden layer to the output layer, the bias vector C of the hidden layer, and the bias vector D of the output layer; Determine whether the training process is over. If |L(Y, Y0)|>ε, return to network initialization and continue training and solving; if |L(Y, Y0)|≤ε, end the training process; if the current training times reaches the maximum training times max_epochs, end the training process; Determine updated network parameters and save the network parameters after training, that is, determine the weight matrix A from the input layer to the hidden layer, the weight matrix B from the hidden layer to the output layer, the bias vector C of the hidden layer, and the bias vector D of the output layer; obtain the trained fully connected neural network according to the above steps, predict the blood growth rate and pain value through the updated network, and select the best ultrasound prescription and usage time according to the prediction results.

7. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The optimization output module comprises: Obtaining the blood growth rate and the duration at the blood growth rate according to the prediction result corresponding to the ultrasound prescription, and generating a statistical list, wherein the statistical list includes the use time of the ultrasound prescription, the ultrasound prescription ID, the average blood growth rate, the use duration, and the blood increment that changes with the use duration; The ultrasound prescription treatment effect score is obtained based on the blood flow in the statistical list, and after sorting by the score, the data can be tracked throughout the process based on the ultrasound prescription ID; The treatment effect score of each ultrasound prescription = blood increment INT that changes with the duration of use, INT = average blood growth rate * duration of use.

8. The wearable gout ultrasound treatment system based on artificial intelligence algorithm according to claim 1 is characterized in that: The wearable ultrasound treatment device comprises: A flexible circuit substrate, on which a plurality of ultrasonic transducer units are distributed to form ultrasonic arrays of different combinations; An ultrasonic transducer unit, wherein the ultrasonic transducer unit is connected to the flexible circuit substrate, and the ultrasonic transducer unit comprises a plurality of ultrasonic transducers and an ultrasonic transducer fixing sleeve adapted to the shape of the ultrasonic transducer, wherein an electrode is arranged on the inner side of the ultrasonic transducer fixing sleeve, and the electrode is connected to the flexible circuit substrate, and at the same time, the electrodes of each ultrasonic transducer are gathered to a bus junction box on the flexible circuit substrate and are connected to an ultrasonic host; the ultrasonic transducer comprises an ultrasonic transducer body, and a through hole is arranged on the upper surface of the ultrasonic transducer body, and when the ultrasonic transducer fixing sleeve is arranged on the ultrasonic transducer and clamped with each other, the ultrasonic transducer body is connected to the electrode, and an anti-fool structure is arranged on the ultrasonic transducer body.

9. A readable storage medium, characterized in that: The readable storage medium stores a processing program, and when the processing program is executed by the processor, the following steps are implemented: obtaining a historical gout case sample of the target object; extracting characteristic data of the target object before and after the use of an ultrasonic prescription from the historical gout case sample, wherein the ultrasonic prescription refers to ultrasonic energy generated by controlling the amplitude and frequency of a wearable ultrasonic treatment device; inputting the extracted characteristic data into a fully connected neural network for training, thereby obtaining a trained fully connected neural network capable of predicting blood growth rate and pain value under different ultrasonic prescriptions; The best target ultrasound prescription result suitable for each target object is selected based on the predicted blood velocity and pain value, including: Detecting the blood flow rate of the target subject before using ultrasound prescription by using a wearable ultrasound therapeutic device; Download and update the corresponding AI ultrasound prescription from the server, where the AI ​​ultrasound prescription is a target ultrasound prescription adapted to the target object after network prediction; Determine whether to select AI ultrasound prescription or default ultrasound prescription. If the AI ​​ultrasound prescription is not selected and the default ultrasound prescription is not selected, the default M sets of ultrasound prescriptions are used, and the current ultrasound prescription list ID = M sets of default ultrasound prescriptions. If AI ultrasound prescription or default ultrasound prescription is selected, it is determined whether N sets of AI ultrasound prescription exist. If yes, N sets of AI ultrasound prescription are used, and the current ultrasound prescription list ID = N sets of AI ultrasound prescription; Select the ultrasound prescription ID corresponding to the scheme in the current ultrasound prescription list, execute the ultrasound operation corresponding to the ID, continuously monitor the blood flow rate and pain value of the target object to generate monitoring results, and determine whether to interrupt the current operation and select the ultrasound prescription ID according to the monitoring results; When the current operation is interrupted, the monitored ultrasound prescription ID, blood flow rate, pain value and interruption mark are uploaded to the server, and the plan in the list is reselected, and the blood flow rate and pain value are continuously monitored to continue to determine whether it is interrupted; If the current operation is not interrupted, it is determined whether the current plan is ended. If so, the currently monitored data is uploaded to the server and the ultrasound prescription is updated; Traverse the ultrasound prescription list ID until the operation is completed, and calculate the treatment effect score under the current ultrasound prescription based on the maximum blood growth rate and duration.

Images