AI-based intelligent analysis system and method for urinary tract functional status
Through the AI-based urinary tract functional status intelligent analysis system, sensors are used to collect urinary flow and urethral pressure data and conduct timing analysis, the problem of insufficient invasiveness and real-time performance of traditional methods is solved, and the efficiency and intelligence of urinary tract functional status is achieved. Monitor and diagnose.
Patent Information
- Application Number
- CN202411173102.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-26
AI Technical Summary
Traditional urinary tract function status analysis methods are highly invasive and cannot reflect the dynamic changes in urinary tract function in real time, resulting in insufficient diagnosis and low patient comfort.
Using an AI-based urinary tract functional state intelligent analysis system, time series data of urine flow and urethral pressure is collected through sensors, and local timing-related features are extracted using sequence encoder and energy timing-decaying dynamic aggregation network to perform timing aggregation and features. Interactive analysis to determine the urinary tract functional status.
It realizes non-invasive and real-time monitoring of the timing dynamic subtle changes in the urinary tract functional status, improves the intelligence level of urinary tract functional status analysis, and reduces the risk of physical discomfort and infection in traditional methods.
Smart Images

Figure CN118692706B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent health data analysis, and specifically to an AI-based intelligent analysis system and method for urinary tract functional status. Background Art
[0002] Urinary tract dysfunction is a far-reaching health problem that can lead to symptoms such as urinary incontinence, frequent urination, and urgency, seriously affecting patients' daily lives and participation in social activities. These symptoms not only bring physical inconvenience, but may also cause emotional problems such as embarrassment, anxiety, and depression, thereby reducing the quality of life.
[0003] Traditional analysis of urinary tract functional status usually relies on urodynamic examinations. Specifically, although traditional urodynamic examinations provide valuable diagnostic information for clinicians, their invasive nature means that patients may feel uncomfortable or painful when undergoing the examination. For example, filling cystometry and pressure flow studies usually need to be performed through medical devices such as catheters, which not only increases the risk of infection, but may also cause physical discomfort to patients due to the use of the equipment. In addition, traditional urodynamic examination methods are mostly intermittent manual examinations or assessments based on static data, which may not be able to reflect the patient's actual situation in real time during different time periods or activity states, such as changes in dynamic indicators such as urination frequency and urine flow, and thus lack the capture of these dynamic subtle changes, resulting in untimely diagnosis of urinary tract dysfunction.
[0004] Therefore, an optimized urinary tract functional status analysis system is desired. Summary of the invention
[0005] This application is made in consideration of the above problems. One purpose of this application is to provide an AI-based intelligent analysis system and method for urinary tract function status.
[0006] The present application provides an AI-based intelligent analysis system for urinary tract function status, which includes:
[0007] A urine flow and urethra pressure time series data acquisition module, used to acquire the urine flow time series and urethra pressure time series acquired by the sensor;
[0008] A urine flow and urethra pressure local time series association module, used for inputting the urine flow time series and the urethra pressure time series into a sequence encoder respectively to obtain a sequence of urine flow local time series association implicit feature vectors and a sequence of urethra pressure local time series association implicit feature vectors;
[0009] A urine flow and urethra pressure time series aggregation module, used for inputting the sequence of urine flow local time series associated implicit feature vectors and the sequence of urethra pressure local time series associated implicit feature vectors into a node feature dynamic aggregation network based on energy time series decay to obtain a urine flow time series aggregation representation vector and a urethra pressure time series aggregation representation vector;
[0010] A urine flow urethra pressure time series aggregation feature interaction module, used for inputting the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector into a feature interaction adaptive response module to obtain a urine flow-urethra pressure time series interaction response representation vector as a urine flow-urethra pressure time series interaction response representation feature;
[0011] A urinary tract function state abnormality judgment module is used to obtain an analysis result based on the urine flow-urethral pressure time series interactive response representation characteristics, and the analysis result is used to indicate whether the urinary tract function state is normal;
[0012] Wherein, the urine flow and urethral pressure time series aggregation feature interaction module includes:
[0013] A position-by-position response calculation unit, used for calculating the position-by-position response between the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector to obtain a urine flow-urethra pressure time series aggregate position-by-position response feature vector;
[0014] A normalization processing unit, used for normalizing the urine flow-urethra pressure time series aggregated position-by-position response feature vector using a Softmax function to obtain a normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector;
[0015] A urine flow-urethral pressure time series aggregate mask calculation unit, used for inputting the normalized urine flow-urethral pressure time series aggregate position-by-position response feature vector into a learnable gating function to obtain a urine flow-urethral pressure time series aggregate response screening weight mask vector;
[0016] An aggregated position-by-position response unit, used for calculating the position-by-position point multiplication between the urine flow-urethra pressure time series aggregated response screening weight mask vector and the normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector to obtain a urine flow-urethra pressure time series aggregated position-by-position response distinguishable weight mask vector;
[0017] The interactive response unit is used to calculate the distinguishable weight mask vector of the urine flow-urethral pressure time series aggregation position-by-position response and the urine flow-urethral pressure time series aggregation position-by-position response feature vector to obtain the urine flow-urethral pressure time series interactive response representation vector.
[0018] For example, according to the AI-based intelligent analysis system for urinary tract function status of the present application, the urine flow and urethral pressure local time series association module is used to:
[0019] The time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder based on an RNN model to obtain a sequence of implicit feature vectors associated with local temporal series of urine flow and a sequence of implicit feature vectors associated with local temporal series of urethral pressure.
[0020] For example, according to the AI-based intelligent analysis system for urinary tract function status of the present application, the urine flow and urethral pressure time series aggregation module includes:
[0021] A urine flow local time series associated implicit feature energy statistical paradigm value calculation unit, used to calculate the energy statistical paradigm value of each urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vectors based on the maximum value, average value and variance of each urine flow local time series associated implicit feature vector to obtain a sequence of urine flow local time series associated energy statistical paradigm values, wherein the energy statistical paradigm value corresponding to the current urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated energy statistical paradigm values is used as the current node energy statistical paradigm value and other energy statistical paradigm values are used as historical node energy statistical paradigm values to obtain a sequence of current urine flow local time series associated energy statistical paradigm values and historical urine flow local time series associated energy statistical paradigm values;
[0022] A node propagation space span value statistical unit is used to count the node propagation space span values between each other urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vector and the current urine flow local time series associated implicit feature vector to obtain a sequence of urine flow local time series associated propagation space span values;
[0023] A urine flow local time series energy propagation attenuation coefficient value calculation unit is used to determine the energy propagation attenuation coefficient values of other urine flow local time series associated implicit feature vectors in the sequence of urine flow local time series associated implicit feature vectors based on the sequence of urine flow local time series associated propagation space span values and the sequence of historical urine flow local time series associated energy statistical paradigm values to obtain a sequence of urine flow local time series associated energy propagation attenuation coefficient values, wherein the urine flow local time series associated energy propagation attenuation coefficient values are inversely correlated with the historical urine flow local time series associated energy statistical paradigm values;
[0024] A weighted sum unit, used to calculate the weighted sum of all other urine flow local time series associated implicit feature vectors in the sequence of urine flow local time series associated implicit feature vectors by using the sequence of urine flow local time series associated energy propagation attenuation coefficient values as a weight sequence to obtain a historical urine flow local time series associated energy attenuation time series aggregated feature vector;
[0025] The aggregation unit is used to fuse the current urine flow local time series associated energy statistical paradigm value to calculate the weighted sum of the historical urine flow local time series associated energy attenuation time series aggregation feature vector and the current urine flow local time series associated implicit feature vector to obtain the urine flow time series aggregation representation vector.
[0026] For example, according to the AI-based intelligent analysis system for urinary tract function status of the present application, the urine flow local time series associated implicit feature energy statistical paradigm value calculation unit is used to:
[0027] Extracting the maximum value of the urine flow local time series associated implicit feature vector to obtain the urine flow local time series associated implicit feature maximum value;
[0028] Respectively calculating the mean value and variance of the urine flow local time series associated implicit feature vector to obtain the urine flow local time series associated implicit feature mean value and the urine flow local time series associated implicit feature variance;
[0029] The value obtained by adding the variance of the urine flow local time series associated implicit feature and the preset hyperparameter is multiplied by four to obtain the first energy statistical paradigm value of the urine flow local time series associated implicit feature;
[0030] Calculating the square of the difference between the maximum value of the urine flow local time series associated implicit feature and the average value of the urine flow local time series associated implicit feature to obtain a urine flow local time series associated implicit feature difference value;
[0031] The second energy statistical paradigm value of the urine flow local time series association implicit feature is obtained by adding the modulated urine flow local time series association implicit feature variance obtained by multiplying the urine flow local time series association implicit feature variance by two, the value obtained by multiplying the preset hyperparameter by two, and the urine flow local time series association implicit feature difference value;
[0032] The first energy statistical paradigm value of the local temporal correlation implicit feature of the urine flow is divided by the second energy statistical paradigm value of the local temporal correlation implicit feature of the urine flow to obtain the energy statistical paradigm value of the local temporal correlation implicit feature of the urine flow.
[0033] For example, according to the AI-based intelligent analysis system for urinary tract function status of the present application, the urine flow local time series energy propagation attenuation coefficient value calculation unit is used to:
[0034] Multiplying each urine flow local time series associated propagation space span value in the sequence of urine flow local time series associated propagation space span values by a first weight hyperparameter to obtain a sequence of urine flow local time series associated first weighted propagation space span values;
[0035] Using each urine flow local time series associated propagation space span value in the sequence of urine flow local time series associated propagation space span values as an exponent of a natural constant to calculate an exponential function value with base e according to position to obtain a sequence of urine flow local time series associated propagation class support space span values;
[0036] Multiplying each urine flow local time series association propagation class support space span value in the sequence of urine flow local time series association propagation class support space span values by a second weight hyperparameter to obtain a sequence of urine flow local time series association second weighted propagation space span values;
[0037] Calculating the positional sum of the sequence of the first weighted propagation space span value associated with the local time series of urine flow and the sequence of the second weighted propagation space span value associated with the local time series of urine flow to obtain a sequence of weighted sum propagation space span value associated with the local time series of urine flow;
[0038] Divide each group of corresponding historical urine flow local time series correlation energy statistical paradigm values and urine flow local time series correlation weighted sum propagation space span values in the sequence of the historical urine flow local time series correlation energy statistical paradigm values and the sequence of the urine flow local time series correlation weighted sum propagation space span values to obtain the sequence of the urine flow local time series correlation energy propagation attenuation coefficient values.
[0039] For example, according to the AI-based intelligent analysis system for urinary tract function status of the present application, the urine flow-urethral pressure time series aggregation mask calculation unit is used to:
[0040] The normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector is input into a sigmoid function to calculate a weight mask value to obtain a urine flow-urethra pressure time series aggregated response screening weight mask vector.
[0041] For example, according to the AI-based intelligent analysis system for urinary tract function status of the present application, the urinary tract function status abnormality judgment module is used to:
[0042] The urine flow-urethra pressure time series interaction response representation vector is input into a classifier-based urinary tract function state intelligent analysis module to obtain the analysis result, and the analysis result is used to indicate whether the urinary tract function state is normal.
[0043] The present application also provides an AI-based intelligent analysis method for urinary tract functional status, which includes:
[0044] Acquire a time series of urine flow and a time series of urethral pressure collected by a sensor;
[0045] Inputting the time series of urine flow and the time series of urethral pressure into a sequence encoder respectively to obtain a sequence of urine flow local time series associated implicit feature vectors and a sequence of urethral pressure local time series associated implicit feature vectors;
[0046] Inputting the sequence of the urine flow local time series associated implicit feature vectors and the sequence of the urethra pressure local time series associated implicit feature vectors into the node feature dynamic aggregation network based on energy time series decay respectively to obtain the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector;
[0047] Inputting the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector into a feature interactive adaptive response module to obtain a urine flow-urethra pressure time series interactive response representation vector as a urine flow-urethra pressure time series interactive response representation feature;
[0048] Based on the urine flow-urethra pressure time series interactive response representation characteristics, an analysis result is obtained, and the analysis result is used to indicate whether the urinary tract function state is normal;
[0049] The step of inputting the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector into a feature interactive adaptive response module to obtain a urine flow-urethra pressure time series interactive response representation vector as a urine flow-urethra pressure time series interactive response representation feature includes:
[0050] Calculating the position-by-position response between the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector to obtain a urine flow-urethra pressure time series aggregate position-by-position response feature vector;
[0051] Using a Softmax function to normalize the urine flow-urethra pressure time series aggregated position-by-position response feature vector to obtain a normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector;
[0052] Inputting the normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector into a learnable gating function to obtain a urine flow-urethra pressure time series aggregated response screening weight mask vector;
[0053] Calculate the position point multiplication between the urine flow-urethra pressure time series aggregate response screening weight mask vector and the normalized urine flow-urethra pressure time series aggregate position-by-position response feature vector to obtain a urine flow-urethra pressure time series aggregate position-by-position response distinguishable weight mask vector;
[0054] The urine flow-urethral pressure time series aggregated position-by-position response is calculated by multiplying the distinguishable weight mask vector by the urine flow-urethral pressure time series aggregated position-by-position response feature vector by the position point to obtain the urine flow-urethral pressure time series interactive response representation vector.
[0055] For example, according to the AI-based intelligent analysis method of urinary tract function status of the present application, the time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder to obtain a sequence of implicit feature vectors associated with local time series of urine flow and a sequence of implicit feature vectors associated with local time series of urethral pressure, including:
[0056] The time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder based on an RNN model to obtain a sequence of implicit feature vectors associated with local temporal series of urine flow and a sequence of implicit feature vectors associated with local temporal series of urethral pressure.
[0057] According to the AI-based intelligent analysis system and method of urinary tract functional status of the present application, it is possible to use AI to monitor and analyze the time series of urine flow and urethral pressure in real time, and timely capture the subtle changes in the temporal dynamics of the urinary tract functional status, thereby significantly improving the intelligence level of urinary tract functional status analysis. BRIEF DESCRIPTION OF THE DRAWINGS
[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.
[0059] Figure 1 A schematic diagram of the application architecture of the AI-based intelligent analysis system for urinary tract function status in an embodiment of the present application is shown;
[0060] Figure 2 The structure diagram of the AI-based intelligent analysis system for urinary tract function status in an embodiment of the present application is shown;
[0061] Figure 3 A schematic diagram showing the structure of a urine flow and urethral pressure time series aggregation module of an AI-based intelligent analysis system for urinary tract function status in an embodiment of the present application is shown;
[0062] Figure 4 A schematic diagram of the structure of the urine flow and urethra pressure time series aggregation feature interaction module of the AI-based urinary tract function status intelligent analysis system in an embodiment of the present application is shown;
[0063] Figure 5 A flow chart of an AI-based intelligent analysis method for urinary tract function status in an embodiment of the present application is shown;
[0064] Figure 6 The application scenario diagram of the AI-based intelligent analysis system for urinary tract functional status in an embodiment of the present application is shown. DETAILED DESCRIPTION
[0065] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.
[0066] The terms used in this specification are those common terms currently widely used in the art in consideration of the functions of the present application, but these terms may vary according to the intentions of those skilled in the art, precedents, or new technologies in the art. In addition, specific terms may be selected, and in this case, their detailed meanings will be described in the detailed description of the present application. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present application.
[0067] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.
[0068] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.
[0069] Figure 1 A schematic diagram of the application architecture of an AI-based intelligent analysis system for urinary tract function status in an embodiment of the present application is shown, including a server 100 and a terminal device 200.
[0070] The terminal device 200 and the server 100 can be connected via the Internet to achieve mutual communication. Optionally, the above-mentioned Internet uses standard communication technology and / or protocol. The Internet is usually the Internet, but it can also be any network, including but not limited to any combination of a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a private network or a virtual private network. In some embodiments, the data exchanged through the network can be represented by technologies and / or formats including Hyper Text Markup Language (HTML), Extensible Markup Language (XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. can be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above-mentioned data communication technologies.
[0071] The server 100 can provide various network services for the terminal device 200, wherein the server 100 can be a single server, a server cluster consisting of several servers, or a cloud computing center. Specifically, the server 100 may include a processor 110 (Center Processing Unit, CPU), a memory 120, an input device 130, and an output device 140, etc. The input device 130 may include a keyboard, a mouse, a touch screen, etc., and the output device 140 may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), a cathode ray tube (Cathode Ray Tube, CRT), etc.
[0072] The memory 120 may include a read-only memory (ROM) and a random access memory (RAM), and provides the processor 110 with program instructions and data stored in the memory 120. In the embodiment of the present application, the memory 120 may be used to store the corresponding method of the AI-based urinary tract function status intelligent analysis system in the embodiment of the present application.
[0073] The processor 110 calls the program instructions stored in the memory 120, and the processor 110 is used to execute the steps of any AI-based intelligent analysis method of urinary tract function status in the embodiments of the present application according to the obtained program instructions.
[0074] In addition, the application architecture diagram in the embodiment of the present application is intended to more clearly illustrate the technical solution in the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present application is also applicable to similar problems.
[0075] The following is a non-restrictive description of the AI-based intelligent analysis system for urinary tract functional status provided according to at least one embodiment of the present application through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present application.
[0076] In response to the above technical problems, the present application proposes an AI-based intelligent analysis system for urinary tract function status, which uses sensors to collect the time series of urine flow and the time series of urethral pressure, and uses AI-based data analysis and processing algorithms to extract local time series correlation features and aggregate the urine flow and urethral pressure in the entire time domain, so as to intelligently judge whether the urinary tract function status is normal based on the responsive interaction characteristics between the aggregated representation features of the urine flow and the urethral pressure in the entire time domain. In this way, non-invasive or minimally invasive sensors can be used to collect urine flow and urethral pressure in real time in different time periods and activity states, reducing the physical discomfort and infection risks caused by the use of catheters and other equipment in traditional urodynamic examinations, and improving the patient's comfort. In addition, AI is used to monitor and analyze the time series of urine flow and urethral pressure in real time, which can timely capture the subtle changes in the time series dynamics of the urinary tract function status, thereby significantly improving the intelligence level of urinary tract function status analysis.
[0077] Figure 2The schematic diagram of the structure of the AI-based intelligent analysis system 800 of urinary tract function status in the embodiment of the present application is shown. The AI-based intelligent analysis system 800 of urinary tract function status includes: a urine flow and urethral pressure time series data acquisition module 810, which is used to obtain the time series of urine flow and the time series of urethral pressure collected by the sensor; a urine flow urethral pressure local time series association module 820, which is used to input the time series of urine flow and the time series of urethral pressure into a sequence encoder respectively to obtain a sequence of urine flow local time series association implicit feature vectors and a sequence of urethral pressure local time series association implicit feature vectors; a urine flow urethral pressure time series aggregation module 830, which is used to aggregate the sequence of urine flow local time series association implicit feature vectors and the sequence of urethral pressure local time series association implicit feature vectors. The sequences are respectively input into the node feature dynamic aggregation network based on energy time series attenuation to obtain the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector; the urine flow urethral pressure time series aggregation feature interaction module 840 is used to input the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector into the feature interaction adaptive response module to obtain the urine flow-urethral pressure time series interaction response representation vector as the urine flow-urethral pressure time series interaction response representation feature; the urinary tract function state abnormality judgment module 850 is used to obtain the analysis result based on the urine flow-urethral pressure time series interaction response representation feature, and the analysis result is used to indicate whether the urinary tract function state is normal.
[0078] Specifically, in the technical solution of the present application, first, the time series of urine flow and the time series of urethral pressure collected by the sensor are obtained. Next, considering that the time series of urine flow and the time series of urethral pressure both have characteristic information in the time dimension, and the implicit associations expressed between different time spans are different. Based on this, in the technical solution of the present application, the time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder based on an RNN model that can capture the local temporal dynamic characteristics and time dependence in the time series to respectively capture and perceive the implicit association relationship between each data in the local temporal series, and obtain a sequence of implicit feature vectors of the local temporal association of urine flow and a sequence of implicit feature vectors of the local temporal association of urethral pressure. In particular, the sequence encoder based on the RNN model described here is a convolutional network model based on the RNN model.
[0079] Correspondingly, the urine flow and urethral pressure local time series association module 820 is used to: input the time series of the urine flow and the time series of the urethral pressure into the sequence encoder based on the RNN model respectively to obtain the sequence of the urine flow local time series association implicit feature vector and the sequence of the urethral pressure local time series association implicit feature vector.
[0080] Then, considering that each local time series association implicit feature vector in the sequence of the urine flow local time series association implicit feature vector and the sequence of the urethral pressure local time series association implicit feature vector respectively represents the time series association feature information of the urine flow and the urethral pressure in the local time, and the importance of different time points in the time series data may change with time, therefore, in order to more reasonably and selectively strengthen and integrate each local time series association feature, in the technical solution of the present application, the sequence of the urine flow local time series association implicit feature vector and the sequence of the urethral pressure local time series association implicit feature vector are processed by using a node feature dynamic aggregation network based on energy time series attenuation to obtain the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector. In particular, the node feature dynamic aggregation network based on energy time series decay utilizes the mechanism of node energy decay to integrate and represent the node features in the time series in the time series dimension, strengthens the model's perception and analysis of the dynamic characteristics of the time series, and realizes the time series spatial aggregation of node features, so that the network can more deeply explore and understand the complex patterns of data on the timeline, which provides richer and more detailed feature support for the intelligent analysis of urinary tract functional status. Specifically, the maximum value, mean value and variance of each node feature vector are first calculated, and these statistics are used to obtain the energy statistical paradigm value of each node feature to quantify the energy level of each node. After that, the propagation space span value between each node is calculated to capture the spatial relationship and topological structure between each node, and these propagation space span values are combined with the historical node energy statistical paradigm value to obtain the node energy propagation attenuation coefficient, so as to reflect the natural attenuation phenomenon of node energy over time and spatial distribution. Finally, these coefficients are used as weights to perform weighted summation on the node feature vector sequence, and combined with the energy statistics of the current node, the weighted summed historical aggregated feature vector is weighted merged with the current feature vector to generate a temporal aggregate representation vector that combines temporal dynamics and spatial layout.
[0081] Accordingly, if Figure 3As shown, the urine flow urethral pressure time series aggregation module 830 includes: a urine flow local time series associated implicit feature energy statistical paradigm value calculation unit 831, which is used to calculate the energy statistical paradigm value of each urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vectors based on the maximum value, average value and variance of each urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vectors to obtain a sequence of urine flow local time series associated energy statistical paradigm values, wherein the urine flow local time series associated implicit feature vector corresponding to the current urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated energy statistical paradigm values is converted into the energy statistical paradigm value of the urine flow local time series associated implicit feature vector. The energy statistical paradigm value is used as the current node energy statistical paradigm value and other energy statistical paradigm values are used as historical node energy statistical paradigm values to obtain a sequence of current urine flow local time series associated energy statistical paradigm values and historical urine flow local time series associated energy statistical paradigm values; the node propagation space span value statistical unit 832 is used to count the node propagation space span values between each other urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vectors and the current urine flow local time series associated implicit feature vector to obtain a sequence of urine flow local time series associated propagation space span values; urine flow local time series associated implicit feature vector The time series energy propagation attenuation coefficient value calculation unit 833 is used to determine the energy propagation attenuation coefficient values of other urine flow local time series associated implicit feature vectors in the sequence of urine flow local time series associated implicit feature vectors based on the sequence of urine flow local time series associated propagation space span values and the sequence of historical urine flow local time series associated energy statistical paradigm values to obtain a sequence of urine flow local time series associated energy propagation attenuation coefficient values, wherein the urine flow local time series associated energy propagation attenuation coefficient values are inversely correlated with the historical urine flow local time series associated energy statistical paradigm values; the weighted sum unit 834 is used Taking the sequence of the urine flow local time series associated energy propagation attenuation coefficient values as the weight sequence, the weighted sum of all other urine flow local time series associated implicit feature vectors in the sequence of the urine flow local time series associated implicit feature vector is calculated to obtain the historical urine flow local time series associated energy attenuation time series aggregated feature vector; the aggregation unit 835 is used to fuse the current urine flow local time series associated energy statistical paradigm value to calculate the weighted sum of the historical urine flow local time series associated energy attenuation time series aggregated feature vector and the current urine flow local time series associated implicit feature vector to obtain the urine flow time series aggregated representation vector.
[0082] The energy statistical paradigm value calculation unit 831 of the local time series association implicit feature of urine flow is used to: extract the maximum value of the local time series association implicit feature vector of urine flow to obtain the maximum value of the local time series association implicit feature of urine flow; respectively calculate the mean value and variance of the local time series association implicit feature vector of urine flow to obtain the mean value of the local time series association implicit feature of urine flow and the variance of the local time series association implicit feature of urine flow; multiply the value obtained by adding the variance of the local time series association implicit feature of urine flow and the preset hyperparameter by four to obtain the first energy statistical paradigm value of the local time series association implicit feature of urine flow; calculate the maximum value of the local time series association implicit feature of urine flow The square of the difference between the value and the average value of the local time series association implicit feature of urine flow is obtained to obtain the difference value of the local time series association implicit feature of urine flow; the modulated urine flow local time series association implicit feature variance obtained by multiplying the variance of the local time series association implicit feature by two is added to the value obtained by multiplying the preset hyperparameter by two and the difference value of the local time series association implicit feature of urine flow to obtain the second energy statistical paradigm value of the local time series association implicit feature of urine flow; the first energy statistical paradigm value of the local time series association implicit feature of urine flow is divided by the second energy statistical paradigm value of the local time series association implicit feature of urine flow to obtain the energy statistical paradigm value of the local time series association implicit feature of urine flow.
[0083] The urine flow local time series energy propagation attenuation coefficient value calculation unit 833 is used to: multiply each urine flow local time series associated propagation space span value in the sequence of urine flow local time series associated propagation space span values by a first weight hyperparameter to obtain a sequence of urine flow local time series associated first weighted propagation space span values; use each urine flow local time series associated propagation space span value in the sequence of urine flow local time series associated propagation space span values as an exponent of a natural constant to calculate an exponential function value with base e according to position to obtain a sequence of urine flow local time series associated propagation class support space span values; and multiply each urine flow local time series associated propagation class support space span value in the sequence of urine flow local time series associated propagation class support space span values by a first weight hyperparameter to obtain a sequence of urine flow local time series associated propagation space span values. The value is multiplied by the second weight hyperparameter to obtain a sequence of second weighted propagation space span values associated with local time series of urine flow; the sequence of first weighted propagation space span values associated with local time series of urine flow and the sequence of second weighted propagation space span values associated with local time series of urine flow are calculated by positional summation to obtain a sequence of weighted sum propagation space span values associated with local time series of urine flow; the sequence of historical urine flow local time series associated energy statistical paradigm values and the sequence of weighted sum propagation space span values associated with local time series of urine flow are divided by each group of corresponding historical urine flow local time series associated energy statistical paradigm values and weighted sum propagation space span values associated with local time series of urine flow to obtain a sequence of energy propagation attenuation coefficient values associated with local time series of urine flow.
[0084] In a specific example, the urine flow urethral pressure time series aggregation module 830 is used to: input the sequence of the urine flow local time series associated implicit feature vector into the node feature dynamic aggregation network based on energy time series decay, and process it with the following dynamic aggregation formula to obtain the urine flow time series aggregation representation vector; wherein the dynamic aggregation formula is:
[0085] ;
[0086] ;
[0087] ;
[0088] ;
[0089] ;
[0090] in, is a sequence of implicit feature vectors associated with the urine flow local time series, represents a sequence of vectors, and are the first and The urine flow local time series associated implicit feature vector, represents the number of feature vectors in the sequence of implicit feature vectors associated with the local temporal sequence of urine flow, and are the first and second eigenvectors of the local temporal association implicit feature vector of urine flow. The mean and variance of the latent feature vectors of the local temporal association of urine flow, For the The maximum value of the latent eigenvector of the local temporal association of urine flow, is the preset hyperparameter, It is The energy statistical norm value of the implicit feature vector of the local temporal association of urine flow, It is The urine flow local time series associated implicit feature vector, It is The first of the hidden feature vectors of the local temporal association of urine flow The eigenvalues at the positions, is the number of eigenvalues in each urine flow local time series associated implicit eigenvector, and They are and The energy statistical norm value of the implicit eigenvector of the local temporal association of urine flow, express and The number of eigenvectors between , , and is the weight hyperparameter, is the urine flow time series aggregate representation vector. Similarly, the encoding method of the sequence of the urethra pressure local time series associated implicit feature vectors is consistent with the encoding method of the sequence of the urine flow local time series associated implicit feature vectors.
[0091] Furthermore, in order to make the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector interact and respond more carefully in the time dimension, so as to deeply analyze and understand the complex time series interaction between urine flow and urethral pressure and its impact on the urinary tract functional state, in the technical solution of the present application, the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector are input into the feature interaction adaptive response module to obtain the urine flow-urethral pressure time series interaction response representation vector. Specifically, the feature interaction adaptive response module captures the dynamic response relationship of urine flow and urethral pressure over time more carefully through the element-by-element interaction analysis and dynamic weight optimization of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector, significantly enhances the ability to capture key time series interaction information, and more accurately reveals the inherent complex interaction pattern of urine flow and urethral pressure in time series, thereby comprehensively improving the recognition accuracy and prediction efficiency of the model, and bringing breakthrough improvements to the intelligent monitoring and evaluation of urinary tract functional state.
[0092] Accordingly, if Figure 4As shown, the urine flow urethra pressure time series aggregation feature interaction module 840 includes: a position-by-position response calculation unit 841, which is used to calculate the position-by-position response between the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector to obtain a urine flow-urethra pressure time series aggregation position-by-position response feature vector; a normalization processing unit 842, which is used to use a Softmax function to normalize the urine flow-urethra pressure time series aggregation position-by-position response feature vector to obtain a normalized urine flow-urethra pressure time series aggregation position-by-position response feature vector; a urine flow-urethra pressure time series aggregation mask calculation unit 843, which is used to convert the normalized urine flow-urethra pressure time series aggregation position-by-position response feature vector into a normalized urine flow-urethra pressure time series aggregation position-by-position response feature vector. A learnable gating function is input into the amount to obtain a urine flow-urethral pressure timing aggregate response screening weight mask vector; an aggregate position-by-position response unit 844 is used to calculate the position point multiplication between the urine flow-urethral pressure timing aggregate response screening weight mask vector and the normalized urine flow-urethral pressure timing aggregate position-by-position response feature vector to obtain a urine flow-urethral pressure timing aggregate position-by-position response distinguishable weight mask vector; an interactive response unit 845 is used to calculate the position point multiplication between the urine flow-urethral pressure timing aggregate position-by-position response distinguishable weight mask vector and the urine flow-urethral pressure timing aggregate position-by-position response feature vector to obtain the urine flow-urethral pressure timing interactive response representation vector.
[0093] Among them, the urine flow-urethral pressure time series aggregation mask calculation unit 843 is used to: input the normalized urine flow-urethral pressure time series aggregation position-by-position response feature vector into the sigmoid function to calculate the weight mask value to obtain the urine flow-urethral pressure time series aggregation response screening weight mask vector.
[0094] In a specific example, the urine flow urethra pressure time series aggregation feature interaction module 840 is used to: input the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector into the feature interaction adaptive response module, and process them with the following adaptive response formula to obtain the urine flow-urethra pressure time series interaction response representation vector; wherein the adaptive response formula is:
[0095] ;
[0096] ;
[0097] ;
[0098] ;
[0099] ;
[0100] in, and are respectively the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector, represents the urine flow-urethral pressure time series aggregated position-by-position response feature vector, yes function, is the normalized urine flow-urethral pressure time series aggregated position-by-position response feature vector, yes function, is the urine flow-urethral pressure time series aggregate response screening weight mask vector, It is the point multiplication by position. is the distinguishable weight mask vector of the urine flow-urethral pressure time series aggregated position-by-position response, is the urine flow-urethra pressure time series interaction response representation vector.
[0101] Then, the urine flow-urethral pressure time series interactive response representation vector is input into the intelligent analysis module of urinary tract function state based on the classifier to obtain the analysis result, and the analysis result is used to indicate whether the urinary tract function state is normal. That is, the urine flow-urethral pressure time series interactive response representation feature obtained by the dynamic interactive response of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector is classified and processed to intelligently determine whether the urinary tract function state is normal. In this way, non-invasive or minimally invasive sensors can be used to collect urine flow and urethral pressure in real time in different time periods and activity states, reducing the physical discomfort and infection risks caused by the use of catheters and other equipment in traditional urodynamic examinations, and improving the comfort of patients. In addition, AI is used to monitor and analyze the time series of urine flow and urethral pressure in real time, which can timely capture the subtle changes in the time series dynamics of the urinary tract function state, thereby significantly improving the intelligence level of urinary tract function state analysis.
[0102] Correspondingly, the urinary tract function status abnormality judgment module 850 is used to: input the urine flow-urethral pressure time series interaction response representation vector into the classifier-based urinary tract function status intelligent analysis module to obtain the analysis result, and the analysis result is used to indicate whether the urinary tract function status is normal.
[0103] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules, and then classify (or predict) unknown data. Logistic regression, SVM, etc. are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification, but this is prone to errors and is not efficient. Commonly used multi-classification methods include the Softmax classification function.
[0104] Preferably, the urinary tract function status abnormality judgment module 850 includes: a correction unit, used to correct the urine flow-urethral pressure timing interaction response representation vector to obtain a corrected urine flow-urethral pressure timing interaction response representation vector; a classification unit, used to input the corrected urine flow-urethral pressure timing interaction response representation vector into a urinary tract function status intelligent analysis module based on a classifier to obtain the analysis result.
[0105] Wherein, the correction unit is used to: input the urine flow-urethral pressure time series interaction response representation vector into the classifier-based urinary tract function state intelligent analysis module to obtain the analysis result, including: calculating the first norm and second norm of the mean vector of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector; calculating the weighted sum of the inverse of the square root of the second norm and the first norm, and performing point multiplication with the point addition sum vector of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector to obtain the first urine flow-urethral pressure time series interaction response correction subvector; inputting the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector into the classifier-based urinary tract function state intelligent analysis module to obtain the analysis result, including: calculating the first norm and second norm of the mean vector of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector; calculating the weighted sum of the inverse of the square root of the second norm and the first norm, and performing point multiplication with the point addition sum vector of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector to obtain the first urine flow-urethral pressure time series interaction response correction subvector; The dot product vector of the timing aggregation representation vector is dotted with the square root of the length of the urine flow-urethral pressure timing interaction response representation vector to obtain a second urine flow-urethral pressure timing interaction response correction subvector; the weighted sum of the first urine flow-urethral pressure timing interaction response correction subvector and the second urine flow-urethral pressure timing interaction response correction subvector is calculated to obtain a urine flow-urethral pressure timing interaction response correction vector; the urine flow-urethral pressure timing interaction response correction vector is dotted with the urine flow-urethral pressure timing interaction response representation vector to obtain the corrected urine flow-urethral pressure timing interaction response representation vector.
[0106] In a specific example, the correction unit is used to calculate the urine flow-urethra pressure time series interaction response correction vector using the following correction formula; wherein the correction formula is expressed as:
[0107] ;
[0108] in, and are respectively the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector, is the mean vector of the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector, is the one-norm of the eigenvector, is the reciprocal of the square root of the two-norm of the eigenvector, It is added by location point. It is the point multiplication by position. is the length of the urine flow-urethral pressure time series interaction response representation vector, and and is the weighted sum weight as a hyperparameter, is the urine flow-urethra pressure timing interaction response correction vector; the urine flow-urethra pressure timing interaction response correction vector is point-multiplied with the urine flow-urethra pressure timing interaction response representation vector to obtain the corrected urine flow-urethra pressure timing interaction response representation vector.
[0109] Here, considering that the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector respectively represent the energy decay time series dynamic aggregation features of urine flow and urethral pressure encoded on local time series nodes based on implicit time series patterns, when they perform feature interactive adaptive response, it is expected to further compensate for the response domain interactive offset adaptive imbalance caused by the correspondence ratio imbalance under the fine-grained distribution of the time series, so as to improve the expression effect of the urine flow-urethral pressure time series interactive response representation vector, thereby improving the accuracy of the classification results.
[0110] Therefore, based on the mean vector norm between the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector, the constrained representation of the structured foreground and background distinction of the supermanifold of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector is used to model the feature-level key correspondence between the feature vectors of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector, and the global correlation relationship between the corresponding features is adjusted, so as to make positive fine-grained correspondence suggestions through the unbalanced proportion control between the corresponding eigenvalues of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector, so as to avoid the imbalance of response domain offset between the feature vectors of the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector by focusing on the focus.
[0111] In this way, by correcting the urine flow-urethral pressure time series interaction response representation vector through the urine flow-urethral pressure time series interaction response correction vector, the expression effect of the urine flow-urethral pressure time series interaction response representation vector can be improved, thereby improving the accuracy of the analysis results obtained by the intelligent analysis module of the urinary tract function state based on the classifier. In this way, non-invasive or minimally invasive sensors can be used to collect urine flow and urethral pressure in real time at different time periods and activity states, reducing the physical discomfort and infection risks caused by the use of catheters and other equipment in traditional urodynamic examinations, and improving the comfort of patients. In addition, AI is used to monitor and analyze the time series of urine flow and urethral pressure in real time, which can timely capture the subtle changes in the time series dynamics of the urinary tract function state, thereby significantly improving the intelligence level of urinary tract function state analysis.
[0112] Based on the above embodiments, see Figure 5 , which is a flow chart of an AI-based intelligent analysis method for urinary tract function status in an embodiment of the present application. For example, the AI-based intelligent analysis method for urinary tract function status can be executed by a server, which can be Figure 1 The server 100 shown in FIG. Figure 5 As shown, according to the embodiment of the present application, the AI-based intelligent analysis method of urinary tract function status includes the following steps: S510, obtaining the time series of urine flow and the time series of urethral pressure collected by the sensor; S520, inputting the time series of urine flow and the time series of urethral pressure into the sequence encoder respectively to obtain the sequence of implicit feature vectors associated with the local time series of urine flow and the sequence of implicit feature vectors associated with the local time series of urethral pressure; S530, inputting the sequence of implicit feature vectors associated with the local time series of urine flow and the sequence of implicit feature vectors associated with the local time series of urethral pressure Input the node feature dynamic aggregation network based on energy time series attenuation respectively to obtain the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector; S540, input the urine flow time series aggregation representation vector and the urethral pressure time series aggregation representation vector into the feature interactive adaptive response module to obtain the urine flow-urethral pressure time series interactive response representation vector as the urine flow-urethral pressure time series interactive response representation feature; S550, based on the urine flow-urethral pressure time series interactive response representation feature, obtain an analysis result, and the analysis result is used to indicate whether the urinary tract function status is normal.
[0113] In one example, the time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder to obtain a sequence of implicit feature vectors associated with local time series of urine flow and a sequence of implicit feature vectors associated with local time series of urethral pressure, including: inputting the time series of urine flow and the time series of urethral pressure into a sequence encoder based on an RNN model to obtain a sequence of implicit feature vectors associated with local time series of urine flow and a sequence of implicit feature vectors associated with local time series of urethral pressure.
[0114] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned AI-based intelligent analysis method of urinary tract function status have been referred to above. Figures 2 to 4 The AI-based urinary tract functional status intelligent analysis system 800 has been introduced in detail, and therefore, its repeated description will be omitted.
[0115] Figure 6 FIG. 1 is an application scenario diagram of an AI-based intelligent analysis system for urinary tract function status according to an embodiment of the present application. Figure 6 As shown, in this application scenario, first, the time series of urine flow and urethral pressure collected by the sensor are obtained (for example, Figure 6 Then, the time series of urine flow and the time series of urethral pressure are input to a server (for example, Figure 6 In S) shown in , the server is able to use the AI-based urinary tract function status intelligent analysis algorithm to process the time series of urine flow and the time series of urethral pressure to obtain an analysis result indicating whether the urinary tract function status is normal.
[0116] Based on the above embodiment, another exemplary embodiment of an electronic device is also provided in the embodiments of the present application. In some possible implementations, the electronic device in the embodiments of the present application may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the AI-based intelligent analysis method for urinary tract function status in the above embodiment when executing the program.
[0117] For example, in the case of electronic equipment Figure 1 Taking the server 100 in the example as an example, the processor in the electronic device is the processor 110 in the server 100, and the memory in the electronic device is the memory 120 in the server 100.
[0118] The embodiment of the present application also provides a computer-readable storage medium, on which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the AI-based intelligent analysis method for urinary tract functional status according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.
[0119] The embodiment of the present application also provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer executable instructions from the computer-readable storage medium, and the processor executes the computer executable instructions, so that the computer device executes the AI-based intelligent analysis method of urinary tract function status according to the embodiment of the present application.
[0120] Those skilled in the art will appreciate that the contents disclosed in this application may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.
[0121] In addition, although the present application makes various references to certain units in the system according to embodiments of the present application, any number of different units can be used and run on the client and / or server. The units are only illustrative, and different aspects of the system and method can use different units.
[0122] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present application is not limited to any particular form of combination of hardware and software.
[0123] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.
[0124] The above is an explanation of the present application and should not be considered as limiting thereof. Although several exemplary embodiments of the present application have been described, those skilled in the art will readily appreciate that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present application.
Claims
1. An AI-based intelligent analysis system for urinary tract function status, characterized in that: include: A urine flow and urethra pressure time series data acquisition module, used to acquire the urine flow time series and urethra pressure time series acquired by the sensor; A urine flow and urethra pressure local time series association module, used for inputting the urine flow time series and the urethra pressure time series into a sequence encoder respectively to obtain a sequence of urine flow local time series association implicit feature vectors and a sequence of urethra pressure local time series association implicit feature vectors; A urine flow and urethra pressure time series aggregation module, used for inputting the sequence of urine flow local time series associated implicit feature vectors and the sequence of urethra pressure local time series associated implicit feature vectors into a node feature dynamic aggregation network based on energy time series decay to obtain a urine flow time series aggregation representation vector and a urethra pressure time series aggregation representation vector; A urine flow urethra pressure time series aggregation feature interaction module, used for inputting the urine flow time series aggregation representation vector and the urethra pressure time series aggregation representation vector into a feature interaction adaptive response module to obtain a urine flow-urethra pressure time series interaction response representation vector as a urine flow-urethra pressure time series interaction response representation feature; A urinary tract function state abnormality judgment module is used to obtain an analysis result based on the urine flow-urethral pressure time series interactive response representation characteristics, and the analysis result is used to indicate whether the urinary tract function state is normal; Wherein, the urine flow and urethral pressure time series aggregation feature interaction module includes: A position-by-position response calculation unit, used for calculating the position-by-position response between the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector to obtain a urine flow-urethra pressure time series aggregate position-by-position response feature vector; A normalization processing unit, used for normalizing the urine flow-urethra pressure time series aggregated position-by-position response feature vector using a Softmax function to obtain a normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector; A urine flow-urethral pressure time series aggregate mask calculation unit, used for inputting the normalized urine flow-urethral pressure time series aggregate position-by-position response feature vector into a learnable gating function to obtain a urine flow-urethral pressure time series aggregate response screening weight mask vector; An aggregated position-by-position response unit, used for calculating the position-by-position point multiplication between the urine flow-urethra pressure time series aggregated response screening weight mask vector and the normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector to obtain a urine flow-urethra pressure time series aggregated position-by-position response distinguishable weight mask vector; An interactive response unit, used for calculating the position-by-position multiplication between the urine flow-urethra pressure time series aggregated position-by-position response distinguishable weight mask vector and the urine flow-urethra pressure time series aggregated position-by-position response feature vector to obtain the urine flow-urethra pressure time series interactive response representation vector; Wherein, the urine flow and urethral pressure time series aggregation module includes: A urine flow local time series associated implicit feature energy statistical paradigm value calculation unit, used to calculate the energy statistical paradigm value of each urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vectors based on the maximum value, average value and variance of each urine flow local time series associated implicit feature vector to obtain a sequence of urine flow local time series associated energy statistical paradigm values, wherein the energy statistical paradigm value corresponding to the current urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated energy statistical paradigm values is used as the current node energy statistical paradigm value and other energy statistical paradigm values are used as historical node energy statistical paradigm values to obtain a sequence of current urine flow local time series associated energy statistical paradigm values and historical urine flow local time series associated energy statistical paradigm values; A node propagation space span value statistical unit is used to count the node propagation space span values between each other urine flow local time series associated implicit feature vector in the sequence of urine flow local time series associated implicit feature vector and the current urine flow local time series associated implicit feature vector to obtain a sequence of urine flow local time series associated propagation space span values; A urine flow local time series energy propagation attenuation coefficient value calculation unit is used to determine the energy propagation attenuation coefficient values of other urine flow local time series associated implicit feature vectors in the sequence of urine flow local time series associated implicit feature vectors based on the sequence of urine flow local time series associated propagation space span values and the sequence of historical urine flow local time series associated energy statistical paradigm values to obtain a sequence of urine flow local time series associated energy propagation attenuation coefficient values, wherein the urine flow local time series associated energy propagation attenuation coefficient values are inversely correlated with the historical urine flow local time series associated energy statistical paradigm values; A weighted sum unit, used to calculate the weighted sum of all other urine flow local time series associated implicit feature vectors in the sequence of urine flow local time series associated implicit feature vectors by using the sequence of urine flow local time series associated energy propagation attenuation coefficient values as a weight sequence to obtain a historical urine flow local time series associated energy attenuation time series aggregated feature vector; an aggregation unit, configured to fuse the current urine flow local time series associated energy statistical paradigm value to calculate the weighted sum of the historical urine flow local time series associated energy attenuation time series aggregation feature vector and the current urine flow local time series associated implicit feature vector to obtain the urine flow time series aggregation representation vector; Wherein, the urine flow local time series associated implicit feature energy statistical paradigm value calculation unit is used for: Extracting the maximum value of the urine flow local time series associated implicit feature vector to obtain the urine flow local time series associated implicit feature maximum value; Respectively calculating the mean value and variance of the urine flow local time series associated implicit feature vector to obtain the urine flow local time series associated implicit feature mean value and the urine flow local time series associated implicit feature variance; The value obtained by adding the variance of the urine flow local time series associated implicit feature and the preset hyperparameter is multiplied by four to obtain the first energy statistical paradigm value of the urine flow local time series associated implicit feature; Calculating the square of the difference between the maximum value of the urine flow local time series associated implicit feature and the average value of the urine flow local time series associated implicit feature to obtain a urine flow local time series associated implicit feature difference value; The second energy statistical paradigm value of the urine flow local time series association implicit feature is obtained by adding the modulated urine flow local time series association implicit feature variance obtained by multiplying the urine flow local time series association implicit feature variance by two, the value obtained by multiplying the preset hyperparameter by two, and the urine flow local time series association implicit feature difference value; Dividing the first energy statistical paradigm value of the urine flow local time series associated implicit feature by the second energy statistical paradigm value of the urine flow local time series associated implicit feature to obtain the energy statistical paradigm value of the urine flow local time series associated; Wherein, the urine flow local time series energy propagation attenuation coefficient value calculation unit is used for: Multiplying each urine flow local time series associated propagation space span value in the sequence of urine flow local time series associated propagation space span values by a first weight hyperparameter to obtain a sequence of urine flow local time series associated first weighted propagation space span values; Using each urine flow local time series associated propagation space span value in the sequence of urine flow local time series associated propagation space span values as an exponent of a natural constant to calculate an exponential function value with base e according to position to obtain a sequence of urine flow local time series associated propagation class support space span values; Multiplying each urine flow local time series association propagation class support space span value in the sequence of urine flow local time series association propagation class support space span values by a second weight hyperparameter to obtain a sequence of urine flow local time series association second weighted propagation space span values; Calculating the positional sum of the sequence of the first weighted propagation space span value associated with the local time series of urine flow and the sequence of the second weighted propagation space span value associated with the local time series of urine flow to obtain a sequence of weighted sum propagation space span value associated with the local time series of urine flow; Divide each group of corresponding historical urine flow local time series correlation energy statistical paradigm values and urine flow local time series correlation weighted sum propagation space span values in the sequence of the historical urine flow local time series correlation energy statistical paradigm values and the sequence of the urine flow local time series correlation weighted sum propagation space span values to obtain the sequence of the urine flow local time series correlation energy propagation attenuation coefficient values.
2. The AI-based intelligent analysis system for urinary tract function status according to claim 1 is characterized in that: The urine flow and urethra pressure local time series association module is used for: The time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder based on an RNN model to obtain a sequence of implicit feature vectors associated with local temporal series of urine flow and a sequence of implicit feature vectors associated with local temporal series of urethral pressure.
3. The AI-based intelligent analysis system for urinary tract function status according to claim 2 is characterized in that: The urine flow-urethra pressure time series aggregation mask calculation unit is used for: The normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector is input into a sigmoid function to calculate a weight mask value to obtain a urine flow-urethra pressure time series aggregated response screening weight mask vector.
4. The AI-based intelligent analysis system for urinary tract function status according to claim 3 is characterized in that: The urinary tract function abnormality judgment module is used to: The urine flow-urethra pressure time series interaction response representation vector is input into a classifier-based urinary tract function state intelligent analysis module to obtain the analysis result, and the analysis result is used to indicate whether the urinary tract function state is normal.
5. An AI-based intelligent analysis method for urinary tract function status, using the AI-based intelligent analysis system for urinary tract function status according to claim 1, characterized in that: include: Acquire a time series of urine flow and a time series of urethral pressure collected by a sensor; Inputting the time series of urine flow and the time series of urethral pressure into a sequence encoder respectively to obtain a sequence of urine flow local time series associated implicit feature vectors and a sequence of urethral pressure local time series associated implicit feature vectors; Inputting the sequence of the urine flow local time series associated implicit feature vectors and the sequence of the urethra pressure local time series associated implicit feature vectors into a node feature dynamic aggregation network based on energy time series decay to obtain a urine flow time series aggregation representation vector and a urethra pressure time series aggregation representation vector; Inputting the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector into a feature interactive adaptive response module to obtain a urine flow-urethra pressure time series interactive response representation vector as a urine flow-urethra pressure time series interactive response representation feature; Based on the urine flow-urethra pressure time series interactive response representation characteristics, an analysis result is obtained, and the analysis result is used to indicate whether the urinary tract function state is normal; The step of inputting the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector into a feature interactive adaptive response module to obtain a urine flow-urethra pressure time series interactive response representation vector as a urine flow-urethra pressure time series interactive response representation feature includes: Calculating the position-by-position response between the urine flow time series aggregate representation vector and the urethra pressure time series aggregate representation vector to obtain a urine flow-urethra pressure time series aggregate position-by-position response feature vector; Using a Softmax function to normalize the urine flow-urethra pressure time series aggregated position-by-position response feature vector to obtain a normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector; Inputting the normalized urine flow-urethra pressure time series aggregated position-by-position response feature vector into a learnable gating function to obtain a urine flow-urethra pressure time series aggregated response screening weight mask vector; Calculate the position point multiplication between the urine flow-urethra pressure time series aggregate response screening weight mask vector and the normalized urine flow-urethra pressure time series aggregate position-by-position response feature vector to obtain a urine flow-urethra pressure time series aggregate position-by-position response distinguishable weight mask vector; The urine flow-urethral pressure time series aggregated position-by-position response is calculated by multiplying the distinguishable weight mask vector by the urine flow-urethral pressure time series aggregated position-by-position response feature vector by the position point to obtain the urine flow-urethral pressure time series interactive response representation vector.
6. The AI-based intelligent analysis method for urinary tract function status according to claim 5, characterized in that: The time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder to obtain a sequence of urine flow local time series associated implicit feature vectors and a sequence of urethral pressure local time series associated implicit feature vectors, including: The time series of urine flow and the time series of urethral pressure are respectively input into a sequence encoder based on an RNN model to obtain a sequence of implicit feature vectors associated with local temporal series of urine flow and a sequence of implicit feature vectors associated with local temporal series of urethral pressure.
