Integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing

Abstract

The invention discloses an integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing, and the system comprises a bladder emptying evaluation module which is used for tracking the form and capacity change of the bladder in real time, and evaluating whether the bladder is completely emptied; the detrusor contraction function evaluation module is used for monitoring the contraction strength and coordination of the detrusor in real time and judging whether the detrusor works normally or not; the bladder pressure evaluation module is used for evaluating the bladder compliance and the detrusor muscle function by monitoring the elastic change and the pressure change of the bladder wall; the urine flow rate measuring module is used for measuring the bladder form change and the urine flow rate in real time and evaluating the urine flow rate; the urethra obstruction and bladder outlet obstruction evaluation module is used for evaluating whether the urethra or the bladder outlet is obstructed or not by monitoring the morphological change and pressure change of the bladder and the urethra in real time; and the data analysis and intelligent diagnosis module is used for providing urination disorder diagnosis and treatment suggestions. The accuracy and operability of urination disorder diagnosis are improved, and the clinical decision-making process is optimized.

Description

Technical Field

[0001] The present invention relates to the field of medical monitoring technology, and in particular to an integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing. Background Art

[0002] Urinary dysfunction is a common urinary system disease, involving multiple physiological processes from bladder storage, urination to urethral discharge. With the aging of the population and changes in lifestyle, the incidence of dysuria and urinary incontinence has gradually increased, which has a significant impact on the quality of life of patients. However, although modern medicine has made certain progress in related diagnostic technologies, accurate diagnosis of urinary dysfunction still faces many challenges. Dysuria and urinary incontinence involve whether the bladder is completely emptied, the contraction function of the detrusor muscle, the assessment of bladder pressure, the determination of urine flow rate, and the assessment of urethral obstruction and bladder outlet obstruction, which are key technical issues.

[0003] First of all, whether the bladder is completely emptied is a fundamental issue in diagnosing dysuria. The normal function of the bladder depends not only on its ability to store urine, but also on its complete emptying during urination. Traditionally, residual urine volume (PVR) assessment is used to measure whether urination is complete, but this method can only provide static results and lacks dynamic and real-time bladder emptying monitoring, making it difficult to accurately reflect the bladder's urination function status. Therefore, how to accurately assess whether the bladder is completely emptied in real time through innovative technologies has become a major challenge in current diagnosis.

[0004] Secondly, the contraction function of the detrusor muscle is a key factor in the bladder urination process. The normal contraction of the detrusor muscle is the prerequisite for maintaining smooth urination, but its contraction strength and coordination are often affected by many factors. The existing ultrasound and electromyography monitoring technology cannot accurately measure the contraction pattern of the detrusor muscle and its relationship with urination, resulting in many causes of dysuria that cannot be identified and treated in time. Therefore, how to combine high-resolution dynamic ultrasound and electromyography monitoring to accurately evaluate the contraction function of the detrusor muscle has become a technical problem that needs to be solved urgently.

[0005] Bladder pressure is another important factor affecting urinary incontinence. The pressure changes of the bladder directly reflect its dynamic ability to store and urinate urine. Traditional bladder pressure assessment usually relies on invasive methods, such as urodynamic examination, which is not only inconvenient for widespread clinical application, but may also cause discomfort to patients. Therefore, how to combine non-invasive ultrasound technology with pressure sensors to monitor bladder pressure changes in real time and then evaluate bladder compliance and detrusor function is an urgent problem to be solved in the diagnosis of urinary disorders.

[0006] In addition, the measurement of urine flow rate is of great significance in the diagnosis of urination disorders, but its accuracy is often affected by factors such as urine flow fluctuations and unstable urine flow rate. Existing ultrasound technology makes it difficult to accurately measure urine flow rate under low or irregular flow rates, resulting in many cases being unable to be accurately diagnosed.

[0007] Finally, urethral obstruction and bladder outlet obstruction are often the root causes of urination disorders, but how to accurately assess the location and type of obstruction and its impact on urination remains a difficult point in clinical practice. Although traditional imaging examinations can provide certain morphological information, they have great limitations in dynamic monitoring and real-time assessment of obstruction conditions. Therefore, how to use ultrasound contrast imaging technology and pressure sensors to monitor the pressure and morphological changes of the urethra and bladder outlets in real time to help assess whether there is obstruction is the key to solving the problem of urination disorders. Summary of the invention

[0008] In order to solve the above problems, the present invention provides an integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing.

[0009] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0010] Integrated monitoring and diagnosis system for urinary difficulty and urinary incontinence based on intelligent sensing, including:

[0011] Bladder emptying assessment module: used to track bladder morphology and capacity changes in real time and assess whether the bladder is completely emptied;

[0012] Detrusor contraction function assessment module: used to monitor the contraction strength and coordination of the detrusor in real time, assess the function of the detrusor, and determine whether it is working normally;

[0013] Bladder pressure assessment module: By monitoring the elasticity and pressure changes of the bladder wall, the pressure inside the bladder is estimated in real time to assess bladder compliance and detrusor function;

[0014] Uroflowmetry module: real-time measurement of bladder morphology changes and urine flow rate, inferring urine flow rate, and evaluating urine flow rate;

[0015] Urethral obstruction and bladder outlet obstruction assessment module: assesses whether there is obstruction in the urethra or bladder outlet by real-time monitoring of the morphological changes and pressure changes of the bladder and urethra;

[0016] Data analysis and intelligent diagnosis module: used to provide urination disorder diagnosis and treatment recommendations.

[0017] Further: the bladder emptying assessment module comprises:

[0018] The bladder is scanned continuously using dynamic three-dimensional ultrasound imaging technology. The three-dimensional structural changes of the bladder are captured in real time through a high-frequency ultrasound probe, thereby obtaining the volume, morphology and capacity data of the bladder at different time points. After obtaining the three-dimensional ultrasound data of the bladder, the three-dimensional model of the bladder is analyzed using computer image processing technology to calculate the real-time volume of the bladder. The bladder area is separated from the background through an image segmentation algorithm, and the boundary points are extracted to calculate the volume of the bladder based on the point cloud data.

[0019] During urination, the bladder volume data is updated in real time to monitor the bladder emptying process. After obtaining the bladder volume data in real time, the three-dimensional morphological changes of the bladder, the changes in bladder volume during urination, and the time points of urination start and end are combined for comprehensive evaluation;

[0020] Combined with the output data of the deep learning model, it predicts whether the bladder is completely emptied and estimates the residual urine volume.

[0021] Further: the detrusor contraction function assessment module includes:

[0022] Use dynamic ultrasound imaging to perform real-time imaging of the detrusor muscle. During urination, the ultrasound probe is used to capture the dynamic changes of the detrusor muscle and its surrounding tissues, and image data of the detrusor muscle during contraction and relaxation are obtained. A linear array ultrasound probe is used in combination with real-time three-dimensional imaging and ultrasound dynamic image processing technology to obtain images of the detrusor muscle, and the images are converted into data on the morphology and thickness of the detrusor muscle.

[0023] Use electromyography monitoring technology to synchronously record the electrical activity of the detrusor muscle. By placing surface electrodes on the patient's perineum or pelvic floor muscle area, the electrical signals of the detrusor muscle are collected in real time to evaluate its electrophysiological response during contraction.

[0024] The morphological and volume changes of the detrusor muscle during contraction provided by ultrasound images, as well as the electrical signals of the detrusor muscle, can be used to comprehensively evaluate the contraction strength and coordination of the detrusor muscle.

[0025] Further: the bladder pressure assessment module includes:

[0026] The elastic characteristics of the bladder wall are obtained through ultrasound elastography. Ultrasound elastography is a technique that estimates the stiffness of the bladder wall by analyzing the changes in the propagation speed of ultrasound in the bladder wall tissue.

[0027] Using a bladder pressure sensor, which is used to measure the direct pressure inside the bladder in real time, capturing the pressure fluctuations inside the bladder through a pressure sensor array;

[0028] Once the real-time intra-bladder pressure data and bladder wall elasticity data are obtained, the bladder compliance model is used to assess the bladder compliance and detrusor function.

[0029] Further: the urine flow rate determination module includes:

[0030] The bladder morphology is monitored in real time through dynamic ultrasound imaging technology, capturing the changes in the bladder volume and morphology during urination. The ultrasound probe provides data on the changes in the bladder during urination by monitoring the three-dimensional structure of the bladder and its capacity changes;

[0031] Based on the bladder morphology change data, a urodynamic monitoring system is set up to monitor the flow rate, flow rate and duration of urine flow through a urine flow sensor to obtain instantaneous flow data of urine flow;

[0032] The ultrasound morphology change data is fused with the urodynamic monitoring data to estimate the urine flow rate and urine flow rate;

[0033] Artificial intelligence analysis will be used to conduct a comprehensive evaluation of the urine flow data obtained to further optimize the urine flow estimation results.

[0034] Further: the urethral obstruction and bladder outlet obstruction assessment module includes:

[0035] The urethral obstruction and bladder outlet obstruction assessment module dynamically monitors the morphology of the bladder and urethra through ultrasound contrast imaging technology;

[0036] The pressure sensor is used to monitor the pressure changes in the bladder and urethra to obtain data related to bladder outlet and urethral obstruction. The pressure sensor is placed in the bladder and urethra outlet area to measure the pressure fluctuations during urine discharge in real time.

[0037] After collecting the morphological and pressure data of the bladder and urethra in real time, the two types of data are analyzed synchronously using a data fusion algorithm to generate a pressure-morphological model of the urethra and bladder outlet, and to infer the specific location and degree of obstruction and its effect on urination.

[0038] Combining the results of fluid dynamics models and data fusion, artificial intelligence algorithms are further used to optimize the diagnosis of urethral and bladder outlet obstruction through machine learning.

[0039] Further: the data analysis and intelligent diagnosis module includes:

[0040] Integrate and preprocess the raw data collected from different modules;

[0041] After data integration and preprocessing, through feature extraction and data modeling, the features meaningful for diagnosis are automatically extracted from the multi-dimensional monitoring data and a mathematical model is constructed;

[0042] After feature extraction is completed, the data is trained using a deep learning algorithm to generate a model for urination disorder diagnosis;

[0043] After the AI ​​deep learning model training is completed, the trained model is used to perform real-time analysis and intelligent diagnosis on the newly input monitoring data. The system can automatically identify and classify different types of urination disorders.

[0044] Compared with the prior art, the present invention has the following technical advances:

[0045] The present invention can effectively realize the integrated monitoring and diagnosis of dysuria and urinary incontinence. First, the system breaks the limitations of traditional single monitoring methods by integrating multiple high-end technologies, such as three-dimensional ultrasound imaging, dynamic ultrasound, ultrasound angiography, pressure sensors, electromyography (EMG) and AI deep learning, and provides a comprehensive and multi-dimensional urination function evaluation method, which not only avoids the errors and limitations that may be caused by a single detection method, but also can track and evaluate multiple key physiological parameters such as bladder, detrusor, urethra in real time, providing more accurate and comprehensive diagnostic support for clinical practice. Secondly, the system greatly improves the comfort and safety of patients through the application of non-invasive technology, avoiding the discomfort and potential risks caused by invasive operations in traditional urodynamic examinations. By real-time monitoring of multiple parameters such as whether the bladder is completely emptied, detrusor contraction function, bladder pressure, urine flow rate and urethral obstruction, the system can comprehensively evaluate the cause of urination disorders and provide a scientific basis for formulating personalized treatment plans. Especially in the measurement of urine flow rate and the assessment of urethral obstruction, the system uses dynamic ultrasound and AI analysis models to accurately identify abnormal urine flow at low or unstable flow rates, making up for the shortcomings of traditional urine flow rate measurement. In addition, the AI ​​deep learning algorithm can integrate multidimensional data from various modules and automatically generate diagnostic reports through intelligent analysis, which greatly improves the efficiency and accuracy of diagnosis and reduces the burden on clinicians. At the same time, the system also has a real-time feedback function, which can provide personalized treatment recommendations based on individual differences of patients, and provide doctors with intuitive diagnostic results and treatment plans through a graphical interface. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0047] In the attached picture:

[0048] Figure 1 It is a system structure diagram of the present invention. DETAILED DESCRIPTION

[0049] The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The embodiments of the present invention will be described below in conjunction with the accompanying drawings.

[0050] like Figure 1 As shown, the present invention discloses an integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing, comprising:

[0051] Specifically, the specific implementation of the bladder emptying assessment module includes:

[0052] The bladder emptying assessment module is designed to accurately assess the completeness of bladder emptying, avoiding reliance on the static indicator of residual urine volume (PVR). At each stage of the urination process, dynamic three-dimensional ultrasound imaging technology is first used to continuously scan the bladder. Through a high-frequency ultrasound probe, the three-dimensional structural changes of the bladder are captured in real time, thereby obtaining data such as the volume, morphology and capacity of the bladder at different time points. The key to this step is to continuously adjust the scanning angle and scanning frequency during urination through the combination of the ultrasound probe and the data acquisition system, thereby obtaining high-resolution three-dimensional ultrasound image data, collecting hundreds of frames of data per second to ensure that the details of every urination moment are not missed.

[0053] After acquiring the three-dimensional ultrasound data of the bladder, computer image processing technology is used to analyze the three-dimensional model of the bladder to calculate the real-time volume of the bladder. This process relies on the image segmentation algorithm to separate the bladder area from the background, extract the boundary points and calculate the volume of the bladder based on the point cloud data.

[0054] V(t)=∫ V(t) f(x,y,z)dxdydz

[0055] Where V(t) represents the volume of the bladder at time point t, f(x,y,z) is the local density function of the bladder surface area, and x,t,z are spatial coordinates.

[0056] During bladder emptying, the system updates the bladder volume data in real time to monitor the bladder emptying process.

[0057] After acquiring the bladder volume data in real time, the deep learning-based AI algorithm is used to further analyze the bladder emptying situation. At this time, the system combines the three-dimensional morphological changes of the bladder, the changes in bladder volume during urination, and the time points of urination start and end for comprehensive evaluation. To this end, this application adopts an innovative hybrid model of convolutional neural network (CNN) and recurrent neural network (RNN).

[0058] Convolutional Neural Network (CNN): First, it is used to analyze the three-dimensional image data of the bladder and extract the key features of the bladder morphology, such as the thickness and shape of the bladder wall and the pattern of bladder wall contraction. CNN can extract spatial features from the image data for preprocessing.

[0059] Recurrent Neural Network (RNN): Next, RNN is used to analyze the dynamic data of bladder volume changes over time. RNN is good at processing time series data and can effectively capture the dynamic characteristics of bladder contraction and emptying during urination. By learning time series, RNN can predict whether the bladder can be completely emptied and determine whether urination is abnormal.

[0060] H(t)=F(H(t-1),X(t),W)

[0061] Among them, H(t) represents the bladder status at time step t (such as bladder capacity, morphology, etc.), X(t) is the input ultrasound image feature, and W is the weight matrix in the convolutional and recurrent neural network.

[0062] Output results: Through the training of the hybrid model, the system can derive the judgment criteria for bladder emptying, output the percentage of bladder emptying, and whether there is incomplete emptying.

[0063] Combined with the output data of the deep learning model, the bladder emptying assessment module can intelligently predict whether the bladder is completely emptied and estimate the residual urine volume. This process comprehensively analyzes the changes in bladder volume and the volume data at the end of urination, and further uses the AI ​​algorithm to infer whether the bladder is incompletely emptied, the amount of residual urine, and whether there are abnormalities during urination (such as difficulty urinating, bladder outlet obstruction, etc.).

[0064] The above process not only breaks through the limitations of traditional static assessment (such as residual urine volume), but also provides a more accurate diagnostic basis for clinical practice, helps to detect bladder emptying problems early, and makes timely interventions in urination disorders.

[0065] The specific implementation of the detrusor contraction function evaluation module includes:

[0066] The detrusor contraction function assessment module combines high-resolution dynamic ultrasound imaging with electromyography (EMG) monitoring technology to monitor the contraction strength and coordination of the detrusor in real time, thereby accurately assessing the function of the detrusor and determining whether it is working properly.

[0067] First, high-resolution dynamic ultrasound imaging technology is used to perform real-time imaging of the detrusor muscle. During urination, a high-frequency ultrasound probe is used to capture the dynamic changes of the detrusor muscle and its surrounding tissues, and image data of the detrusor muscle during contraction and relaxation is obtained. In order to accurately capture the contraction pattern of the detrusor muscle, the ultrasound probe will closely cooperate with the dynamic changes of the bladder and urethra, and scan and generate two-dimensional and three-dimensional images of the detrusor muscle in real time.

[0068] This module uses a linear array ultrasound probe, combined with real-time three-dimensional imaging and ultrasound dynamic image processing technology to ensure the acquisition of high-resolution images and convert the images into detrusor morphology and thickness data. The system can monitor the contraction state of the detrusor in real time and optimize the image quality through image enhancement technology.

[0069] Assuming that at time t, the volume of the detrusor muscle is V(t), its thickness change can be expressed by the following formula:

[0070] V(t)=∫ 肌肉区域 f(x,y,z)dxdydz

[0071] Among them, f(x, y, z) is the density function of the detrusor area, x, y, z are spatial coordinates, and the thickness and volume changes of the detrusor will be used for subsequent contraction strength analysis.

[0072] Next, electromyography (EMG) monitoring technology is used to synchronously record the electrical activity of the detrusor muscle. By placing surface electrodes on the patient's perineum or pelvic floor muscle area, the electrical signals of the detrusor muscle are collected in real time to evaluate its electrophysiological response during contraction. The intensity and frequency of the EMG signal are directly related to the contraction strength and activity state of the muscle, and can provide important information about the detrusor contraction function.

[0073] This process uses surface EMG electrodes, combined with high-pass filters to remove stray signals, and transmits EMG signals in real time through a data acquisition system. The system monitors the electrical activity of the detrusor muscle in real time, records the muscle's contraction pattern and activity intensity, and the EMG signal is output in the form of time series data, providing the electromyogram amplitude (unit: microvolt) at each time point.

[0074] EMG(t)=∫ 肌肉区域 A(x,y,z,t)dxdydz

[0075] Among them, A(x,y,z,t) is the amplitude of electrical activity at a point in the detrusor muscle at time t, and x,y,z are spatial coordinates.

[0076] By synchronously combining high-resolution dynamic ultrasound imaging data and EMG signals, the contraction strength of the detrusor muscle can be comprehensively evaluated. Specifically, ultrasound images provide the morphological and volume changes of the detrusor muscle during contraction, while EMG signals reflect the electrical activity intensity of the muscle. Combining these two data can achieve a comprehensive evaluation of the contraction strength and coordination of the detrusor muscle.

[0077] This module uses a multimodal data fusion algorithm to combine the detrusor morphological changes in ultrasound images with the electrical activity intensity of EMG signals to calculate the detrusor contraction strength index (CCI). This process uses a weighted average algorithm to generate a detrusor contraction strength score based on the temporal correlation between ultrasound images and EMG signals. The contraction strength index formula is:

[0078]

[0079] Where V(t) is the volume of the bladder detrusor muscle at time t, EMG(t) is the amplitude of the EMG signal at time t, and V max and EMG max are the maximum volume and maximum EMG amplitude respectively, ω1 and ω2 are the weight coefficients of ultrasound and EMG signals (adjusted according to the specific situation). By calculating the contraction intensity index CCI(t), the contraction intensity of the detrusor muscle can be quantitatively evaluated.

[0080] By further analyzing the coordination of detrusor contraction and combining the timing relationship between ultrasound data and EMG signals, we can evaluate whether the detrusor contraction is normally coordinated. Under normal circumstances, the detrusor contraction should have a certain degree of synchronization, and its intensity should gradually increase to the maximum value and then recover quickly. In the case of detrusor dysfunction, insufficient contraction or poor coordination may occur.

[0081] A timing analysis algorithm, such as the dynamic time warping (DTW) algorithm, is used to time-align the ultrasound data with the EMG signal and assess the timing consistency between the two. If the detrusor contraction in the ultrasound image is not synchronized with the electrical activity of the EMG signal, the system flags a possible coordination problem.

[0082] The specific implementation of the bladder pressure assessment module includes:

[0083] The bladder pressure assessment module uses ultrasound elastic imaging technology combined with bladder pressure sensors to monitor the elasticity changes of the bladder wall and the pressure changes in the bladder in real time, thereby calculating the internal pressure of the bladder. This module combines dynamic ultrasound and pressure sensor data to not only assess bladder compliance, but also effectively replace traditional urodynamic examinations to provide real-time and accurate bladder function assessment for clinicians.

[0084] First, the bladder pressure assessment module obtains the elastic characteristic data of the bladder wall through ultrasonic elastography technology. Ultrasonic elastography estimates the stiffness (elasticity) of the bladder wall by analyzing the changes in the propagation speed of ultrasound in the bladder wall tissue. The stiffness of the bladder wall is closely related to the compliance of the bladder (that is, the bladder's ability to respond to pressure). Therefore, this technology can help determine the compliance of the bladder and the function of the detrusor muscle.

[0085] This module uses a focused ultrasound array probe and ultrasound elastic imaging algorithms (such as Shear Wave Elastography, SWE) to measure the deformation of the bladder wall under different pressures. SWE technology measures the shear wave speed (SWS) of the bladder wall under stress and infers the elastic modulus (E) of the bladder wall, that is, the stiffness of the bladder wall:

[0086] E=ρ·SWS 2

[0087] Among them, ρ is the density of the tissue, SWS is the shear wave velocity, and by monitoring different pressure changes, the system can calculate the elastic modulus of the bladder wall in real time, thereby indirectly reflecting the bladder compliance.

[0088] At the same time, bladder pressure sensors are used to measure the direct pressure inside the bladder in real time. These sensors are placed in the bladder or through a urethral catheter, capturing pressure fluctuations in the bladder through a pressure sensor array and transmitting data to a computing system in real time. Bladder pressure data can provide specific information about bladder compression status, detrusor function, and bladder capacity changes.

[0089] The bladder pressure sensor uses micro sensors (such as pressure micro sensors or membrane sensors). These sensors can accurately sense the pressure changes in the bladder and have the characteristics of high sensitivity and rapid response. The sensor data is synchronized with the main control system in real time through the wireless transmission module to ensure the real-time and accuracy of the pressure data.

[0090] The change of bladder internal pressure P(t) with time t can be expressed by the following formula:

[0091] P(t)=P0+ΔP(t)

[0092] Where P0 is the static pressure (baseline) of the bladder and ΔP(t) is the instantaneous change in pressure within the bladder.

[0093] Through the data fusion algorithm, the bladder pressure assessment module combines the bladder wall elasticity data provided by ultrasonic elastography and the pressure data provided by the bladder pressure sensor to calculate the pressure state inside the bladder in real time. This combination can effectively improve the accuracy and real-time performance of bladder pressure assessment, while being able to dynamically observe changes in the bladder wall and pressure fluctuations inside the bladder.

[0094] The system uses a weighted fusion algorithm to calculate the comprehensive pressure state of the bladder based on the elastic modulus (E) of the bladder wall measured by ultrasound and the real-time pressure data (P) provided by the pressure sensor. The algorithm sets the weight between the two data sources and dynamically adjusts the weight according to the reliability and accuracy of each data source.

[0095] Bladder internal pressure P total (t) can be calculated comprehensively by the following formula:

[0096] P total (t)=ω1·E(t)+ω2·P(t)

[0097] Among them, E(t) is the elastic modulus of the bladder wall, P(t) is the internal pressure of the bladder, ω1 and ω2 are the weight coefficients of ultrasound data and pressure data respectively. Through this weighted method, the system can comprehensively consider the influence of both and obtain a more accurate bladder pressure estimate.

[0098] Once the real-time pressure data in the bladder and the elasticity data of the bladder wall are obtained, the system will use the bladder compliance model to evaluate the compliance of the bladder and the function of the detrusor muscle. Bladder compliance refers to the ability of the bladder to adapt to pressure during urination, which can reflect the health of the bladder wall elasticity and urination function. Low bladder compliance may indicate problems such as detrusor dysfunction or bladder outlet obstruction. Bladder compliance (Compliance, C) can be calculated by the following formula:

[0099]

[0100] Where ΔV is the change in bladder volume (measured by ultrasound imaging) and ΔP is the change in bladder pressure (measured by a pressure sensor). Low compliance may indicate a stiff bladder wall or abnormal detrusor contraction function.

[0101] Based on the changes in bladder pressure and compliance, the system can infer the function of the detrusor muscle. If the bladder compliance is poor and the pressure rises too quickly, it may indicate detrusor dysfunction or bladder outlet obstruction.

[0102] The specific implementation of the uroflowmetry module includes:

[0103] The uroflowmetry module combines dynamic ultrasound imaging with urodynamic monitoring technology to measure changes in bladder morphology and urine flow rate in real time. It can also accurately calculate urine flow based on fluid dynamics models and artificial intelligence (AI) analysis technology, thereby comprehensively evaluating urine flow rate.

[0104] The first step of the uroflowmetry module is to monitor the bladder morphology in real time through dynamic ultrasound imaging technology to capture the volume and morphological changes of the bladder during urination. The ultrasound probe provides data on the changes of the bladder during urination by monitoring the three-dimensional structure of the bladder and its capacity changes.

[0105] This module uses a high-frequency linear array ultrasound probe, combined with three-dimensional ultrasound imaging technology, to obtain real-time data on bladder morphology and capacity changes. During urination, the system can capture the contraction pattern of the bladder and record the real-time changes in bladder volume. Through the three-dimensional ultrasound image processing algorithm, changes in bladder morphology can be mapped to volume data V(t) in real time, that is, changes in bladder capacity at different time points. The changes in bladder volume V(t) can be expressed by the following formula:

[0106] V(t)=∫ 膀胱区域 f(x,y,z,t)dxdydz

[0107] Among them, g(x,y,z,t) represents the density distribution of each position of the bladder at time t, and x, y, z are three-dimensional coordinates.

[0108] Based on the bladder morphology change data, the urodynamic monitoring system is used to measure the urine flow rate in real time. The system uses a urine flow sensor to monitor the flow rate, flow rate and duration of urine flow. By accurately measuring the time curve of urine flow rate, the system can obtain the instantaneous flow data of urine flow and provide raw data for subsequent urine flow calculation.

[0109] The urodynamic monitoring system is equipped with a high-precision urine flow sensor, which records the urine flow rate (unit: m / s) and flow rate (unit: ml / s) in real time through the sensor array. The urine flow data is transmitted to the central control unit in real time through the digital signal processor (DSP). The urine flow sensor can capture the urine flow velocity u(t) and urine flow rate Q(t) in real time and provide a highly accurate flow rate time curve. The relationship between the urine flow rate u(t) and the urine flow rate Q(t) can be expressed by the following formula:

[0110] Q(t)=A·u(t)

[0111] Where A is the cross-sectional area of ​​the urethra (in cm 2 ), u(t) is the urine flow velocity (unit: cm / s), and Q(t) is the urine flow rate (unit: ml / s).

[0112] The ultrasound morphology change data is fused with the urodynamic monitoring data to achieve accurate calculation of urine flow and urine flow rate. By combining these two types of data simultaneously, the relationship between the flow rate and bladder morphology changes during urine discharge can be more accurately evaluated.

[0113] This module uses a multimodal data fusion algorithm to combine the bladder volume change V(t) and urine flow rate u(t) data. Through a dynamic time series analysis algorithm (such as a Kalman filter or a particle filter), the bladder morphology change and urine flow rate data are integrated to calculate the instantaneous urine flow Q(t). Through this fusion process, the system can calculate the urine discharge rate in real time and consider the influence of factors such as bladder pressure, morphology, and urethral resistance on urine flow. Taking into account the bladder capacity V(t) and urine flow rate u(t), the urine flow Q(t) can be expressed as:

[0114]

[0115] Among them, T(t) is the urination time period, which indicates the duration of bladder urination. The system uses data fusion and time alignment algorithms to ensure the synchronization and accuracy of these data on the same time scale.

[0116] To further improve the accuracy of urine flow estimation, the system will use a fluid dynamics model to model and simulate factors such as pressure, urine flow rate, and bladder wall compliance during urination. The fluid dynamics model can effectively simulate the flow state of urine in the bladder and the relationship between bladder pressure and urine flow rate.

[0117] This module uses the Navier-Stokes equation to simulate the flow of urine in the bladder, taking into account factors such as the morphology of the urethra, bladder pressure, and urine flow rate. In the fluid dynamics model, the pressure and morphology of the bladder affect the flow rate and flow of urine. Therefore, the model combines the pressure-volume relationship and the urodynamic equation to accurately calculate the flow state of urine. In the model, the control equation for urine flow can be expressed by the Navier-Stokes equation:

[0118]

[0119] Among them, u is the flow velocity vector, p is the urine pressure, μ is the liquid viscosity, and ρ is the density of urine. By solving this equation, the system can accurately simulate the flow state of urine and calculate the urine flow rate Q(t).

[0120] Finally, the system will use artificial intelligence (AI) analysis to conduct a comprehensive evaluation of the acquired urine flow data to further optimize the urine flow estimation results. AI models (such as deep neural networks or support vector machines) can learn the complex relationship between bladder morphological changes, urine flow rate and fluid dynamics models, thereby improving the prediction accuracy of urine flow rate.

[0121] The AI ​​model is trained on historical data to optimize the predictive relationship between urine flow rate and bladder capacity, bladder pressure, and urethral morphology. During the training process, the system uses a supervised learning algorithm to adjust the model parameters by minimizing the error function so that it can accurately predict urine flow rate. During the prediction process, the AI ​​model can predict the changing trend of urine flow rate based on the current bladder volume and pressure changes. The AI ​​prediction model can be expressed as:

[0122]

[0123] in, is the urine flow rate predicted by the AI ​​model, f AI It is a mapping function generated by an AI algorithm, and its input includes factors such as bladder capacity V(t), bladder pressure P(t), and urine flow rate u(t).

[0124] The specific implementation of the urethral obstruction and bladder outlet obstruction assessment module includes:

[0125] The urethral obstruction and bladder outlet obstruction assessment module combines ultrasound imaging technology with pressure sensors to provide accurate assessment of urethral and bladder outlet obstruction by real-time monitoring of morphological changes and pressure changes in the bladder and urethra. The core goal of this module is to timely discover and quantify obstruction problems through comprehensive analysis of the structural and functional data of the urethra and bladder outlet, and provide decision support for clinical treatment.

[0126] First, the urethral obstruction and bladder outlet obstruction assessment module dynamically monitors the morphology of the bladder and urethra through contrast-enhanced ultrasound (CEUS). Contrast-enhanced Ultrasound technology combines ultrasound imaging with contrast agents. By injecting microbubble contrast agents, the contrast of ultrasound images is significantly enhanced, thereby more clearly depicting the structural details of the bladder and urethra and helping to identify obstruction areas.

[0127] An appropriate amount of ultrasound contrast agent (such as barium sulfate microbubble solution) is injected into the bladder, and a high-resolution ultrasound probe is used for real-time scanning. The microbubble signal in the ultrasound image can enhance the visualization of the bladder and urethra, and clearly show the movement of the bladder wall and the patency of the urethra. The ultrasound image processing software uses image segmentation and dynamic tracking algorithms to accurately extract the morphological information of the bladder and urethra. By comparing the morphological changes under normal and abnormal conditions, it can be inferred whether there is obstruction. The change in bladder volume V(t) can be expressed as:

[0128] V(t)=∫ 膀胱区域 f(x,y,z,t)dxdydz

[0129] Wherein, f(x,y,z,t) is the density distribution of bladder tissue, and x,y,z are three-dimensional coordinates.

[0130] By tracking the spread of urine-flowing contrast media through the bladder and urethra, the system can assess the urethra for significant areas of obstruction and quantify the extent of the obstruction.

[0131] At the same time, the module monitors the pressure changes in the bladder and urethra through pressure sensors to obtain key data related to bladder outlet and urethral obstruction. The pressure sensors are placed in the bladder and urethral outlet areas to measure the pressure fluctuations of urine during discharge in real time.

[0132] In this module, membrane pressure sensors (such as micro pressure sensor arrays) are used. These sensors can detect real-time pressure fluctuations in the bladder and urethra with extremely high accuracy. Through the wireless transmission module, the pressure data is transmitted to the central processing unit in real time. By analyzing the changing trend of bladder and urethra pressure, the system can identify whether there is an abnormal pressure increase, indicating possible obstruction. The pressure change formula is: the instantaneous pressure P in the bladder bladder (t) and urethral pressure P irethra (t) can be expressed by the following formula:

[0133] P(t)=P0+ΔP(t)

[0134] Among them, P0 is the baseline pressure (i.e., the pressure when there is no urine flow), and ΔP(t) is the pressure increment of the bladder or urethra during urination. If there is an abnormal increase in pressure in the bladder or urethra entrance area (such as a sharp increase in pressure in a short period of time), it may indicate obstruction.

[0135] After collecting the morphological and pressure data of the bladder and urethra in real time, the system will use a data fusion algorithm to synchronously analyze these two types of data. By combining ultrasound contrast imaging data with pressure sensor data, the system can more accurately assess the obstruction of the urethra or bladder outlet.

[0136] The system uses a multimodal data fusion algorithm (such as Kalman filter or particle filter) to jointly analyze the three-dimensional morphological information of the bladder and urethra obtained by ultrasound imaging and the real-time pressure data provided by the pressure sensor. Through this fusion process, the system can generate a pressure-morphological model of the urethra and bladder outlet, and infer the specific location and degree of obstruction and its impact on urination.

[0137] In order to improve the accuracy of obstruction assessment, the system introduces a fluid dynamics model to simulate the flow of urine in the bladder. The model combines the morphological changes of the bladder, the pressure data of the urethra, and the flow characteristics of urine to accurately infer the presence or absence of obstruction.

[0138] The fluid dynamics model simulates urine flow based on the Navier-Stokes equations and the Hagen-Bocelli formula (used to calculate the flow resistance of fluid through a pipe). The system estimates the flow rate of urine based on the pressure data in the bladder and urethra, and identifies whether there is a sharp change in urine flow, indicating an obstruction area. The fluid dynamics model during bladder urination can be described by the following Navier-Stokes equation:

[0139]

[0140] Among them, u is the urine flow rate, ρ is the pressure, μ is the urine viscosity, and ρ is the urine density. By solving this equation, the system can determine the fluid properties of the bladder and urethra and identify the flow rate changes caused by obstruction.

[0141] Combining the results of the fluid dynamics model and data fusion, the system further uses artificial intelligence (AI) algorithms to optimize the diagnosis of urethral and bladder outlet obstruction through machine learning. The AI ​​algorithm can identify different types of obstruction based on training data and provide quantitative assessment.

[0142] The AI ​​model uses convolutional neural networks (CNN) or support vector machines (SVM) to learn the complex relationship between bladder morphology, pressure data, and urine flow status. Through training on a large amount of historical case data, the AI ​​system can accurately classify the type of obstruction (such as urethral stenosis, bladder outlet obstruction, etc.) and provide the corresponding obstruction severity score. The output of the AI ​​algorithm can be expressed as:

[0143]

[0144] in, is the degree of obstruction predicted by the AI ​​model, f AI The diagnostic function generated for the AI ​​model, the input includes the comprehensive pressure P total (t), bladder capacity V(t), urine flow rate u(t) and other data.

[0145] The specific implementation of the data analysis and intelligent diagnosis module includes:

[0146] The data analysis and intelligent diagnosis module is the core component of the integrated urinary difficulty and urinary incontinence monitoring and diagnosis system. It mainly uses AI deep learning algorithms to intelligently analyze the monitoring data from other modules and generate detailed diagnostic reports to provide clinicians with accurate urinary disorder diagnosis and personalized treatment recommendations. The core goal of this module is to provide real-time, accurate and actionable clinical decision support based on large-scale data analysis.

[0147] In the initial stage of the data analysis and intelligent diagnosis module, it is necessary to first integrate and preprocess the raw data collected from different modules (such as bladder emptying assessment module, detrusor contraction function assessment module, bladder pressure assessment module, uroflowmetry module, urethral obstruction and bladder outlet obstruction assessment module). The data provided by each module is usually presented in different formats, including image data, time series data, and scalar data.

[0148] The module receives data streams from various modules through data interfaces (such as JSON or Protobuf format) and uses data cleaning algorithms (such as missing value filling, outlier detection, standardization and normalization) to ensure data consistency and quality. For example, for bladder capacity and urine flow rate data, the z-score standardization method is used to unify the data to the same dimension. For ultrasound image data, image preprocessing algorithms (such as denoising and edge detection) are used to improve image quality.

[0149] After data integration and preprocessing, the module automatically extracts diagnostically meaningful features from multi-dimensional monitoring data through feature extraction and data modeling techniques, and constructs a mathematical model. These features include the rate of change of bladder capacity, fluctuation of urine flow rate, contraction strength of detrusor muscle, pressure changes in bladder and urethra, etc. These features will be used as input for subsequent AI deep learning model training and prediction.

[0150] The feature extraction process includes using time series analysis (such as Fourier transform or wavelet transform) to extract the dynamic change characteristics of the bladder and urethra; image analysis (such as convolutional neural network (CNN)) to extract the morphological characteristics of the bladder and urethra in ultrasound images; and using pressure waveform analysis (such as dynamic time warping (DTW)) to extract the characteristics of pressure changes during urine flow.

[0151] Assume that the frequency spectrum characteristics of the time series data of bladder emptying velocity v(t) and urine flow rate u(t) obtained by Fourier transform are F v (ω) and F u (ω), then:

[0152] F v(ω)=F[v(t)]

[0153] F u (ω)=F[u(t)]

[0154] Where F represents the Fourier transform operation and ω is the frequency variable.

[0155] After feature extraction is completed, the module uses deep learning algorithms (such as convolutional neural networks (CNN), long short-term memory networks (LSTM) or graph neural networks (GNN)) to train the data and generate a model for urination disorder diagnosis. The training process of the deep learning model uses a labeled data set, including known urination disorder symptoms and normal urination samples. The system continuously adjusts the model's weight parameters to minimize the loss function.

[0156] First, use the cross-validation method to evaluate the generalization ability of the model. Second, use the Adam optimization algorithm or the SGD optimization algorithm to train the deep learning network, and adjust the learning rate and regularization parameters according to the training data. During the training process, the Dropout technology can be used to prevent overfitting and improve the stability and robustness of the model. The loss function L(θ) of the AI ​​model can be expressed as:

[0157]

[0158] Where N is the number of samples, y i is the actual label, f(x i ,θ) is the model output, λ is the regularization parameter, ||θ|| 2 is the L2 regularization term.

[0159] After the AI ​​deep learning model training is completed, the system uses the trained model to perform real-time analysis and intelligent diagnosis of the newly input monitoring data. The system can automatically identify and classify different types of urination disorders, such as poor urine flow, detrusor dysfunction, bladder outlet obstruction, etc., and generate corresponding diagnostic results. In addition, the system also has an abnormality detection function that can identify potential abnormal urination behavior or potential pathological problems.

[0160] The system inputs real-time monitoring data (such as bladder capacity changes, urine flow rate fluctuations, bladder pressure, etc.) into the AI ​​model to obtain output results. The model learns the characteristics of different urination disorders from the input data through multiple layers of nonlinear transformations and maps them to specific diagnostic categories, such as "urethral obstruction" and "overactive bladder". The model also analyzes the data in conjunction with anomaly detection algorithms (such as Isolation Forest or One-Class SVM) to identify potential abnormal patterns. Assume that the diagnostic result output by the model is The corresponding category is C, then:

[0161]

[0162] is the predicted output of the model, X is the input data set, and the diagnostic category C obtained by the model through deep learning can be "normal", "urethral obstruction", "bladder outlet obstruction" or other types of urination disorders.

[0163] After completing the intelligent diagnosis, the module generates a detailed diagnostic report. The report not only includes the diagnostic results, but also automatically generates personalized treatment recommendations based on the patient's specific conditions (such as age, gender, medical history, etc.). The treatment recommendations include recommended examination items, possible drug treatments, surgical interventions or physical therapy, etc. After the report is generated, the system will display it to the doctor through a graphical user interface (GUI) to support further clinical decision-making.

[0164] Treatment recommendations are generated based on a combination of a rule engine and an AI model. The rule engine generates treatment recommendations based on diagnostic categories and existing medical literature. For example, if the diagnosis is "bladder outlet obstruction," "bladder outlet pressure measurement" or "bladder dilation therapy" is automatically recommended.

[0165] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the claims of the present invention.

Claims

1. An integrated monitoring and diagnostic system for urinary difficulty and urinary incontinence based on intelligent sensing, characterized in that: include: Bladder emptying assessment module: used to track bladder morphology and capacity changes in real time and assess whether the bladder is completely emptied; Detrusor contraction function assessment module: used to monitor the contraction strength and coordination of the detrusor in real time, assess the function of the detrusor, and determine whether it is working normally; Bladder pressure assessment module: By monitoring the elasticity and pressure changes of the bladder wall, the pressure inside the bladder is estimated in real time to assess bladder compliance and detrusor function; Uroflowmetry module: real-time measurement of bladder morphology changes and urine flow rate, inferring urine flow rate, and evaluating urine flow rate; Urethral obstruction and bladder outlet obstruction assessment module: assesses whether there is obstruction in the urethra or bladder outlet by real-time monitoring of the morphological changes and pressure changes of the bladder and urethra; Data analysis and intelligent diagnosis module: used to provide urination disorder diagnosis and treatment recommendations.

2. The integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing according to claim 1 is characterized in that: The bladder emptying assessment module includes: The bladder is scanned continuously using dynamic three-dimensional ultrasound imaging technology. The three-dimensional structural changes of the bladder are captured in real time through a high-frequency ultrasound probe, thereby obtaining the volume, morphology and capacity data of the bladder at different time points. After obtaining the three-dimensional ultrasound data of the bladder, the three-dimensional model of the bladder is analyzed using computer image processing technology to calculate the real-time volume of the bladder. The bladder area is separated from the background through an image segmentation algorithm, and the boundary points are extracted to calculate the volume of the bladder based on the point cloud data. During urination, the bladder volume data is updated in real time to monitor the bladder emptying process. After obtaining the bladder volume data in real time, the three-dimensional morphological changes of the bladder, the changes in bladder volume during urination, and the time points of urination start and end are combined for comprehensive evaluation; Combined with the output data of the deep learning model, it predicts whether the bladder is completely emptied and estimates the residual urine volume.

3. The integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing according to claim 2 is characterized in that: The detrusor contraction function assessment module comprises: Use dynamic ultrasound imaging to perform real-time imaging of the detrusor muscle. During urination, the ultrasound probe is used to capture the dynamic changes of the detrusor muscle and its surrounding tissues, and image data of the detrusor muscle during contraction and relaxation are obtained. A linear array ultrasound probe is used in combination with real-time three-dimensional imaging and ultrasound dynamic image processing technology to obtain images of the detrusor muscle, and the images are converted into data on the morphology and thickness of the detrusor muscle. Use electromyography monitoring technology to synchronously record the electrical activity of the detrusor muscle. By placing surface electrodes on the patient's perineum or pelvic floor muscle area, the electrical signals of the detrusor muscle are collected in real time to evaluate its electrophysiological response during contraction. The morphological and volume changes of the detrusor muscle during contraction provided by ultrasound images, as well as the electrical signals of the detrusor muscle, can be used to fully evaluate the strength and coordination of detrusor contraction.

4. The integrated urination difficulty and urinary incontinence monitoring and diagnosis system based on intelligent sensing according to claim 3 is characterized in that: The bladder pressure assessment module comprises: The elastic characteristics of the bladder wall are obtained through ultrasound elastography. Ultrasound elastography is a technique that estimates the stiffness of the bladder wall by analyzing the changes in the propagation speed of ultrasound in the bladder wall tissue. Using a bladder pressure sensor, which is used to measure the direct pressure inside the bladder in real time, capturing the pressure fluctuations inside the bladder through a pressure sensor array; Once the real-time intra-bladder pressure data and bladder wall elasticity data are obtained, the bladder compliance model is used to assess the bladder compliance and detrusor function.

5. The integrated monitoring and diagnosis system for dysuria and urinary incontinence based on intelligent sensing according to claim 4 is characterized in that: The urine flow rate measurement module comprises: The bladder morphology is monitored in real time through dynamic ultrasound imaging technology, capturing the changes in the bladder volume and morphology during urination. The ultrasound probe provides data on the changes in the bladder during urination by monitoring the three-dimensional structure of the bladder and its capacity changes; Based on the bladder morphology change data, a urodynamic monitoring system is set up to monitor the flow rate, flow rate and duration of urine flow through a urine flow sensor to obtain instantaneous flow data of urine flow; The ultrasound morphology change data is fused with the urodynamic monitoring data to estimate the urine flow rate and urine flow rate; Artificial intelligence analysis will be used to conduct a comprehensive evaluation of the urine flow data obtained to further optimize the urine flow estimation results.

6. The integrated urination difficulty and urinary incontinence monitoring and diagnosis system based on intelligent sensing according to claim 5 is characterized in that: The Urethral Obstruction and Bladder Outlet Obstruction Assessment Module includes: The urethral obstruction and bladder outlet obstruction assessment module dynamically monitors the morphology of the bladder and urethra through ultrasound contrast imaging technology; The pressure sensor is used to monitor the pressure changes in the bladder and urethra to obtain data related to bladder outlet and urethral obstruction. The pressure sensor is placed in the bladder and urethra outlet area to measure the pressure fluctuations during urine discharge in real time. After collecting the morphological and pressure data of the bladder and urethra in real time, the two types of data are analyzed synchronously using a data fusion algorithm to generate a pressure-morphological model of the urethra and bladder outlet, and to infer the specific location and degree of obstruction and its effect on urination. Combining the results of fluid dynamics models and data fusion, artificial intelligence algorithms are further used to optimize the diagnosis of urethral and bladder outlet obstruction through machine learning.

7. The integrated urination difficulty and urinary incontinence monitoring and diagnosis system based on intelligent sensing according to claim 6 is characterized in that: The data analysis and intelligent diagnosis module includes: Integrate and preprocess the raw data collected from different modules; After data integration and preprocessing, through feature extraction and data modeling, the features meaningful for diagnosis are automatically extracted from the multi-dimensional monitoring data and a mathematical model is constructed; After feature extraction is completed, the data is trained using a deep learning algorithm to generate a model for urination disorder diagnosis; After the AI ​​deep learning model training is completed, the trained model is used to perform real-time analysis and intelligent diagnosis on the newly input monitoring data. The system can automatically identify and classify different types of urination disorders.

Images