Non-invasive bladder pressure detection device and method based on electrical impedance

Abstract

The invention provides a non-invasive bladder pressure detection device and method based on electrical impedance. The non-invasive bladder pressure detection device comprises a waistband assembly and a pressure detection assembly. Wherein the pressure detection assembly comprises a pressure detection host and an electrode limiting piece, and the pressure detection host comprises an electrical impedance acquisition module used for acquiring electrical impedance data of the skin on the outer surface of the bladder area of a patient and a data processor installed on the surface of the electrical impedance acquisition module; the electrode limiting piece comprises a supporting piece and two electrode slice limiting pieces which are installed on the supporting piece and used for limiting the position of an electrode slice. The bladder pressure monitoring system has the advantages that the problems that an existing bladder pressure monitoring technology is high in invasiveness, complex in operation, dependent on professional places, high in cost and the like can be solved, and a neurogenic bladder patient can be helped to achieve comfortable data collection and urination time planning so as to maintain a low-pressure bladder environment; the concurrent risk is reduced, the burden of the patient is relieved, and the medical work efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of bladder pressure detection, and particularly relates to a non-invasive bladder pressure detection device and method based on electrical impedance. BACKGROUND

[0002] Neurogenic bladder is a chronic functional disorder caused by nervous system injury, which is common in patients with spinal cord injury, multiple sclerosis, Parkinson's disease and diabetes. Due to the blockage of signal transmission between the bladder and the nerve center, patients often have symptoms such as delayed urination, urinary retention or high bladder pressure. If not managed in time, it may further cause damage to the bladder wall structure, reflux nephropathy, and even endanger life safety. Effective management of neurogenic bladder depends on accurate measurement and maintenance of the safe range of bladder filling.

[0003] The main technical means of the currently commonly used bladder pressure detection technology is:

[0004] 1) Water column method for measuring bladder pressure: In the existing clinical process, after the neurogenic bladder patient is admitted to the hospital, the doctor will first conduct a preliminary assessment to determine whether bladder pressure and volume measurement is needed according to the assessment results. If needed, the detection process will be completed by the nurse, which may last for 3 hours or more. Before measurement, it is necessary to determine whether the patient has the ability to perceive bladder filling, check whether he has taken any drugs that may affect bladder function, and whether there are any contraindications, such as bladder infection with systemic symptoms, bleeding tendency, autonomic reflex or urethral stricture, etc. The patient needs to prepare to empty the bladder, avoid holding urine, and avoid smoking and drinking. Then prepare the equipment, including adjustable infusion stand, pressure gauge, bladder irrigator with three-way tube, 500ml normal saline, graduated measuring cup or urine pot, etc. After the patient urinates, a sterile urinary catheter is inserted to ensure that the urine is completely drained, and the residual urine volume is recorded. Then the urinary catheter is fixed and connected to the three-way tube. Before measurement, adjust the infusion stand so that the zero point of the pressure gauge is at the same level as the patient's pubic symphysis. During the formal measurement, the normal saline warmed to 35-37℃ is injected into the bladder at an appropriate speed, and the water column fluctuation value is recorded every 50ml of liquid injected. When the water column height rises to 40cm or urine leakage occurs at the urethral orifice, the injection is stopped. After the measurement is completed, the bladder capacity, pressure and residual urine volume are analyzed by recording the data to guide the subsequent bladder training, reduce the risk of infection and improve the ability to control urination. Finally, the measuring device is removed and the bladder is emptied. After the measurement is completed, the doctor will develop a personalized treatment plan based on the results and adjust it in time in combination with the review. The current clinical solution is to measure the bladder pressure using a simple water column method, and then measure the bladder volume using an ultrasonic device to determine whether it is empty. The combination of the two is used to estimate the patient's urination interval to prevent kidney damage due to failure to urinate in time.

[0005] 2)Urodynamic testing: Urodynamic testing is a professional medical device used to comprehensively evaluate the physiological function of the bladder and urethra. It mainly records the changes of multiple physiological parameters during the storage and voiding process to assist in the diagnosis of urinary dysfunction, especially for neurogenic bladder, overactive bladder, detrusor dysfunction, stress urinary incontinence, etc. The core function of this device is to synchronously collect multiple channel data such as intravesical pressure, abdominal pressure, detrusor activity, urine flow rate, and sphincter electromyography to evaluate the compliance of the bladder, the coordination between the detrusor and sphincter, and whether the patient has difficulty urinating or high-pressure storage risk factors. In the specific use process, a catheter is inserted into the bladder under sterile conditions to measure intravesical pressure, and a second catheter is placed in the rectum or vagina to monitor abdominal pressure. The doctor will control the slow infusion of normal saline into the bladder through the device to simulate the natural storage process and record the patient's subjective feelings at different filling stages, such as the first filling sensation and strong urge to urinate. When the patient has a strong urge to urinate, the system guides them to naturally void, and records data such as urine flow rate, voiding duration, and sphincter electromyographic activity during this process. Through a complete detection process, doctors can analyze the relationship between bladder pressure and volume, the compliance of the bladder wall, whether the detrusor is overactive, whether the sphincter has dysfunction, and the coordination between the bladder and the nervous system, thereby providing a scientific basis for subsequent treatment.

[0006] 3)Cystometric testing: When using, a catheter is usually inserted into the patient's bladder under sterile conditions, and a pressure sensor built into the device is connected through the catheter. Some devices integrate infusion functions and can slowly infuse preheated normal saline into the bladder through system control. During the infusion process, the system records the corresponding pressure change in the bladder after each increase in liquid volume in real time and displays it in numerical or chart form on the instrument interface. Medical personnel can judge the compliance of the bladder and whether the bladder has high pressure during the filling process according to the pressure-volume curve, thereby evaluating the safety of urine storage. In addition, key indicators such as maximum bladder capacity and residual urine volume after emptying can be measured during detection to provide a basis for subsequent development of bladder function training or intervention programs. Compared with more complex urodynamic tests, cystometric testing has a relatively simplified operation process and more flexible usage, and generally does not include multi-channel synchronous measurement function. However, this device still requires catheterization and has a certain degree of invasiveness and relies on professional operation, with relatively high usage cost.

[0007] 4) Ultrasound technology: The principle of ultrasound monitoring of the bladder is to use the reflection characteristics of high-frequency sound waves in different tissue interfaces (such as skin, fat, muscle, bladder wall, urine). Specifically, the ultrasound sensor in the device emits high-frequency sound waves into the body, which are reflected when they encounter different density tissue boundaries. The reflected echo is received by the sensor and converted into an electrical signal. In bladder monitoring, the ultrasound device can detect the boundary between the bladder wall and the internal liquid (urine), and then determine the position, shape and filling state of the bladder through the echo intensity and time difference. Some devices also model the three-dimensional structure of the bladder based on multiple angle scan images, to estimate the current bladder volume. The essence of this method is to use the penetration ability and echo characteristics of sound waves to identify soft tissue interfaces, which is commonly used to estimate urine volume, monitor bladder expansion, and determine whether the urination threshold has been reached.

[0008] 5) Optical / chemical sensing technology: Optical sensing technology uses changes in light reflection, transmission or scattering in tissues to infer bladder expansion or tissue tension. Near-infrared spectroscopy, fiber-optic sensing or interferometry are commonly used. This technology has high sensitivity and fast response, and in theory can be used for real-time detection of bladder volume or pressure. However, this technology usually relies on implantable or catheter-based sensors, which are complex and difficult to integrate, and the signal is easily disturbed by body fluids, movement and individual differences. Currently, it is still in the experimental research stage and has not yet formed a mature clinical or commercial solution. Therefore, its application in daily scenarios still has certain limitations.

[0009] The existing technical means for bladder pressure monitoring and function assessment are widely used in clinical practice, but there are obvious limitations in the face of the long-term, autonomous, non-professional daily management needs of neurogenic bladder patients. The traditional water column method, although simple equipment and low cost, is cumbersome to operate, requires a urinary catheter, physiological saline, and other cooperation, the detection process is time-consuming (often more than 3 hours), and there is a risk of infection and patient discomfort, which cannot meet the requirements of frequent or continuous use. Urodynamic detector can provide comprehensive and accurate function assessment data and is considered as the "gold standard", but the detection process highly depends on various catheters and sensors, is highly invasive, complex, and has high operation threshold, and the equipment is expensive, which is only suitable for medical institutions and difficult to promote in family or rehabilitation scenarios. The bladder pressure assessment instrument has been simplified in the process, but it still needs to be inserted into the tube, which is invasive and uncomfortable for patients, and the device is not portable and has high cost, which does not have the feasibility of wide promotion. The ultrasonic technology provides a visual capacity estimation method with certain intuitiveness, but most devices can only realize instantaneous sampling and cannot continuously and dynamically track the bladder state, and the data accuracy is easily affected by factors such as operator technology, body position, and fat layer thickness. The optical sensing technology has shown certain detection potential in experiments, but it often relies on invasive methods and complex structures, the signal is easily disturbed, and it has not yet formed a mature commercial form, and the actual accessibility and stability are limited.

[0010] In summary, the current clinical bladder pressure detection techniques are mostly invasive operations, such as water column method or urodynamic assessment, although they have high data accuracy, but generally have problems such as complex operation process, high cost, strong professional nature, high risk of infection, and obvious patient discomfort. In the subsequent stabilization stage, the urination record tool for assisting medical staff to timely judge the patient's condition and optimize the intervention opportunity is also in paper form, and the medical staff has a heavy burden. Therefore, the core technical problem to be solved by the present application is: how to design a neurogenic bladder management product system with simple structure, low cost, convenient operation, and non-invasive, to help neurogenic bladder patients to realize comfortable data acquisition and urination time planning, to maintain a low-pressure bladder environment, to reduce the risk of complications, and to reduce the burden of patients and improve the work efficiency of medical staff. SUMMARY

[0011] In view of the above-mentioned shortcomings of the prior art, the purpose of the present application is to provide a non-invasive bladder pressure detection device and method based on electrical impedance, which solves the problems of high invasiveness, complex operation, dependence on professional places, and high cost of existing bladder pressure monitoring techniques.

[0012] To achieve the above object and other related objects, the present application provides the following technical solutions:

[0013] The application discloses a non-invasive bladder pressure detection device based on electric impedance, which comprises a waistband assembly and a pressure detection assembly which can be installed on the waistband assembly; wherein the pressure detection assembly comprises a pressure detection host and an electrode limiting piece which can be installed on the pressure detection host; the pressure detection host comprises a protective shell, an electric impedance acquisition module which is installed in the protective shell and is used for acquiring electric impedance data of the skin outside the bladder area of a patient, and a data processor which is installed on the surface of the electric impedance acquisition module and is electrically connected with the electric impedance acquisition module through a wire; the electrode limiting piece comprises a support and two electrode sheet limiting pieces which are installed on the support and are used for limiting the position of electrode sheets.

[0014] In an embodiment of the application, the waistband assembly comprises a skin-friendly waistband which is used for being worn on the waist of a patient and a host support fixing piece which is movably connected with the skin-friendly waistband; a recess which is matched with the shape of the pressure detection host is formed in the host support fixing piece.

[0015] In an embodiment of the application, the protective shell comprises a first shell which is used for placing the electric impedance acquisition module and the data processor and a second shell which can be installed on the first shell; an electrode limiting piece connecting port which can be clamped with the support is formed on the first shell; an electrode wire insertion port which is used for inserting an electrode wire plug is formed on the second shell; a power key, an alarm lamp and a prompt lamp are arranged on the surface of the second shell.

[0016] In an embodiment of the application, a Bluetooth transmission module which is electrically connected with the data processor through a wire is further arranged on the surface of the electric impedance acquisition module; the data processor is connected with a smart terminal device through the Bluetooth transmission module; a power module which is used for providing power supply for each module is further arranged in the first shell.

[0017] In an embodiment of the application, the support comprises a first support block and a second support block which is installed on the first support block and is integrally formed with the first support block; the end of the second support block which is far away from the first support block can be clamped with the electrode limiting piece connecting port; two electrode sheet limiting pieces are arranged on the first support block; a storage box which is used for storing electrode wires is arranged on the first support block and is located between the two electrode sheet limiting pieces; a plurality of electrode wire limiting pieces which are used for limiting the position of electrode wires are arranged on the second support block.

[0018] A non-invasive bladder pressure detection method based on electrical impedance, a non-invasive bladder pressure detection device based on electrical impedance, and applied to a data processor, the method comprising the following steps: obtaining an electrical impedance signal collected in real time by an electrical impedance collection module, and preprocessing the electrical impedance signal; extracting multi-dimensional data features from the preprocessed electrical impedance signal, wherein the multi-dimensional data features include impedance change rate, fluctuation amplitude, signal stability, multi-frequency ratio, phase change pattern, and periodicity characteristics;

[0019] Inputting the multi-dimensional data features into a trained bladder pressure detection model, outputting an estimation result of bladder pressure through the trained bladder pressure detection model, and obtaining an estimation result of filling speed and abnormal state according to the real-time collected electrical impedance signal and the estimation result of bladder pressure.

[0020] In an embodiment of the present application, before obtaining the electrical impedance signal collected in real time by the electrical impedance collection module, the method further comprises: constructing an initial bladder pressure detection model, wherein the initial bladder pressure detection model is a nonlinear regression model constructed using a random forest algorithm; automatically adjusting parameters of the bladder pressure detection model using a Bayesian optimization method, and optimizing the process with cross-validation error as the objective function to obtain the optimal parameter combination through multiple rounds of iterative search, wherein the parameters of the bladder pressure detection model include the number of decision trees, the maximum depth, and the feature sampling ratio.

[0021] The optimized bladder pressure detection model is trained using synchronous clinical pressure measurement or subjective urine intention grade as label data, and in the training process, the random forest model learns the mapping relationship between impedance features and bladder pressure level, thereby obtaining the trained bladder pressure detection model.

[0022] In an embodiment of the present application, the preprocessing includes filtering and denoising processing, drift and baseline correction, and signal quality evaluation.

[0023] In an embodiment of the present application, the outputting of the estimation result of bladder pressure through the trained bladder pressure detection model, and the obtaining of the estimation result of filling speed and abnormal state according to the real-time collected electrical impedance signal and the estimation result of bladder pressure, comprises: identifying the smooth change trend of impedance through the trained bladder pressure detection model, thereby calculating the rising or falling process of bladder pressure, and fitting the dynamic curve of pressure over time; judging the filling speed according to the impedance change rate, confirming the size relationship between the estimation result of bladder pressure and the individual average level, and obtaining the estimation result of abnormal state according to the confirmation result.

[0024] In one embodiment of the present invention, after the estimated result of bladder pressure is output through the trained bladder pressure detection model, and the estimated result of filling speed and abnormal state is obtained based on the real-time acquired impedance signal and the estimated result of bladder pressure, the method further includes: generating an individualized voiding management plan based on the impedance signal acquired in real time by the impedance acquisition module, combined with the estimated results of bladder pressure, filling speed and abnormal state. During the execution of the individualized voiding management plan, the individualized voiding management plan is adjusted in real time based on the real-time updated impedance signal, the new filling trend detected, and the records written by the user on the smart terminal device.

[0025] As described above, the non-invasive bladder pressure detection device and method based on electrical impedance of the present invention has the following beneficial effects:

[0026] 1) The present invention has a reasonable wearing structure and a non-invasive collection method, resulting in a better user experience: The present invention uses an abdominal electrode attachment design, combined with a limiting component and an elastic structure, to ensure consistent wearing position and stable signal each time; no catheterization or professional operation is required, and users can complete the wearing and activation independently, which avoids the discomfort caused by invasive detection and greatly reduces the operation threshold, making it more suitable for patients who use it frequently and for a long time.

[0027] 2) This invention supports continuous monitoring and intelligent analysis, and has strong data transformation capabilities: This invention can not only collect electrical impedance signals in real time, but also has a built-in signal processing module (including filtering, denoising, time series correction, etc.) to ensure stable data quality; on this basis, combined with the random forest regression model, the hyperparameters are adjusted through Bayesian optimization methods, which can dynamically identify the bladder fullness state, predict the timing of urination, and generate personalized urination suggestions, realizing the transformation from "passive monitoring" to "active management";

[0028] 3) This invention constructs a closed-loop interactive system that supports remote viewing and behavior guidance: The accompanying App of this invention integrates urination reminder, recording and voice input functions, allowing users to easily record urination behavior; the data can be synchronized to the cloud in real time, and doctors can view the matching between the patient's actual behavior and the prediction of this invention, which facilitates remote intervention and personalized adjustments, thus constructing a complete "monitoring-reminder-recording-management" closed loop;

[0029] 4) The materials used in this invention are universal, the structure is simple, the cost is controllable, and it is easy to promote and apply: The electrode sheets and electrode wires used in this invention are common materials on the market, and the manufacturing and maintenance costs are low. The core costs are concentrated in the structural design and algorithm development. Compared with existing equipment that is expensive and has a complex structure, this invention is easier to mass-produce and is particularly suitable for low-resource scenarios such as community rehabilitation, telemedicine and home management. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of a non-invasive bladder pressure detection device based on electrical impedance according to the first embodiment of the present invention;

[0031] Figure 2 This is a top view of the various components of the non-invasive bladder pressure detection device based on electrical impedance in this invention;

[0032] Figure 3 This is a three-dimensional schematic diagram of the waist belt component in the non-invasive bladder pressure detection device based on electrical impedance in this invention;

[0033] Figure 4 This is a three-dimensional schematic diagram of the waist belt assembly and pressure detection host in the non-invasive bladder pressure detection device based on electrical impedance in this invention;

[0034] Figure 5 This is a top-view perspective view of the electrical limit positioner with electrode wires and electrode plates installed in the non-invasive bladder pressure detection device based on electrical impedance in this invention.

[0035] Figure 6 This is a side perspective three-dimensional schematic diagram of the electrical limit positioner with electrode wires and electrode plates installed in the non-invasive bladder pressure detection device based on electrical impedance in this invention.

[0036] Figure 7 This is a physical diagram of the non-invasive bladder pressure detection device based on electrical impedance in this invention;

[0037] Figure 8 This is a flowchart of the non-invasive bladder pressure detection method based on electrical impedance according to the second embodiment of the present invention;

[0038] Figure 9 This is a diagram illustrating the internal processing of the bladder pressure detection model in this invention.

[0039] Figure 10 This is a flowchart illustrating the use of the bladder pressure monitoring device by a patient during the initial stage of hospitalization in this invention.

[0040] Figure 11 This is a schematic diagram of the urination management plan generation mechanism in this invention;

[0041] Figure 12 This is a schematic diagram of an APP installed on a smart terminal device in this invention.

[0042] Component designation explanation

[0043] 1. Waist belt assembly; 11. Skin-friendly waist belt; 12. Main unit support and fixing component; 13. Groove; 2. Pressure detection assembly; 21. Pressure detection main unit; 211. Protective shell; 2111. First shell; 2112. Second shell; 212. Impedance acquisition module; 213. Data processor; 214. Bluetooth transmission module; 215. Power module; 216. Electrical limit positioner connection port; 217. Electrode wire socket; 218. Power button; 219. Alarm light and indicator light; 22. Electrical limit positioner; 221. Support component; 2211. First support block; 2212. Second support block; 222. Electrode plate limiting component; 223. Storage box; 224. Electrode wire limiting component. Detailed Implementation

[0044] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. It should be noted that, unless otherwise specified, the following embodiments and features described herein can be combined with each other.

[0045] The first embodiment of the present invention relates to a non-invasive bladder pressure detection device based on electrical impedance. Please refer to [link to relevant documentation]. Figures 1 to 2 The device includes a waist belt assembly 1 and a pressure detection assembly 2 that can be mounted on the waist belt assembly 1. The waist belt assembly 1 includes a skin-friendly waist belt 11 for wearing on the patient's waist and a main unit support and fixing member 12 that is movably connected to the skin-friendly waist belt 11. The main unit support and fixing member 12 has a groove 13 that matches the shape of the pressure detection main unit 21.

[0046] The pressure detection component 2 includes a pressure detection host 21 and an electrical limit switch 22 that can be installed on the pressure detection host 21. The pressure detection host 21 includes a protective shell 211, an impedance acquisition module 212 installed inside the protective shell 211 for collecting impedance data of the skin surface of the patient's bladder area, and a data processor 213 installed on the surface of the impedance acquisition module 212. The data processor 213 is electrically connected to the impedance acquisition module 212 via a wire. The surface of the impedance acquisition module 212 is also provided with a Bluetooth transmission module 214 electrically connected to the data processor 213 via a wire. The data processor 213 is connected to a smart terminal device via the Bluetooth transmission module 214.

[0047] The protective housing 211 includes a first housing 2111 for housing the impedance acquisition module 212 and the data processor 213, and a second housing 2112 that can be installed on the first housing 2111. The first housing 2111 also has a power supply module 215 for providing power to each module. The first housing 2111 has an electrical limit positioner connection port 216 that can be snapped into the support member 221. The second housing 2112 has an electrode wire socket 217 for inserting the electrode wire plug. The surface of the second housing 2112 has a power button 218 and an alarm light and an indicator light 219.

[0048] The electrical limit positioner 22 includes a support member 221 and two electrode plate limiting members 222 mounted on the support member 221 for limiting the position of the electrode plates. The support member 221 includes a first support block 2211 and a second support block 2212 integrally formed with the first support block 2211. The end of the second support block 2212 away from the first support block 2211 can be engaged with the electrical limit positioner connection port 216. The two electrode plate limiting members 222 are disposed on the first support block 2211. The first support block 2211 is also provided with a storage box 223 located between the two electrode plate limiting members 222 for storing the electrode wires. The second support block 2212 is provided with a plurality of electrode wire limiting members 224 for limiting the position of the electrode wires.

[0049] Specifically, electrical impedance tomography (EIT) is a detection method that reflects changes in the internal structure, physiological state, or composition of tissues by measuring changes in their impedance to weak currents. Its core principle is that different tissues or fluids have varying conductivity; when the structure or fluid content of a region in an organism changes, its corresponding electrical impedance value also changes. In actual measurements, the device injects a safe, low-frequency, or multi-frequency current into the target area through one or more pairs of electrodes and measures the voltage change to calculate the tissue's electrical impedance value. This method can be used to assess changes in the morphology, fluid content, or elasticity of tissues in vivo and is widely applied in fields such as cardiopulmonary monitoring, body fluid analysis, and tumor detection.

[0050] More specifically, please see Figures 3 to 6When using the non-invasive bladder pressure monitoring device based on electrical impedance to perform the following steps: First, wear the skin-friendly waist belt 11 around the waist, with its center approximately level with the navel. The main unit support and fixing component 12 is made of soft silicone, facilitating the fixation of the main unit and reducing discomfort. Second, after turning on the pressure monitoring main unit 21, place it into the main unit support and fixing component 12. Third, the electrical limit stop component 22 can be used to fix and store the electrode wires, and helps the user ensure that the electrode pads are placed in similar positions each time, meaning the collected data originates from the same location. Fourth, after placing the electrode wires into the electrode wire limit stop component 224, insert the electrode wires into the pressure monitoring main unit 21, attach the electrode pads according to the position of the limit stop component, and fasten the electrode wires onto the electrode pads. Open the app to begin use. This app can be installed on a smart terminal device. Please refer to the actual image of the non-invasive bladder pressure monitoring device based on electrical impedance to view. Figure 7 .

[0051] The second embodiment of the present invention relates to a non-invasive bladder pressure detection method based on electrical impedance, applied to a data processor 213, the process of which is as follows: Figure 8 As shown, the specific steps include:

[0052] Step 101: Construct an initial bladder pressure detection model.

[0053] Specifically, this invention uses the random forest algorithm to construct a nonlinear regression model. By training multiple decision trees and combining the results, the random forest can effectively handle noisy data, cope with individual differences, and is suitable for complex correlations in physiological signals.

[0054] Step 102: The parameters of the bladder pressure detection model are automatically adjusted using the Bayesian optimization method. The optimization process uses the cross-validation error as the objective function and obtains the optimal parameter combination through multiple rounds of iterative search.

[0055] Specifically, to improve prediction accuracy, the parameters of the bladder pressure detection model are automatically adjusted using a Bayesian optimization method. The parameters of the bladder pressure detection model include the number of decision trees, the maximum depth, and the feature sampling ratio. The optimization process uses the cross-validation error as the objective function and obtains the optimal parameter combination through multiple rounds of iterative search.

[0056] Step 103: The optimized bladder pressure detection model is trained using synchronous clinical pressure measurement or subjective urination level as label data. During the training process, the random forest model learns the mapping relationship between impedance characteristics and bladder pressure level, thereby obtaining the trained bladder pressure detection model.

[0057] Specifically, in the initial model training phase, synchronous clinical pressure measurements or subjective urination urgency levels were used as labeled data, and the random forest model learned the mapping relationship between impedance characteristics and pressure levels. In the model validation phase, performance was evaluated using independent datasets of subjects, including mean error, trend correlation, and high pressure identification accuracy. Results showed that the model could predict bladder pressure change trends with moderate accuracy. For details, please refer to [link to relevant documentation]. Figure 9 .

[0058] Step 104: Obtain the impedance signal acquired in real time by the impedance acquisition module 212, and preprocess the acquired impedance signal.

[0059] Specifically, this invention utilizes multi-electrode impedance measurement technology to inject a weak, safe alternating current into the subcutaneous tissue of the abdomen and collect changes in electrical signals in real time. As the bladder gradually increases in volume and internal fluid content during filling, the conductivity distribution in the pelvic region changes accordingly, thereby causing changes in the amplitude and phase of transabdominal impedance. This change has a stable correlation with the increase in intrabladder pressure. Therefore, the filling state and pressure trend of the bladder can be reflected by changes in impedance signals.

[0060] More specifically, to ensure signal stability and analysis accuracy, the device performs multi-level preprocessing on the impedance signal after acquisition, including: 1) Filtering and denoising: notch and bandpass filtering are applied to the impedance signal to remove power frequency interference and high-frequency noise, while retaining low-frequency components related to respiration, heartbeat, and slow filling; 2) Drift and baseline correction: moving average and trend removal algorithms are used to eliminate slow drift caused by changes in body position, temperature, or skin humidity, so that impedance changes can truly reflect changes in the bladder's conductive environment; 3) Signal quality assessment: signal stability, noise level, and electrode contact status are calculated for each time window to form a quality score, and data below a set threshold are not included in the model calculation.

[0061] Step 105: Extract multidimensional data features from the preprocessed impedance signal.

[0062] Specifically, multiple features for analysis are extracted from the preprocessed impedance signal. These features are multidimensional data characteristics that can reflect the trend of bladder pressure changes, providing a basis for prediction and strategy formulation. These features include impedance change rate, fluctuation amplitude, signal stability, multi-frequency ratio, phase change pattern, and periodicity. These features are used as input to subsequent models to reflect the dynamic changes in bladder status.

[0063] Step 106: Input the multidimensional data features into the trained bladder pressure detection model, output the estimated bladder pressure through the trained bladder pressure detection model, and obtain the estimated results of filling speed and abnormal state based on the real-time acquired impedance signal and the estimated bladder pressure.

[0064] Specifically, in real-time operation, the present invention inputs multidimensional data features into a trained model and outputs an estimation result of bladder pressure. Specifically, the model deduces the process of bladder pressure increase or decrease by identifying the smooth change trend of impedance, thereby fitting a dynamic curve of pressure over time.

[0065] More specifically, the filling speed and abnormal state are estimated based on the real-time collected impedance signal and the estimated bladder pressure. Specifically: 1) Filling speed assessment: The data processor 213 judges the filling speed based on the impedance change rate. When filling is too fast, it indicates that there may be an abnormality in urine storage function. This trend serves as the basis for generating urination reminders and optimizing the plan; 2) Abnormal state identification: When the data processor 213 detects that the pressure change rate is significantly higher than the individual average level and lasts for a certain period of time, it is determined to be a high pressure trend, and a urination reminder is generated in advance; When there is periodic and violent fluctuation or sudden rise in impedance before the bladder has approached its capacity limit, the data processor 213 identifies it as an abnormal activity state and suggests that the user record the symptoms or communicate with a doctor; When the contact impedance increases or the signal noise is too large, the data processor 213 determines that the patch is loose or has poor contact, and prompts the user to re-attach it; Through the above automatic identification rules, the present invention can effectively distinguish between physiological changes and non-physiological interferences in daily monitoring, ensuring the accuracy and safety of urination plan formulation.

[0066] Step 107: Based on the impedance signal collected in real time by the impedance acquisition module 212, and combined with the estimation results of bladder pressure, filling speed and abnormal conditions, an individualized voiding management plan is generated.

[0067] Specifically, the urination management plan mechanism includes:

[0068] 1) Plan Generation: Urinating behavior is influenced by multiple factors, including the severity of neurogenic bladder symptoms, physician experience, and user personal habits. To establish a unified decision-making logic among multi-source heterogeneous data, this invention introduces a knowledge graph method. The knowledge graph defines a unified semantic framework through an ontology layer, which can associate data from different sources and of different types (such as sensor signals, user feedback, and doctor-set rule thresholds) in the same structure, thereby avoiding information fragmentation and improving reasoning consistency and system interpretability. The data processor 213 monitors impedance signals in real time through a pressure detection device and generates an individualized urination management plan by combining the bladder filling trend analysis results. In this process, Protégé is first used to construct a knowledge graph ontology of "patient characteristics - bladder status - strategy rules" to clarify key attributes such as bladder filling speed, urination urge perception, time period, and safety threshold and their hierarchical relationships. After modeling, the data processor 213 imports the ontology into the Neo4j graph database and instantiates it based on actual data to form a semantic network that can be used for reasoning.

[0069] The electrical impedance model provides real-time numerical predictions, while the knowledge graph performs semantic reasoning on this basis, realizing joint decision-making of "data-driven + knowledge-driven". When there is uncertainty or data anomaly in the model output, the data processor 213 can fall back to the explicit rule layer in the graph for judgment, thereby maintaining the stability and security of plan generation. The knowledge graph can continuously record the plan generation, execution and feedback history of each user, forming a knowledge network with a time dimension. The data processor 213 can automatically mine patterns based on graph data, generate new rule nodes and incorporate them into use after verification, so that the urination strategy can continuously self-optimize in long-term operation. At the same time, physicians can directly view the source of the strategy and the relationship between nodes in the graph, understand the decision logic of the data processor 213, and adjust the threshold or modify the rules through the graphical interface without retraining the model, thereby realizing human-machine collaborative safety management.

[0070] 2) Plan Adaptation: When a user first uses this invention, a short-term data collection phase of three days is set to establish the individual's basic physiological mapping relationship and initial instance of the knowledge model. During this period, the pressure detection device continuously collects impedance signals, and the algorithm analyzes the bladder pressure change trend, filling speed, and signal quality. The data processor 213 summarizes the results daily to form the user's baseline filling curve, calculates the average filling time, safe pressure range, and typical filling speed to reflect the individual's natural urination pattern. If the user can perceive the urge to urinate, the urge level and urination time are recorded in the App. The data processor 213 automatically captures the impedance characteristics of the corresponding time period and establishes a correspondence between "urge level - impedance change - pressure level". If the user cannot clearly perceive the urge to urinate, only the urination time needs to be recorded. The data processor 213 automatically extracts the physiological threshold segment based on the impedance change before and after urination to calibrate the individual's urination threshold.

[0071] After standardization, all collected data is written into the knowledge graph database to form personalized nodes, including individual characteristics, impedance characteristic distribution, typical filling curves, safe threshold ranges, and basic strategy templates. Based on this, the data processor 213 performs rule reasoning to generate the first reference voiding plan template, which serves as the baseline input for subsequent daily plan generation. After the three-day phase, the model parameters and graph instances are fixed as the initial state for long-term learning of the individual. The main purpose of this phase is to allow the data processor to fully understand the individual's filling patterns, so that the daily plans generated by the knowledge graph are more in line with the individual's physiological rhythm and threshold distribution. After the third day, the data processor 213 will use the model parameters and ontology instances formed during the sampling period as the initial state for long-term learning. From the fourth day onwards, the user will directly receive the daily voiding plan generated by the system and provide simplified feedback through the App after voiding according to the plan or unplanned, so that the invention can be continuously optimized.

[0072] 3) Plan Adjustment: During plan execution, the pressure detection device continuously updates impedance data, and the data processor 213 monitors new filling trends in real time using a scrolling window, writing the analysis results into the knowledge graph and comparing them with the time windows yet to arrive. When the predicted time to reach the safety threshold is earlier than expected, the data processor 213 automatically moves the next urination time window forward or narrows the current reminder interval. When the filling speed is lower than expected, the time window is appropriately delayed to reduce unnecessary reminders. The adjustment only applies to the remaining time windows of the day, ensuring that the plan is both flexible and stable. After completing urination according to the plan, the user makes a brief record through the App. If unplanned urination occurs, the user actively registers and selects the reason (such as "sudden strong urge to urinate"). These records are standardized by the data processor 213 as execution result events, written into the knowledge graph with a timestamp, and associated with the plan, rule version, and model inference results at that time. The deviation analysis module on the graph side automatically compares the difference between the plan and the actual execution, identifies situations such as early, delayed, or unplanned urination, and records the deviation magnitude and cause. The data processor performs two levels of optimization based on the deviation pattern:

[0073] Daily fine-tuning: If delays occur repeatedly, the data processor 213 increases the priority of subsequent time windows and moves them forward appropriately; if unplanned urination is related to the activity scenario, the reminder method is adjusted or an alternative time window is provided; Periodic optimization: The data processor 213 periodically summarizes deviation information to form new knowledge instances; if similar deviations occur repeatedly in a certain situation, the data processor 213 automatically generates rule revisions, which are then incorporated into the graph after verification in the background, achieving self-evolution; In this way, the graph continuously accumulates and updates the individual's physiological state, planning decisions, and execution results, preserving long-term change trajectories and providing knowledge basis for each plan generation. Therefore, this invention can automatically perform reasoning and pattern recognition on this basis, realizing collaborative decision-making driven by data and knowledge.

[0074] In practical applications, please refer to Figure 10This diagram illustrates the usage process of the present invention in the early stages of hospitalization, used to identify patients' urination patterns and abnormalities. The patient first wears a detection belt 11 based on electrical impedance technology to collect electrical impedance data from the skin of the bladder area, analyzing bladder pressure changes, filling speed, and abnormal conditions. In the early stages of hospitalization, the patient needs to continuously wear the pressure detection device for data collection. The data processor 213 will determine whether there are any abnormalities based on the monitoring results. If no abnormalities are found, a preliminary urination plan is generated based on the collected data and continuously refined and adjusted in conjunction with feedback from the APP. The doctor can discontinue wearing the device after evaluation. If abnormalities are found, such as drastic pressure changes or frequent fluctuations, it is recommended to continue wearing the device for further observation. The data at this stage will be synchronized to the urination management APP to assist in developing personalized urination and hydration plans, while also providing real-time reminders and recording functions to guide subsequent interventions. Following the "monitoring-judgment-feedback-adjustment" process, the scientific nature and individual adaptability of urination management are improved.

[0075] To elaborate further, please refer to Figure 11 This process is divided into three parts: basic examination at the beginning of hospitalization, equipment wearing monitoring, and APP data input, which together build a precise voiding management plan; through short-term testing and data-driven methods, a scientific, safe, and personalized voiding management plan is developed for patients with neurogenic bladder.

[0076] Phase Division: 1) Basic Admission Examination: Develop a basic hydration plan based on the doctor's advice; 2) Three-Day Wearing + Monitoring Period: Day 1: Confirm the appropriate urination pressure range: Assess the correlation between pressure values ​​and urination intention; Data Collection and Preliminary Plan Generation: Build a preliminary urination strategy by combining the pressure-urination relationship; 3) APP-Assisted Optimization: Input hydration and urination data (frequency, time, etc.); The algorithm continuously optimizes the plan; The urination plan is continuously evaluated and fine-tuned based on APP records to improve the accuracy of the plan.

[0077] To elaborate further, please see Figure 12The accompanying app can be used as follows: After connecting to the main unit, the app helps users systematically manage their urination behavior through two main modules: "Urination Record" and "Water Drinking Record." Unlike general data recording applications, it emphasizes device integration, alignment with medical needs, and continuous guidance of behavior. 1) Main Unit Connection and Real-time Monitoring: The app can connect to wearable devices to display bladder pressure data in real time, helping users and medical staff to keep track of the current situation and improve the timeliness and accuracy of management. 2) Urination Record and Abnormal Feedback: Users can record urination on and off schedule. The data processor 213 automatically organizes the time and frequency distribution. At the same time, it provides an abnormal feedback entry point, making it convenient for users to report phenomena such as frequent urination at night, urgency, or leakage, assisting doctors in judging the development of the condition. 3) Water Drinking Record and Habit Formation: By scientifically recording the time, type, and amount of water consumed each day, the app can help users establish regular drinking habits and cross-reference them with urination data to adjust plans and improve treatment compliance.

[0078] In summary, this invention is an intelligent bladder pressure monitoring and urination management system based on electrical impedance acquisition. It covers a complete system from physiological signal acquisition, status recognition, behavior reminders to data recording and remote management, and has core advantages such as non-invasiveness, user-friendliness, high integration and mass production capability.

[0079] 1) Integrated hardware structure and wearable solution: This invention ensures that the surface electrodes stably cover the bladder area during use by setting a specific electrode attachment area on the abdomen, combined with a limiting structure and elastic material, thereby improving signal consistency and data accuracy. This structure also optimizes wearing comfort, enabling users to wear the device independently and collect data continuously during daily activities, laying the foundation for the long-term use and widespread deployment of this invention.

[0080] 2) Electrical impedance data acquisition and individualized state modeling method: This invention obtains the electrical impedance changes in the bladder region through surface electrodes. The built-in data processing module can perform preprocessing operations such as filtering, noise reduction and time series correction. Subsequently, a machine learning-based modeling method is used to dynamically identify the bladder filling trend and combine it with the user's individual characteristics to determine the timing of urination, forming a personalized urination prediction model to improve prediction accuracy and adaptability.

[0081] 3) Data-driven interactive closed-loop system: This invention is accompanied by a mobile APP, which constructs a complete usage process: After receiving the urination reminder, the user completes urination and records the behavior (supporting multiple methods such as click selection and voice input). The data will be automatically linked and analyzed with the information on the device and synchronized to the cloud; the doctor can remotely view the matching degree between the prediction and the actual record, intervene and adjust the patient's urination pattern, and truly realize an integrated user experience from collection → recognition → reminder → feedback → remote management;

[0082] 4) Low cost and adaptability design for telemedicine scenarios: This invention uses conventional electronic components and replaceable electrode materials to control hardware costs, while achieving a high degree of integration of core functions through algorithm and structural design; This invention is suitable for application scenarios with limited resources such as rehabilitation institutions, home environments or telemedicine, and has good scalability and promotion prospects.

[0083] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. All equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this invention should still be covered by the claims of this invention.

Claims

1. A non-invasive bladder pressure detection device based on electrical impedance, characterized in that: Includes a belt assembly (1) and a pressure detection assembly (2) that can be mounted on the belt assembly (1); The pressure detection component (2) includes a pressure detection host (21) and an electrical limit device (22) that can be installed on the pressure detection host (21). The pressure detection host (21) includes a protective shell (211), an impedance acquisition module (212) installed inside the protective shell (211) for collecting impedance data of the skin surface of the patient's bladder area, and a data processor (213) installed on the surface of the impedance acquisition module (212). The data processor (213) is electrically connected to the impedance acquisition module (212) via a wire. The electrical limit device (22) includes a support (221) and two electrode limiters (222) installed on the support (221) for limiting the position of the electrode pads.

2. The non-invasive bladder pressure detection device based on electrical impedance as described in claim 1, characterized in that: The waist belt assembly (1) includes a skin-friendly waist belt (11) for wearing on the patient's waist and a main unit support fastener (12) movably connected to the skin-friendly waist belt (11). The main unit support fastener (12) has a groove (13) that matches the shape of the pressure detection main unit (21).

3. The non-invasive bladder pressure detection device based on electrical impedance as described in claim 1, characterized in that: The protective housing (211) includes a first housing (2111) for housing the impedance acquisition module (212) and the data processor (213) and a second housing (2112) that can be mounted on the first housing (2111); The first housing (2111) has an electrical limit switch connection port (216) that can be engaged with the support member (221), the second housing (2112) has an electrode wire socket (217) for inserting the electrode wire plug, and the surface of the second housing (2112) has a power button (218) and an alarm light and an indicator light (219).

4. The non-invasive bladder pressure detection device based on electrical impedance as described in claim 3, characterized in that: The surface of the impedance acquisition module (212) is also provided with a Bluetooth transmission module (214) electrically connected to the data processor (213) via a wire. The data processor (213) is connected to the smart terminal device through the Bluetooth transmission module (214). The first housing (2111) is also provided with a power module (215) for providing power to each module.

5. A non-invasive bladder pressure detection device based on electrical impedance as described in claim 3, characterized in that: The support member (221) includes a first support block (2211) and a second support block (2212) that is mounted on the first support block (2211) and integrally formed with the first support block (2211); The end of the second support block (2212) away from the first support block (2211) can be engaged with the connection port (216) of the electric limit positioner. The two electrode plate limiting members (222) are disposed on the first support block (2211). The first support block (2211) is also provided with a storage box (223) located between the two electrode plate limiting members (222) for storing the electrode wire. The second support block (2212) is provided with a plurality of electrode wire limiting members (224) for limiting the position of the electrode wire.

6. A non-invasive bladder pressure detection method based on electrical impedance, characterized in that: The method, based on the non-invasive bladder pressure detection device based on electrical impedance according to any one of claims 1-5 and applied to a data processor (213), comprises the following steps: The impedance signal acquired in real time by the impedance acquisition module (212) is obtained, and the impedance signal is preprocessed. Multidimensional data features are extracted from the preprocessed impedance signal, including impedance change rate, fluctuation amplitude, signal stability, multi-frequency ratio, phase change mode, and periodicity. The multidimensional data features are input into the trained bladder pressure detection model, which outputs an estimate of bladder pressure. Based on the real-time acquired impedance signal and the estimate of bladder pressure, the estimation results of filling speed and abnormal state are obtained.

7. The non-invasive bladder pressure detection method based on electrical impedance as described in claim 6, characterized in that: Before acquiring the impedance signal acquired in real time by the impedance acquisition module (212), the following is also included: An initial bladder pressure detection model is constructed, wherein the initial bladder pressure detection model is a nonlinear regression model constructed using the random forest algorithm; The parameters of the bladder pressure detection model are automatically adjusted using the Bayesian optimization method. The optimization process uses the cross-validation error as the objective function and obtains the optimal parameter combination through multiple rounds of iterative search. The parameters of the bladder pressure detection model include the number of decision trees, the maximum depth, and the feature sampling ratio. The optimized bladder pressure detection model was trained using synchronous clinical pressure measurement or subjective urination urge level as label data. During the training process, the random forest model learned the mapping relationship between impedance characteristics and bladder pressure level, thus obtaining the trained bladder pressure detection model.

8. A non-invasive bladder pressure detection method based on electrical impedance as described in claim 6, characterized in that: The preprocessing includes filtering and denoising, drift and baseline correction, and signal quality assessment.

9. A non-invasive bladder pressure detection method based on electrical impedance as described in claim 6, characterized in that: The process involves outputting an estimated bladder pressure result through a trained bladder pressure detection model, and obtaining estimates of filling speed and abnormal states based on real-time acquired impedance signals and the estimated bladder pressure result, including: By identifying the smooth change trend of impedance through a trained bladder pressure detection model, the process of bladder pressure increase or decrease can be deduced, and a dynamic curve of pressure over time can be fitted. The filling rate is determined by the rate of change of impedance, and the relationship between the estimated bladder pressure and the individual average level is confirmed. Based on the confirmation results, the estimated result of the abnormal state is obtained.

10. A non-invasive bladder pressure detection method based on electrical impedance as described in claim 6, characterized in that: After the trained bladder pressure detection model outputs an estimate of bladder pressure, and the estimation results of filling speed and abnormal states are obtained based on the real-time acquired impedance signal and the bladder pressure estimation results, the method further includes: Based on the impedance signal collected in real time by the impedance acquisition module (212), combined with the estimation results of bladder pressure, filling speed and abnormal state, an individualized voiding management plan is generated. During the execution of the individualized voiding management plan, the individualized voiding management plan is adjusted in real time based on the real-time updated impedance signal, the new filling trend detected, and the records written by the user on the smart terminal device.

Images