Wearable catheter pressure monitoring system and method based on AI

Abstract

The invention discloses an AI-based wearable catheter pressure monitoring system, which comprises an initial starting module, which is used for determining initial starting time by using a classification model according to obtained initial urethral pressure data of a target patient and an original medical data set and the initial urethral pressure data of the target patient, and sending a pressure detection starting instruction; the pressure sensor monitors the internal pressure of the balloon of the catheter in real time at a certain monitoring frequency; and the dynamic monitoring module is used for receiving the internal pressure of the catheter balloon and dynamically adjusting the monitoring frequency of the pressure sensor in different nursing stages of the target patient by using a time sequence model according to the internal pressure of the catheter balloon. The system provided by the invention can intelligently adjust the monitoring frequency according to the dynamic change of the urethral pressure of the patient, predicts the service life of the equipment, realizes low-power-consumption operation and improves the monitoring efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment, and in particular to an AI-based wearable urinary catheter pressure monitoring system and method. Background Art

[0002] In the medical field, the care and monitoring of patients undergoing indwelling urinary catheterization remains a significant and challenging issue. During the use of traditional urinary catheters, medical staff struggle to accurately and in real time monitor the pressure within the catheter's balloon. This can lead to abnormal balloon pressure, causing patient discomfort and even complications such as urethral mucosal damage and infection. Furthermore, existing urinary catheters lack effective pressure monitoring and early warning mechanisms, making it impossible to promptly detect ruptures or leaks within the balloon, which can lead to serious consequences such as pain, urinary retention, urethral bleeding, and permanent bladder damage.

[0003] While there are some pressure monitoring devices currently available on the market, these implantable devices suffer from issues such as high power consumption, limiting their application in catheter balloon pressure monitoring. Therefore, designing a system that can accurately monitor catheter balloon pressure while maintaining low power consumption is an urgent need. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an AI-based wearable catheter pressure monitoring system and method to address the shortcomings of the existing technology. The system can intelligently adjust the monitoring frequency according to the dynamic changes of the patient's urethral pressure, achieve low-power operation and improve monitoring efficiency.

[0005] To achieve the above objectives, according to one aspect of the present invention, a wearable urinary catheter pressure monitoring system based on AI is provided, comprising: An initial startup module obtains the initial urethral pressure data of the target patient and uses a classification model to determine the initial startup time based on the original medical data set and the initial urethral pressure data of the target patient, and then enters a dormant state; after the initial startup time is reached, a pressure detection startup instruction is sent to enable the pressure sensor to monitor the internal pressure of the catheter balloon in real time at a certain monitoring frequency; The dynamic monitoring module is used to receive and dynamically adjust the monitoring frequency of the pressure sensor at different nursing stages of the target patient based on the internal pressure of the catheter balloon using a time series model.

[0006] In the above solution, the AI-based wearable urinary catheter pressure monitoring system also includes: A battery life prediction module is used to use the pressure data in the catheter balloon to establish a linear regression equation to predict the remaining battery life of the AI-based wearable catheter pressure monitoring system.

[0007] In the above solution, the AI-based wearable catheter pressure monitoring system also includes a cloud, which can receive the raw pressure data in the catheter balloon monitored in real time for incremental training and strategy iterative optimization of the classification model and the time series model.

[0008] In the above solution, the method of using the classification model to confirm the initial start time and send the pressure detection start instruction so that the pressure sensor monitors the pressure inside the catheter balloon in real time at a certain monitoring frequency is: Extracting multidimensional features from the medical big data of the original medical data set as initial feature vectors, and using the time of the first occurrence of catheter abnormality in the original medical data set as label data, to construct a training set through feature engineering; wherein the multidimensional features include the patient's initial urethral pressure value, the patient's age, the patient's gender, the catheter model, and the patient's previous complication record; The classification model is used to discretize the time of the first occurrence of the urinary catheter abnormality into multiple time windows, and the classification model is trained using the training set to output a probability distribution of each time window, wherein the input of the classification model is the initial feature vector; The initial urethral pressure data of the target patient is collected in real time, and the trained classification model is inputted in combination with the multi-dimensional features of the target patient to obtain the probability distribution of each time window, and the time window with the highest probability is selected as the initial start time. A pressure detection start instruction is sent so that the pressure sensor monitors the internal pressure of the catheter balloon in real time at a certain monitoring frequency.

[0009] In the above scheme, the method of dynamically adjusting the monitoring frequency of the pressure sensor at different nursing stages of the target patient using the time series model is as follows: Collecting time series data of pressure in the catheter balloon in a sliding window manner, and calculating the pressure mean, standard deviation, and pressure change rate; defining stage labels of different nursing stages based on the original medical data set, wherein the different nursing stages include an initial stage, a stable stage, and a critical stage; Constructing and training a time series model, inputting the pressure time series data in the catheter balloon within the sliding window, the pressure mean, the pressure change rate, and the standard deviation into the trained time series model, and outputting the probability distribution of the sliding window being in the initial period, the stable period, and the critical period; The monitoring frequency of the pressure sensor is dynamically adjusted according to the output probability distribution results: the monitoring frequency is increased in the initial stage, the monitoring frequency is reduced in the stable period, and the monitoring frequency is significantly increased in the dangerous period and an early warning is triggered.

[0010] In the above scheme, the initial stage is defined as the pressure fluctuation rate not exceeding 5% or the pressure value is always lower than the clinical safety threshold; The stable period is defined as: the pressure fluctuation rate is between 5% and 10% or the duration of the pressure value exceeding the clinical safety threshold does not exceed the first time; The critical period is defined as the pressure fluctuation rate exceeding 10% or the duration of the pressure value exceeding the clinical safety threshold exceeding the first time.

[0011] In the above solution, the method for using the pressure data in the catheter balloon to establish a linear regression equation to predict the remaining life of the AI-based wearable catheter pressure monitoring system is: The average power consumption of the AI-based wearable catheter pressure monitoring system is obtained by using the pressure data test inside the catheter balloon in the early stage. Combined with the current remaining power of the AI-based wearable catheter pressure monitoring system, a linear regression equation is established to predict the remaining battery life of the AI-based wearable catheter pressure monitoring system.

[0012] In addition, the present invention also proposes an AI-based wearable urinary catheter pressure monitoring method, comprising: Acquire the initial urethral pressure data of the target patient, and use the classification model to determine the initial start time based on the original medical data set and the initial urethral pressure data of the target patient, and then enter a dormant state; after the initial start time is reached, send a pressure detection start instruction so that the pressure sensor monitors the pressure inside the catheter balloon in real time at a certain monitoring frequency; The pressure sensor's monitoring frequency at different stages of care for target patients is dynamically adjusted using a time series model based on the catheter balloon's internal pressure.

[0013] Another aspect of the present invention further provides a device for monitoring the internal pressure of a urinary catheter balloon, comprising: A pressure detection sensor module is used to convert the pressure signal in the catheter balloon detected by the pressure detection sensor into an electrical signal pressure sensor; A main control chip module is used to obtain pressure data from the pressure detection sensor module to complete the above-mentioned AI-based wearable catheter pressure monitoring method; The power supply module is used to provide power support.

[0014] In the above solution, the catheter balloon internal pressure monitoring device further includes: a wireless communication module for transmitting data with an associated cloud or mobile terminal; The time series model is deployed on the main control chip module, which can quickly realize the mutual transmission of data; The power supply module has a built-in micro button battery.

[0015] In the above solution, the wireless communication module adopts a Bluetooth module based on the BLE 5.1 ​​protocol.

[0016] In the above solution, the pressure detection sensor module is used to convert the pressure signal in the catheter balloon into an electrical signal through a Wheatstone bridge circuit; the measurement accuracy of the pressure sensor is ≤150Pa.

[0017] In the above solution, the mobile terminal can display the pressure data and warning information in the catheter balloon in real time, and remind the user that the pressure monitoring device in the catheter balloon is low on power and needs to be replaced.

[0018] In this invention, AI technology is used to optimize the power management mechanism during catheter pressure monitoring. Based on the analysis, the sleep time and monitoring frequency of the monitoring system are dynamically adjusted to achieve low-power operation. The specific mechanism is as follows: (1) Optimization of initial startup time and initial monitoring frequency Medical big data experience shows that the time when catheter abnormalities begin to appear is generally related to the patient's urethral pressure. After the AI-based wearable catheter pressure monitoring system of the present invention is activated, it first performs an initial urethral pressure monitoring to obtain initial pressure data. By analyzing this initial data and combining it with an empirical data set of medical big data on the distribution characteristics of urethral pressure in different patients and the time patterns of catheter abnormalities, the initial startup time and initial monitoring frequency are dynamically determined. Because the probability of catheter abnormalities in the early stage is low, the present invention greatly extends the endurance of the monitoring system by setting an initial startup time and reducing the initial monitoring frequency.

[0019] (2) Adjustment of monitoring frequency during the stable period Based on medical big data experience, the stable period refers to the time when the catheter is operating normally and smoothly. During this period, the likelihood of catheter anomalies is low, resulting in a low probability of missed detection, allowing for a reduced monitoring frequency. During the monitoring process, by analyzing the initial measurement data, the monitoring frequency is automatically reduced when the pressure data stabilizes and, based on medical big data experience, the catheter enters a stable period with a low probability of anomalies. This strategy significantly reduces the system's power consumption during the stable period while maintaining basic monitoring capabilities for anomalies.

[0020] (3) Adjustment of monitoring frequency during the critical period The critical period is when, based on medical experience, the likelihood of catheter abnormalities increases significantly. During monitoring, the system analyzes measurement data from the initial and stable periods, combined with the patient's urethral pressure trends, to predict when the catheter will enter the critical period. During this period, the system automatically increases monitoring frequency to ensure timely detection of abnormalities and trigger early warnings, thereby improving patient safety.

[0021] (4) Equipment life prediction The system can also analyze the device's power consumption data and remaining power, and combine historical usage data to predict when the device will run out of power. The system reminds users in advance through mobile terminals to replace the device, ensuring the continuous operation of the monitoring system and avoiding monitoring interruptions due to insufficient power.

[0022] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects compared with the prior art: (1) The present invention provides an AI-based wearable catheter pressure monitoring system. The system solves the problems of fixed monitoring frequency and high power consumption of existing catheter pressure monitoring systems by dynamically adjusting the monitoring frequency. In the initial stage, the system quickly obtains the patient's urethral pressure characteristic data through high-frequency monitoring and establishes a pressure baseline. In the stable period, the monitoring frequency is reduced, which significantly reduces power consumption. In the critical period, the monitoring frequency is increased to ensure timely detection of abnormalities. This dynamic adjustment mechanism reduces the overall energy consumption of the system through causal relationships, while improving monitoring efficiency and patient safety.

[0023] (2) The AI-based wearable catheter pressure monitoring system provided by the present invention uses a micro button battery and combines low-power sleep technology to ensure long-term battery life; at the same time, by optimizing the power management strategy, the system reduces the monitoring frequency during the stable period, further reducing power consumption, so that the system can maintain long-term operation while being miniaturized, significantly improving the patient's wearing comfort and ease of use.

[0024] (3) The AI-based wearable catheter pressure monitoring system provided by the present invention can timely warn of catheter abnormalities (such as balloon rupture or catheter detachment) by real-time analysis of pressure data, effectively preventing patients from complications caused by abnormal pressure. In addition, the system can predict the life of the device by combining power consumption data and remaining power, and remind users to replace the device in advance through mobile terminals to avoid monitoring interruptions caused by insufficient power, significantly improving the reliability of the system and the safety of patients.

[0025] (4) The AI-based wearable catheter pressure monitoring system provided by the present invention adopts a high-precision pressure sensor, which can accurately capture the pressure changes in the catheter balloon, ensuring that the system can detect abnormal conditions in a timely manner, avoiding misjudgment or missed detection due to monitoring errors, and thus improving the monitoring quality.

[0026] (5) The AI-based wearable catheter pressure monitoring system provided by the present invention continuously learns and optimizes the analysis model. As the monitoring data accumulates, the prediction accuracy gradually improves, enabling the system to dynamically adjust the monitoring strategy and warning threshold according to the individual differences and clinical needs of patients, providing more accurate medical protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. Throughout the drawings, the same reference figures denote the same components. In the drawings: Figure 1 This is a structural diagram of an AI-based wearable urinary catheter pressure monitoring system in Example 1 of the present invention.

[0028] Figure 2 The figure is a schematic diagram of the workflow of an AI-based wearable urinary catheter pressure monitoring system in Example 1 of the present invention. DETAILED DESCRIPTION

[0029] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0030] It should be understood that the size of the serial numbers of the steps in the embodiment does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0031] Example 1 The present application embodiment provides an AI-based wearable urinary catheter pressure monitoring system, please refer to Figure 1 and Figure 2 , Figure 1 Schematic diagram of an AI-based wearable urinary catheter pressure monitoring system in an embodiment of the present application. Figure 2 The following is a schematic diagram of the workflow of an AI-based wearable urinary catheter pressure monitoring system in an embodiment of the present application. The AI-based wearable urinary catheter pressure monitoring system in an embodiment of the present application includes: The initial startup module obtains the initial urethral pressure data of the target patient and uses the classification model to determine the initial startup time based on the original medical data set and the initial urethral pressure data of the target patient, and then enters a dormant state. After the initial startup time is reached, a pressure detection startup instruction is sent to enable the pressure sensor to monitor the pressure inside the catheter balloon in real time at a certain monitoring frequency; The dynamic monitoring module is used to receive and dynamically adjust the monitoring frequency of the pressure sensor at different nursing stages of the target patient based on the internal pressure of the catheter balloon using a time series model.

[0032] It can be understood that, in this embodiment, Figure 1As shown, the catheter used is an improved double-lumen catheter, which is a conventional double-lumen catheter with a third catheter added. The front end of the third catheter is connected to the balloon and the tail end is sealed. A pressure sensor is provided at the tail end of the third catheter of the improved double-lumen catheter. The pressure sensor can obtain the air pressure in the balloon by measuring the air pressure of the third catheter. The initial startup module and the dynamic monitoring module in this embodiment are provided in the AI ​​module, and the AI ​​module can interact with the associated APP for data.

[0033] Specifically, in this embodiment, the classification model adopts a random forest classification model. The random forest classification model is used to confirm the initial start time and send a pressure detection start instruction so that the pressure sensor monitors the catheter balloon internal pressure in real time at a certain monitoring frequency. The method is as follows: S11, Data Preparation and Feature Construction Multi-dimensional features are extracted from the medical big data of the original medical dataset as the initial feature vector, including static features such as the patient's initial urethral pressure value P0, patient age, patient gender, catheter model, and patient's previous complication record, as well as the time of the first abnormality of the catheter T0 as label data. A training set is constructed through feature engineering, with the goal of establishing a mapping relationship between the initial features and the abnormal time window.

[0034] S12, model training and classification rule generation A random forest classification model was used to discretize the time of the first catheter abnormality into multiple time windows (specifically, 0-12 hours, 12-24 hours, 24-48 hours, and 48-72 hours in this example). The random forest classification model takes as input the initial feature vector and outputs the probability distribution Pwindow for each time window. During training, the Gini coefficient was used to optimize the splitting rules of the decision tree to generate classification boundaries. Finally, the probability distribution was determined by ensembling multiple decision trees and voting. The random forest classification model was validated using a cross-entropy loss function and K-fold cross-validation to ensure generalization performance.

[0035] S13, dynamic decision on startup time After the AI-based wearable urinary catheter pressure monitoring system of this embodiment is initialized, the pressure sensor first collects the target patient's initial urethral pressure value P0. Combined with other characteristics of the target patient (target patient age, target patient gender, catheter model, target patient's previous complication record), the trained random forest model is input to obtain the probability distribution of each time window, and the time window with the highest probability is selected as the initial startup time of the system. T 开机 , the formula is as follows:

[0036] After confirming the initial start time, the system enters a dormant state, and sends a pressure detection start instruction after the initial start time is reached, so that the pressure sensor monitors the pressure inside the catheter balloon in real time at a certain monitoring frequency.

[0037] Specifically, in this embodiment, the time series model adopts an LSTM time series model, receives and dynamically adjusts the monitoring frequency of the pressure sensor at different nursing stages of the target patient based on the internal pressure of the catheter balloon using the LSTM time series model: S21, collecting the pressure time series data P in the catheter balloon in a sliding window manner (in this embodiment, a period of 10 minutes) t , calculate the mean pressure , standard deviation σ P and pressure change rate ; Based on the original medical data set, the stage labels of each time period are defined, including the initial stage, stable stage, and critical stage; Specifically, in this embodiment, the initial stage is defined as the pressure fluctuation rate not exceeding 5% or the pressure value is always lower than the clinical safety threshold (2500 Pa in this embodiment); The stable period is defined as: the pressure fluctuation rate is between 5% and 10% or the duration of the pressure value exceeding the clinical safety threshold does not exceed the first time (5 minutes in this embodiment); The critical period was defined as the pressure fluctuation rate exceeding 10% or the duration of the pressure value exceeding the clinical safety threshold for more than the first time.

[0038] S22, construct and train a time series model, input the pressure time series data, pressure mean, pressure change rate, and standard deviation in the catheter balloon within the sliding window into the trained time series model, and output the probability distribution of the sliding window being in the initial period, stable period, and critical period; Specifically, in this embodiment, the network structure of the LSTM time series model is as follows: the input layer receives the pressure time series data and related data in the sliding window (dimension: T×4, T is the time step, 3 is the pressure time series data P in the catheter balloon). t , pressure mean , pressure change rate , standard deviation of pressure data σ P Features); LSTM layer, double hidden layer structure (64 units per layer), captures the long-term dependency and short-term fluctuation characteristics of the pressure sequence through the gating mechanism; output layer, Softmax function outputs the probability distribution of the sliding window in the initial period, stable period, and dangerous period (P stage ∈ [0,1] 3 ).

[0039] The training strategy of the LSTM time series model is as follows: the loss function uses weighted cross-entropy loss, giving higher weights to the dangerous period label to enhance sensitivity to high-risk stages; the optimizer uses the Adam optimizer (learning rate η = 0.001) combined with the early stopping method (patience = 10) to prevent overfitting; data augmentation improves model robustness by adding Gaussian noise (σ = 50 Pa) and time series interpolation.

[0040] S23, dynamically adjust the monitoring frequency of the catheter balloon internal pressure monitoring device according to the output probability distribution result: increase the monitoring frequency in the early stage, reduce the monitoring frequency in the stable period, and significantly increase the monitoring frequency in the dangerous period and trigger an early warning.

[0041] Specifically, in this embodiment, the AI-based wearable catheter pressure monitoring system dynamically adjusts the monitoring frequency based on the probability distribution output by the LSTM time series model. In the initial stage, when the LSTM time series model predicts a high initial probability, the system increases the monitoring frequency to quickly obtain a baseline of the target patient's pressure characteristics. After entering the stable period, if the LSTM time series model determines that the stable period probability is high and the pressure fluctuation rate is low, the system reduces the monitoring frequency to reduce the power consumption of the catheter balloon internal pressure monitoring device while maintaining basic monitoring capabilities. During the critical period, if the LSTM time series model predicts a high probability of a critical period or detects a sudden change in pressure, the system immediately increases the monitoring frequency significantly and triggers an alert. A mobile terminal associated with the system (specifically, a mobile app in this embodiment) can display pressure data in real time and receive alerts, notifying relevant medical staff or the target patient to ensure that abnormal conditions are detected promptly.

[0042] In this embodiment, the AI-based wearable urinary catheter pressure monitoring system further includes: The battery life prediction module is used to use the pressure data in the catheter balloon to establish a linear regression equation to predict the remaining battery life of the AI-based wearable catheter pressure monitoring system. The specific method is as follows: The average power consumption of the catheter balloon internal pressure monitoring device was obtained by using the pressure data of the previous catheter balloon test. Combined with the current remaining power of the catheter balloon internal pressure monitoring device, a linear regression equation was established to predict the remaining battery life of the catheter balloon internal pressure monitoring device. The specific formula is as follows:

[0043] in, T 剩余 To predict the remaining life (battery exhaustion time) of the catheter balloon internal pressure monitoring device, E 剩余 The current remaining power of the catheter balloon internal pressure monitoring device. P 平均is the average historical power consumption of the catheter balloon internal pressure monitoring device.

[0044] Another aspect of the present application provides an AI-based wearable urinary catheter pressure monitoring method, comprising: By acquiring the initial urethral pressure data of the target patient and using the classification model to confirm the initial start time based on the original medical data set and the initial urethral pressure data of the target patient, a pressure detection start instruction is sent to enable the pressure sensor to monitor the pressure inside the catheter balloon in real time at a certain monitoring frequency; The pressure sensor's monitoring frequency at different stages of care for target patients is dynamically adjusted using a time series model based on the catheter balloon's internal pressure.

[0045] Another embodiment of the present application provides a device for monitoring the internal pressure of a urinary catheter balloon, comprising: The pressure detection sensor module is used to convert the pressure signal in the urinary catheter balloon detected by the pressure detection sensor into an electrical signal pressure sensor; in this embodiment, the pressure detection sensor module is used to convert the pressure signal in the urinary catheter balloon into an electrical signal through a Wheatstone bridge circuit; the measurement accuracy of the pressure sensor is ≤150Pa; The main control chip module is used to obtain pressure data from the pressure detection sensor module to complete the above-mentioned AI-based wearable catheter pressure monitoring method; A power supply module with a built-in micro button battery for providing power support; The wireless communication module is used for data transmission with the associated cloud and mobile terminals. In this embodiment, the wireless communication module adopts a Bluetooth module based on the BLE 5.1 ​​protocol.

[0046] In particular, in this embodiment, the user's mobile terminal (such as a mobile phone APP) can display the pressure data and warning information inside the catheter balloon in real time, and remind the user that the pressure monitoring device inside the catheter balloon is low on power and needs to be replaced. In summary, the embodiments of the present application provide an AI-based wearable catheter pressure monitoring system that can intelligently adjust the monitoring frequency according to the dynamic changes in the patient's urethral pressure, while predicting the device life, achieving low-power operation and improving monitoring efficiency.

[0047] Example 2 On one hand, an embodiment of the present application provides an AI-based wearable urinary catheter pressure monitoring system, wherein the method of this embodiment is substantially the same as that of the first embodiment, except that the method further comprises: The time series model is deployed on the main control chip module of the catheter balloon internal pressure monitoring device. It can transmit data between the time series model and the main control chip, realize local real-time reasoning, and ensure rapid response. The AI-based wearable catheter pressure monitoring system of this embodiment also includes a cloud, which can receive the raw pressure data inside the catheter balloon monitored in real time for incremental training and strategy iterative optimization of the classification model and time series model, thereby continuously improving the model performance and improving the accuracy of the AI-based wearable catheter pressure monitoring system.

[0048] It should be pointed out that, according to the needs of implementation, the various steps described in this application can be split into more steps, or two or more steps or partial operations of the steps can be combined into new steps to achieve the purpose of the present invention.

[0049] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A wearable urinary catheter pressure monitoring system based on AI, characterized in that: include: An initial startup module obtains initial urethral pressure data of a target patient, and uses a classification model to determine an initial startup time based on the original medical data set and the initial urethral pressure data of the target patient, and then enters a dormant state; After the initial start time is reached, a pressure detection start instruction is sent to enable the pressure sensor to monitor the internal pressure of the catheter balloon in real time at a certain monitoring frequency; The dynamic monitoring module is used to receive and dynamically adjust the monitoring frequency of the pressure sensor at different nursing stages of the target patient based on the internal pressure of the catheter balloon using a time series model.

2. The AI-based wearable urinary catheter pressure monitoring system according to claim 1 is characterized in that: The AI-based wearable urinary catheter pressure monitoring system also includes: A battery life prediction module is used to use the pressure data in the catheter balloon to establish a linear regression equation to predict the remaining battery life of the AI-based wearable catheter pressure monitoring system.

3. The AI-based wearable urinary catheter pressure monitoring system according to claim 1, characterized in that: The AI-based wearable catheter pressure monitoring system also includes a cloud, which can receive the raw pressure data in the catheter balloon monitored in real time for incremental training and strategy iterative optimization of the classification model and the time series model.

4. The AI-based wearable urinary catheter pressure monitoring system according to claim 1, characterized in that: The method of using the classification model to confirm the initial start time and send a pressure detection start instruction so that the pressure sensor monitors the internal pressure of the catheter balloon in real time at a certain monitoring frequency is as follows: Extracting multidimensional features from the medical big data of the original medical data set as initial feature vectors, and using the time of the first occurrence of catheter abnormality in the original medical data set as label data, to construct a training set through feature engineering; wherein the multidimensional features include the patient's initial urethral pressure value, the patient's age, the patient's gender, the catheter model, and the patient's previous complication record; The classification model is used to discretize the time of the first occurrence of the urinary catheter abnormality into multiple time windows, and the classification model is trained using the training set to output a probability distribution of each time window, wherein the input of the classification model is the initial feature vector; The initial urethral pressure data of the target patient is collected in real time, and the trained classification model is inputted in combination with the multi-dimensional features of the target patient to obtain the probability distribution of each time window, and the time window with the highest probability is selected as the initial start time. A pressure detection start instruction is sent so that the pressure sensor monitors the internal pressure of the catheter balloon in real time at a certain monitoring frequency.

5. The AI-based wearable urinary catheter pressure monitoring system according to claim 1, characterized in that: The method for dynamically adjusting the monitoring frequency of the pressure sensor at different nursing stages of the target patient using the time series model is as follows: Collecting time series data of pressure in the catheter balloon in a sliding window manner, and calculating the pressure mean, standard deviation, and pressure change rate; defining stage labels of different nursing stages based on the original medical data set, wherein the different nursing stages include an initial stage, a stable stage, and a critical stage; Constructing and training a time series model, inputting the pressure time series data in the catheter balloon within the sliding window, the pressure mean, the pressure change rate, and the standard deviation into the trained time series model, and outputting the probability distribution of the sliding window being in the initial period, the stable period, and the critical period; The monitoring frequency of the pressure sensor is dynamically adjusted according to the output probability distribution results: the monitoring frequency is increased in the initial stage, the monitoring frequency is reduced in the stable period, and the monitoring frequency is significantly increased in the dangerous period and an early warning is triggered.

6. The AI-based wearable urinary catheter pressure monitoring system according to claim 5, characterized in that: The initial stage is defined as the pressure fluctuation rate not exceeding 5% or the pressure value being consistently below the clinical safety threshold; The stable period is defined as: the pressure fluctuation rate is between 5% and 10% or the duration of the pressure value exceeding the clinical safety threshold does not exceed the first time; The critical period is defined as the pressure fluctuation rate exceeding 10% or the duration of the pressure value exceeding the clinical safety threshold exceeding the first time.

7. The AI-based wearable urinary catheter pressure monitoring system according to claim 2, characterized in that: The method for using the pressure data in the catheter balloon to establish a linear regression equation to predict the remaining life of the AI-based wearable catheter pressure monitoring system is: The average power consumption of the AI-based wearable catheter pressure monitoring system is obtained by using the pressure data test inside the catheter balloon in the early stage. Combined with the current remaining power of the AI-based wearable catheter pressure monitoring system, a linear regression equation is established to predict the remaining battery life of the AI-based wearable catheter pressure monitoring system.

8. A wearable urinary catheter pressure monitoring method based on AI, characterized in that: include: obtaining initial urethral pressure data of a target patient, and confirming an initial start time using a classification model based on the original medical data set and the initial urethral pressure data of the target patient, and then entering a dormant state; After the initial start time is reached, a pressure detection start instruction is sent to enable the pressure sensor to monitor the internal pressure of the catheter balloon in real time at a certain monitoring frequency; The pressure sensor's monitoring frequency at different stages of care for target patients is dynamically adjusted using a time series model based on the catheter balloon's internal pressure.

9. A catheter balloon internal pressure monitoring device, characterized in that: include: A pressure detection sensor module is used to convert the pressure signal in the catheter balloon detected by the pressure detection sensor into an electrical signal pressure sensor; A main control chip module, used to obtain pressure data from the pressure detection sensor module to complete the AI-based wearable catheter pressure monitoring method of claim 8; The power supply module is used to provide power support.

10. The device for monitoring the internal pressure of a urinary catheter balloon according to claim 9, characterized in that: The catheter balloon internal pressure monitoring device further includes: a wireless communication module for transmitting data with an associated cloud or mobile terminal; The time series model is deployed on the main control chip module; The power supply module has a built-in micro button battery.

Images