Peritoneal drainage flushing device based on multi-dimensional parameter monitoring and negative pressure regulation and control system

Abstract

The invention discloses a peritoneal drainage flushing device based on multi-dimensional parameter monitoring and a negative pressure regulation and control system, and belongs to the technical field of peritoneal drainage. The peritoneal drainage flushing device and negative pressure regulation and control system based on multi-dimensional parameter monitoring comprises a device support, a multi-channel rotary valve and two negative pressure pumps, a flow guide pipe and a backflow pipe are arranged between the two negative pressure pumps and the multi-channel rotary valve, and pipeline detection modules are installed on the flow guide pipe and the backflow pipe. The peritoneal drainage device solves the problems that an existing peritoneal drainage device is not ideal in treatment effect and affects the treatment effect when a drainage tube is blocked, early intervention is carried out by adjusting a negative pressure pump, a multi-channel rotary valve and a three-way electromagnetic valve, targeted treatment is automatically triggered at the early stage of complications through continuous and multi-dimensional monitoring, and the treatment effect is effectively improved. The tube blocking rate, the abdominal infection rate and the abdominal hypertension syndrome occurrence rate are remarkably reduced, the pain of a patient is relieved, the recovery time is shortened, and personalized and self-adaptive adjustment is carried out according to the physical state of the patient.

Description

Technical Field

[0001] This invention relates to the field of abdominal drainage technology, specifically to an abdominal drainage and irrigation device and a negative pressure control system based on multidimensional parameter monitoring. Background Technology

[0002] A drainage device is a medical device used in clinical surgery to drain pus, blood, and fluid accumulated between human tissues or in body cavities to the outside of the body, preventing postoperative infection and promoting wound healing.

[0003] Chinese Patent CN120771372A discloses an abdominal drainage device for tumor surgery, including a negative pressure drainage device. The beneficial effects of this patent are as follows: by designing a rigid adjustment branch tube below the main abdominal drainage tube, and cooperating with the design of an adjustment fixing ring and adjustment fixing components, the movable pushing pad makes the movable adjustment tube rotate as a whole. Pressing and pushing the pad pulls the tension line, causing the gate to squeeze the adjustment protrusion, thereby causing the hollow tube above the rigid adjustment branch tube to shift and the hollow tube below to deflect. The deflection of the rigid adjustment branch tube below drives the adjustment of the position of the rigid connecting branch tube and the abdominal drainage branch tube, making the drainage position easy to adjust.

[0004] In actual use, the abdominal drainage device described in the above patent has an unsatisfactory effect on the treatment of drainage tube blockage, which affects the treatment effect; therefore, it does not meet the existing needs. In response, we have proposed an abdominal drainage and irrigation device and negative pressure control system based on multi-dimensional parameter monitoring. Summary of the Invention

[0005] The purpose of this invention is to provide an abdominal drainage and irrigation device and a negative pressure control system based on multi-dimensional parameter monitoring, which solves the problem in the above-mentioned background art that the treatment effect is not ideal when the drainage tube is blocked during actual use, thus affecting the treatment effect.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an abdominal drainage and irrigation device based on multi-dimensional parameter monitoring, comprising a device support, a multi-channel rotary valve and two negative pressure pumps, wherein the two negative pressure pumps are installed at the rear end of the device support, the multi-channel rotary valve is located at the front end of the device support, and a three-way solenoid valve is installed between the multi-channel rotary valve and the two negative pressure pumps. A guide pipe and a return pipe are provided between the two negative pressure pumps and the multi-channel rotary valve. A pipeline detection module is installed on the guide pipe and the return pipe. The pipeline detection module includes a negative pressure sensor, a temperature sensor, a volume sensor, a biochemical marker detection module, and a fiber optic imaging module.

[0007] Preferably, a protective cover is fixedly installed on the device support, and a mounting bracket is welded to the middle of the device support. The mounting bracket is connected to the multi-channel rotary valve by fixing screws.

[0008] Preferably, the device bracket is equipped with an electrical box and an alarm module. The electrical box is equipped with a wireless transmission module and a processor module, and the processor module is electrically connected to the alarm module and the pipeline detection module, respectively.

[0009] Preferably, the bottom of the device support is equipped with a flushing fluid tank, the upper end of the flushing fluid tank is equipped with a drainage collection box, the drainage collection box is equipped with a sterilization module, and the bottom of the device support is equipped with casters.

[0010] Preferably, the flushing fluid tank is connected to the negative pressure pump via a pipeline, a drainage pipe is provided between the drainage collection tank and the multi-channel rotary valve, the drainage pipe adopts an easy-to-disassemble structure, the front end of the multi-channel rotary valve is provided with a quick-release head for the drainage pipe, and the front end of the quick-release head for the drainage pipe is provided with an abdominal drainage pipe connector.

[0011] Preferably, a pulse generator is installed on one side of the multi-pass rotary valve, and a pulse gas guide tube is provided between the pulse generator and the multi-pass rotary valve. The pulse gas guide tube guides the pulsed gas emitted by the pulse generator into the multi-pass rotary valve to impact the blockage.

[0012] Preferably, a hydraulic high-precision rotary disc is installed on one side of the multi-pass rotary valve, a hydraulic pump is installed at one end of the hydraulic high-precision rotary disc, and the hydraulic pump is connected to two swing cylinders. The two swing cylinders are connected to the hydraulic high-precision rotary disc through gear transmission, and the hydraulic high-precision rotary disc is fixedly connected to the valve body of the multi-pass rotary valve through a flange.

[0013] Preferably, the three-way solenoid valve is sealed to the guide pipe and the return pipe, and a sterile electrical connector is installed between the guide pipe and the return pipe and the pipeline detection module, and the sterile electrical connector is electrically connected to the electrical box.

[0014] Preferably, the electrical box is equipped with a power supply, and an operation panel and a display screen are mounted on the outside of the device bracket. The operation panel and the display screen are connected to the processor module inside the electrical box, and the display screen is used to display the monitoring data of the pipeline detection module, while the operation panel is used by medical personnel to control the equipment.

[0015] The negative pressure control system for an abdominal drainage and irrigation device based on multidimensional parameter monitoring includes: The bedside equipment layer is used to perform real-time monitoring and rapid response to the abdominal drainage medical care process. It is implemented based on the display tubing detection module and electrical box. The equipment layer, used for integrated intelligent drainage, employs a multi-channel rotary valve and two negative pressure pumps for abdominal drainage. The intelligent management layer, used for intelligent monitoring and adaptive control, is based on the data collected by the pipeline detection module and is intelligently adjusted through the processor module. The intelligent management system includes a conventional and stable drainage unit, an early warning and automatic intervention unit, and a manual intervention unit. The conventional and stable drainage unit maintains stable negative pressure drainage based on the monitoring parameters of the pipeline detection module. Based on the monitoring parameters of the pipeline detection module, the early warning and automatic intervention unit automatically triggers the following cascaded responses: Primary blockage warning: decreased drainage flow and fluctuating tube end pressure, with the negative pressure pump briefly increasing negative pressure; Intermediate blockage alarm: Primary unblocking is ineffective, differential pressure continues to rise, triggering the pulse unblocking program, starting the pulse generator, rapidly switching the multi-channel rotary valve, and performing pulse flushing; High-risk infection warning: When the turbidity of the drainage fluid continues to rise and the pH value decreases, the multi-channel rotary valve for therapeutic irrigation is automatically activated, the negative pressure pump draws the irrigation fluid from the irrigation fluid tank, the pulse generator starts a high-frequency low-amplitude sine wave pulse, and the negative pressure pump draws the drainage tube. Emergency safety alarm: If the intra-abdominal pressure exceeds the safety threshold, the system will automatically stop immediately, the multi-channel rotary valve will switch to closed, all negative pressure pumps will stop, and an alarm will be triggered via the alarm module. Active bleeding warning: When the drainage fluid color suddenly turns bright red and the flow rate per unit time increases sharply, the system will automatically switch to safe mode, reduce the negative pressure pump to a very low negative pressure to maintain drainage, avoid aggravating bleeding, and turn off any automatic flushing or unblocking operations.

[0016] Preferably, the intelligent management layer is configured as follows: Multiple frames of fiber optic images are acquired through the fiber optic imaging module in the pipeline detection module; A convolutional neural network is used to segment the fiber optic image and extract the morphological feature parameters of the blockage. The morphological feature parameters of the blockage include: the proportion of the blockage coverage area, the edge sharpness index of the blockage, the texture feature vector of the blockage, and the color distribution histogram of the blockage. Time-series analysis based on multi-frame fiber optic images was performed to calculate the motion characteristics of the blockage, including the deformation rate of the blockage as a function of negative pressure and the adhesion strength index between the blockage and the pipe wall. Pressure waveform data is collected using the negative pressure sensor in the pipeline detection module; Based on the pressure waveform data, extract the time-frequency domain features of pressure fluctuations and construct a pressure-flow relationship model; Intelligent classification of blockage conditions is performed based on blockage morphology parameters, time-frequency domain characteristics of pressure fluctuations, and pressure-flow relationship models to determine the blockage situation. The optimal pulse parameters are automatically matched based on the blockage situation; The pulse generator monitors changes in intra-abdominal pressure in real time during pulse execution based on optimal pulse parameters. Pulse intervention is performed based on changes in intra-abdominal pressure; pulse intervention is a safety mechanism for dynamically adjusting or terminating the operation during pulse execution. Based on the optimal pulse parameters and pulse intervention results, a reinforcement learning algorithm is used to optimize the subsequent optimal pulse parameter matching strategy.

[0017] Preferably, based on the optimal pulse parameters and the pulse intervention results, a reinforcement learning algorithm is used to optimize the subsequent optimal pulse parameter matching strategy, including: When pulse intervention is triggered, pressure is released through the micro pressure buffer chamber built into the pulse air tube, and the pulse parameters are adjusted according to the safety offset. Calculate the pulse effect score based on the clearing results and the maximum safe offset; Construct a reinforcement learning optimization framework, define the state space, action space, and reward function, and update the Q-value table; Based on the updated Q-value table, the weight coefficients of the impulse parameter determination model are adjusted; The optimal pulse parameter matching strategy is determined by determining the weight coefficients of the model based on the adjusted pulse parameters.

[0018] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention provides proactive early warning through the pipeline detection module and early intervention by adjusting the negative pressure pump, multi-channel rotary valve, and three-way solenoid valve. Through continuous and multi-dimensional monitoring, it automatically triggers targeted treatment in the early stages of complications, significantly reducing the rates of tube blockage, abdominal infection, and intra-abdominal hypertension syndrome. This alleviates patient suffering, shortens recovery time, and allows for personalized and adaptive adjustments based on the patient's physical condition. The system automatically executes and adjusts flushing and unblocking strategies based on real-time data on the characteristics, flow rate, and pressure of each patient's drainage fluid, achieving adaptive and precise control. This improves the consistency of treatment outcomes and effectively solves the problem in existing technologies where complications such as blockage, infection, and intra-abdominal hypertension rely on nurses' regular rounds for detection.

[0019] 2. This invention utilizes the sensor components of the pipeline detection module for comprehensive detection, combined with the processor module to control the equipment for dynamic and precise pressure control. It monitors and adjusts negative pressure in real time to maintain a constant and safe drainage environment. A pulse generator, combined with a negative pressure pump, extracts flushing fluid for cleaning, which is more physiologically sound, with controllable impact force, better patient tolerance, and a more stable and comfortable treatment experience. This minimizes secondary damage or discomfort caused by improper operation and avoids the need for manual adjustment of negative pressure, which can lead to pressure fluctuations due to changes in body position or blockage of drainage material, causing pain or tissue damage. The invention automatically completes the mechanical work of monitoring, flushing, and unblocking, reducing manual labor. Based on the actual working status of the abdominal drainage system, it can generate objective quantitative data, providing real-time curves, trend analysis, and multi-parameter fusion-based early warning reports, offering strong data support for subsequent medical care. Attached Figure Description

[0020] Figure 1 This is a top-view axonometric view of the present invention; Figure 2 This is an isometric view of the front view of the present invention; Figure 3 This is an isometric view of the side view of the present invention; Figure 4 For the present invention Figure 3 Enlarged view of a portion of area A in the middle; Figure 5 This is a system schematic diagram of the present invention.

[0021] In the diagram: 1. Device support; 101. Protective cover; 102. Alarm module; 2. Mounting bracket; 3. Electrical box; 4. Fluid tank; 401. Drainage collection box; 5. Multi-channel rotary valve; 501. Pulse generator; 502. Quick-release head for drainage tube; 503. Hydraulic high-precision rotary disc; 504. Abdominal drainage tube connector; 505. Three-way solenoid valve; 506. Guide tube; 507. Pulse air guide tube; 508. Return tube; 509. Drainage discharge tube; 6. Pipeline detection module; 7. Negative pressure pump. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] To address the issue of drainage tube blockage in existing abdominal drainage devices during practical use, resulting in unsatisfactory treatment outcomes and negatively impacting therapeutic results, please refer to... Figure 1 - Figure 4This embodiment provides the following technical solution: In this embodiment, the abdominal drainage and irrigation device based on multi-dimensional parameter monitoring includes a device support 1, a multi-channel rotary valve 5 and two negative pressure pumps 7. The two negative pressure pumps 7 are installed at the rear end of the device support 1, the multi-channel rotary valve 5 is located at the front end of the device support 1, and a three-way solenoid valve 505 is installed between the multi-channel rotary valve 5 and the two negative pressure pumps 7. A guide tube 506 and a return tube 508 are provided between the two negative pressure pumps 7 and the multi-channel rotary valve 5. A pipeline detection module 6 is installed on the guide tube 506 and the return tube 508. The pipeline detection module 6 includes a negative pressure sensor, a temperature sensor, a volume sensor, a biochemical marker detection module, and a fiber optic imaging module. It monitors the temperature of the drainage fluid and provides early warning of infection or bleeding. The biochemical marker detection module is used to more accurately determine complications such as infection and anastomotic leakage. The fiber optic imaging module is used to directly observe the condition of the front end of the drainage tube or to perform image analysis on the characteristics of the drainage fluid.

[0024] It should be noted that the three-way solenoid valve 505 is sealed to the guide pipe 506 and the return pipe 508, and a sterile electrical connector is installed between the guide pipe 506 and the return pipe 508 and the pipeline detection module 6. The sterile electrical connector is electrically connected to the electrical box 3. By operating the three-way solenoid valve 505, the passage of the guide pipe 506 and the return pipe 508 is changed, and the negative pressure drainage route is controlled in conjunction with the multi-pass rotary valve 5.

[0025] The device bracket 1 houses an electrical box 3 and an alarm module 102. The electrical box 3 contains a wireless transmission module and a processor module, and the processor module is electrically connected to the alarm module 102 and the pipeline detection module 6, respectively. The electrical box 3 is equipped with a power supply. An operation panel and a display screen are mounted on the outside of the device bracket 1. The operation panel and display screen are connected to the processor module inside the electrical box 3. The display screen is used to display the monitoring data of the pipeline detection module 6, and the operation panel is used by medical personnel to control the equipment. The display screen displays real-time curves, parameters, and system status, and provides simple operation buttons. The alarm module 102 is composed of an alarm light and a buzzer.

[0026] In addition, a pulse generator 501 is installed on one side of the multi-pass rotary valve 5. A pulse air guide pipe 507 is provided between the pulse generator 501 and the multi-pass rotary valve 5. The pulse air guide pipe 507 introduces the pulse gas emitted by the pulse generator 501 into the multi-pass rotary valve 5 to impact the blockage.

[0027] Specifically, the sensor components of the pipeline detection module 6 perform thorough detection, and the processor module controls the equipment to dynamically and precisely control the pressure, monitor and adjust the negative pressure in real time to maintain a constant and safe drainage environment. The pulse generator 501, together with the negative pressure pump 7, draws out the flushing fluid for cleaning, which is more physiological, the impact force is controllable, and the patient's tolerance is better, providing a more stable and comfortable treatment experience. It minimizes secondary damage or discomfort caused by improper operation, avoids manual adjustment of negative pressure, and avoids pressure fluctuations caused by changes in body position or blockage of drainage material, which may cause pain or tissue damage. It automatically completes the mechanical labor of monitoring, flushing, and unblocking, reducing manual labor. It can generate objective quantitative data based on the actual working status of the abdominal drainage system, and provide real-time curves, trend analysis, and multi-parameter fusion early warning reports, providing strong data support for subsequent medical care.

[0028] To address the issue of drainage tube blockage in existing abdominal drainage devices during practical use, resulting in unsatisfactory treatment outcomes and negatively impacting therapeutic results, please refer to... Figure 1 - Figure 4 This embodiment provides the following technical solution: In this embodiment, a protective cover 101 is fixedly installed on the device support 1, and a mounting bracket 2 is welded to the middle of the device support 1. The mounting bracket 2 is connected to the multi-channel rotary valve 5 by fixing screws.

[0029] The bottom of the device support 1 is equipped with a flushing fluid tank 4, and the top of the flushing fluid tank 4 is equipped with a drainage collection box 401. The drainage collection box 401 is equipped with a sterilization module. The bottom of the device support 1 is equipped with casters. Moving the device support 1 will move the entire device, which is convenient for operation.

[0030] It should be noted that the flushing fluid tank 4 is connected to the negative pressure pump 7 through a pipeline. A drainage and discharge pipe 509 is provided between the drainage collection box 401 and the multi-channel rotary valve 5. The drainage and discharge pipe 509 adopts an easy-to-disassemble structure. A quick-release head 502 for the drainage tube is provided at the front end of the multi-channel rotary valve 5, and an abdominal drainage tube connector 504 is provided at the front end of the quick-release head 502 to temporarily close the drainage end of the pipeline. The flushing end is connected to a pre-filled sterile flushing fluid bag.

[0031] Furthermore, a high-precision hydraulic rotary disc 503 is installed on one side of the multi-way rotary valve 5. A hydraulic pump is installed at one end of the high-precision hydraulic rotary disc 503, and the hydraulic pump is connected to two swing cylinders. The two swing cylinders are connected to the high-precision hydraulic rotary disc 503 through gear transmission. The high-precision hydraulic rotary disc 503 is fixedly connected to the valve body of the multi-way rotary valve 5 through a flange. The position angle of the two swing cylinders is adjusted by controlling the hydraulic pump, and the adjustment effect is precise, which is more accurate to meet the working requirements of the multi-way rotary valve 5.

[0032] Specifically, the system proactively issues warnings through the pipeline detection module 6 and intervenes early by adjusting the negative pressure pump 7, multi-channel rotary valve 5, and three-way solenoid valve 505. Through continuous and multi-dimensional monitoring, targeted treatment is automatically triggered in the early stages of complications, significantly reducing the rates of tube blockage, abdominal infection, and intra-abdominal hypertension syndrome. This alleviates patient suffering, shortens recovery time, and allows for personalized and adaptive adjustments based on the patient's individual physical condition. The system automatically executes and adjusts flushing and unblocking strategies based on real-time data on the characteristics, flow rate, and pressure of each patient's drainage fluid, achieving adaptive and precise control. This improves the consistency of treatment outcomes and effectively solves the problem of existing technologies where complications such as blockage, infection, and intra-abdominal hypertension rely on nurses' regular rounds for detection.

[0033] Please see Figure 5 A negative pressure control system for an abdominal drainage and irrigation device based on multidimensional parameter monitoring includes: The bedside equipment layer, used for real-time monitoring and rapid response of the abdominal drainage medical care process, is implemented based on the display tubing detection module 6 and the electrical box 3; The equipment layer, used for integrated intelligent drainage, employs a multi-channel rotary valve 5 and two negative pressure pumps 7 for abdominal drainage. The intelligent management layer, which is used for intelligent monitoring and adaptive control, is based on the data collected by the pipeline detection module 6 and is intelligently adjusted through the processor module. The intelligent management system includes a regular stable drainage unit, an early warning and automatic intervention unit, and a manual intervention unit. The regular stable drainage unit maintains stable negative pressure drainage based on the monitoring parameters of the pipeline detection module 6. Based on the monitoring parameters of pipeline detection module 6, the early warning and automatic intervention unit automatically triggers the following cascaded responses: Primary blockage warning: decreased drainage flow and fluctuating tube end pressure; negative pressure pump 7 briefly increases negative pressure. Intermediate blockage alarm: Primary unblocking is ineffective, differential pressure continues to rise, triggering the pulse unblocking program, starting the pulse generator 501, and quickly switching the multi-channel rotary valve 5 to perform pulse flushing; High risk of infection warning: The turbidity of the drainage fluid continues to rise and the pH value decreases. The therapeutic irrigation multi-channel rotary valve 5 is automatically activated to switch, the negative pressure pump 7 draws the irrigation fluid in the irrigation fluid tank 4, the pulse generator 501 starts a high frequency low amplitude sine wave pulse, and the negative pressure pump 7 draws the drainage tube. Emergency safety alarm: When the intra-abdominal pressure exceeds the safety threshold, the system will automatically stop immediately, the multi-channel rotary valve 5 will switch to closed, all negative pressure pumps 7 will stop, and the alarm module 102 will activate the alarm function. Active bleeding warning: When the drainage fluid color suddenly turns bright red and the flow rate per unit time increases sharply, the system will automatically switch to safe mode. The negative pressure pump 7 will be reduced to an extremely low negative pressure to maintain drainage and prevent further bleeding. Any automatic flushing or unblocking operations will be turned off.

[0034] Working Principle: During use, move the device support 1 to the working position, confirm that the power and network connections are normal, align the sterile electrical connector of the tubing with the corresponding interface of the tubing detection module 6, completing both physical and electrical connections simultaneously. Temporarily close the drainage end of the tubing, connect the flushing end to the drainage collection box 401, perform sensor calibration, valve repositioning, and pump function testing. Set initial parameters according to the surgical type and patient's physical characteristics. After connecting the drainage tube to the abdominal drainage tube connector 504, place the intubated end in the predetermined position in the abdominal cavity, rotate it to the drainage channel via the multi-channel rotary valve 5, and the negative pressure pump 7 starts smoothly with initial negative pressure. Detection is performed through the tubing detection module 6, establishing baseline data and real-time data. The curve is displayed on the screen. The system continuously collects physical, biochemical, and image data of the drainage fluid and calculates the intra-abdominal pressure. When the monitored data deviates from the normal range, the system automatically triggers the device response. The intelligent management layer automatically adjusts the device and intervenes early by adjusting the negative pressure pump 7, the multi-channel rotary valve 5, and the three-way solenoid valve 505. The processor module controls the device to dynamically and accurately control the pressure, monitors and adjusts the negative pressure in real time, and maintains a constant and safe drainage environment. The pulse generator 501, together with the negative pressure pump 7, draws out the flushing fluid for cleaning, which is more physiological, the impact force is controllable, and the patient's tolerance is better. It provides a more stable and comfortable treatment experience and minimizes secondary damage or discomfort caused by improper operation.

[0035] In this embodiment, the intelligent management layer of the negative pressure control system of the abdominal drainage and irrigation device based on multi-dimensional parameter monitoring is configured as follows: Multiple frames of fiber optic images are acquired through the fiber optic imaging module in pipeline detection module 6; In this embodiment, the multi-frame fiber optic images are a sequence of images of the inside of the drainage tube acquired at different time points; A convolutional neural network is used to segment the fiber optic image and extract the morphological feature parameters of the blockage. The morphological feature parameters of the blockage include: the proportion of the blockage coverage area, the edge sharpness index of the blockage, the texture feature vector of the blockage, and the color distribution histogram of the blockage. In this embodiment, the convolutional neural network is a deep learning algorithm architecture used for image recognition and feature extraction. Here, it is used to perform pixel-level segmentation processing on the fiber optic image, automatically identify and mark the blockage area, and eliminate interference from the tube wall and normal fluid. The blockage coverage area ratio (the percentage of the area occupied by the blockage in the cross-section of the drainage tube, ranging from 0-100%); the blockage edge sharpness index (a value characterizing the distinction between the blockage edge and the tube wall, ranging from 0-1, with a value closer to 1 indicating a sharper edge); the blockage texture feature vector (a mathematical vector extracted through the gray-level co-occurrence matrix, quantifying the roughness, uniformity, and other texture information of the blockage surface); and the blockage color distribution histogram (statistical analysis of the blockage color distribution to distinguish different types, such as blood clots exhibiting a red distribution characteristic, and fibrin clots exhibiting a white / transparent distribution characteristic).

[0036] Time-series analysis based on multi-frame fiber optic images was performed to calculate the motion characteristics of the blockage, including the deformation rate of the blockage as a function of negative pressure and the adhesion strength index between the blockage and the pipe wall. In this embodiment, the deformation rate of the blockage as a function of negative pressure is calculated as deformation amount / original size, which reflects the elasticity or hardness of the blockage. Soft blockages have a high deformation rate, while hard blockages have a low deformation rate. The adhesion strength index between the blockage and the pipe wall is calculated by measuring the displacement of the blockage after applying a small test pulse. The larger the value, the stronger the adhesion, and the more difficult it is to remove by simple changes in negative pressure. Pressure waveform data is collected using the negative pressure sensor in pipeline detection module 6; In this embodiment, the pressure waveform data is continuous pressure time-series data collected by the negative pressure sensor; Based on the pressure waveform data, extract the time-frequency domain features of pressure fluctuations and construct a pressure-flow relationship model; In this embodiment, the time-frequency domain characteristics of pressure fluctuations are features extracted from the pressure waveform through time-frequency analysis (such as fast Fourier transform), such as: differential pressure amplitude, standard deviation of pressure fluctuations, pressure harmonic spectrum distribution, and pressure response delay time; the pressure-flow relationship model is a mathematical model describing the relationship between pressure and flow rate, the core of which is the calculation of the flow resistance coefficient: flow resistance coefficient = (proximal pressure - distal pressure) / instantaneous flow rate. By analyzing the rate of change of the flow resistance coefficient, progressive blockage and sudden blockage can be distinguished. Intelligent classification of blockage conditions is performed based on blockage morphology parameters, time-frequency domain characteristics of pressure fluctuations, and pressure-flow relationship models to determine the blockage situation. In this embodiment, the blockage situations include: Type A: Loose partial blockage (coverage area <40%, blurred edges, and increased flow resistance coefficient <30%). Type B: Adhesive partial clogging (coverage area 40-70%, clear edges, flow resistance coefficient increased by 30-60%). Type C: Completely dense blockage (coverage area >70%, pressure fluctuations disappear, flow resistance coefficient increases >60%). Type D: Elastic biofilm blockage (fluctuating coverage area, with periodic pressure fluctuations); When determining the blockage situation, the blockage morphology parameters, pressure fluctuation time-frequency domain features, and pressure-flow relationship model multidimensional feature parameter set are input into a pre-trained blockage situation classifier, which outputs the blockage situation category. The optimal pulse parameters are automatically matched based on the blockage situation; In this embodiment, when automatically matching the optimal pulse parameters according to the blockage situation, the optimal pulse parameters are obtained by inputting the blockage morphology parameters, pressure fluctuation time-frequency domain characteristics, and flow resistance coefficient change rate into the pulse parameter determination model corresponding to the blockage situation. In this embodiment, the pulse parameter determination model corresponding to the blockage situation is trained and obtained based on the historical pulse control records of the corresponding blockage situation. The historical pulse control records are marked with historical blockage morphological feature parameters, historical pressure fluctuation time-frequency domain features, historical flow resistance coefficient change rate, and their corresponding historical pulse parameters. During training, the historical blockage morphological feature parameters, historical pressure fluctuation time-frequency domain features, and historical flow resistance coefficient change rate are used as input to the preset CNN model, and the corresponding historical pulse parameters are used as the model output. The pulse intensity weight coefficient, pulse frequency weight coefficient, pulse duration weight coefficient, and pulse waveform selection weight coefficient of the model are iterated based on the minimum gradient optimization method until the model reaches the preset convergence criterion. The pulse generator 501 monitors changes in intra-abdominal pressure in real time during pulse execution according to the optimal pulse parameters; In this embodiment, intra-abdominal pressure change refers to the real-time fluctuation of intra-abdominal pressure when the pulse generator 501 executes pulses according to the optimal pulse parameters.

[0037] Pulse intervention is performed based on changes in intra-abdominal pressure; pulse intervention is a safety mechanism for dynamically adjusting or terminating the operation during pulse execution. In this embodiment, pulse intervention refers to a safety mechanism that dynamically adjusts or stops the operation during pulse execution. For example, when the instantaneous pressure is detected to exceed the set pressure threshold, the pressure is automatically released through the micro pressure buffer chamber set in the pulse air guide tube 507. Based on the optimal pulse parameters and pulse intervention results, the subsequent optimal pulse parameter matching strategy is optimized through reinforcement learning algorithm; In this embodiment, when optimizing the subsequent optimal pulse parameter matching strategy, the system learns through trial and error: the pulse parameters are used as actions, and the intervention results (such as the success rate of blockage clearance and intra-abdominal pressure safety) are used as rewards to iteratively optimize the strategy; the subsequent optimal pulse parameter matching strategy is the optimized decision rule, which is used for future blockage events. For example, if a certain type of blockage often triggers pressure relief, the strategy will automatically reduce the initial pulse intensity.

[0038] This invention first utilizes the fiber optic imaging module of the pipeline detection module 6 to continuously acquire multi-frame image sequences of the drainage tube's interior. Pixel-level segmentation is then performed using a convolutional neural network to extract morphological parameters such as the proportion of the blockage area, edge sharpness index, texture feature vector, and color distribution histogram. Simultaneously, based on time-series image analysis, the deformation rate of the blockage under negative pressure and the tube wall adhesion strength index are calculated. Combined with pressure waveform data acquired by a negative pressure sensor, time-frequency domain features (such as differential pressure amplitude and harmonic spectrum) are extracted, and a pressure-flow relationship model (with dynamic calculation of the flow resistance coefficient as its core) is constructed. This model then integrates multi-source parameters to intelligently classify blockage conditions (such as loose partial blockage, adhesive partial blockage, dense complete blockage, or elastic biofilm blockage). Subsequently, the optimal pulse parameters are automatically matched based on a historical training model, and the pulse generator 501 executes the pulse operation. During the process, changes in intra-abdominal pressure are monitored in real time. If the instantaneous pressure exceeds a safety threshold, automatic pressure relief is initiated through the micro-pressure buffer chamber built into the pulse airway 507 for safety intervention. Finally, by combining the pulse parameter execution results and intervention feedback, a reinforcement learning algorithm is used to iteratively optimize subsequent pulse strategies, forming an adaptive closed loop. This invention achieves precise blockage treatment through multi-dimensional parameter collaborative monitoring and AI-driven closed-loop optimization; multi-dimensional feature fusion enables millisecond-level classification of blockage types, reducing the risk of misjudgment; real-time abdominal pressure monitoring and automatic pressure relief mechanisms effectively prevent iatrogenic injury; reinforcement learning-driven parameter optimization continuously improves clearance efficiency and reduces energy consumption. Intelligent control significantly reduces the operational burden on medical staff and improves the stability of the abdominal drainage process.

[0039] In one embodiment, based on the optimal pulse parameters and the pulse intervention result, a reinforcement learning algorithm is used to optimize the subsequent optimal pulse parameter matching strategy, including: When pulse intervention is triggered, pressure is released through the miniature pressure buffer chamber built into the pulse air delivery tube 507, and simultaneously, based on the safety offset... Adjust pulse parameters: ; ; ; ; in, , For the set pressure threshold, Intra-abdominal pressure over time The function of change; , , and These are the pulse intensity, pulse frequency, pulse duration, and pulse waveform type before adjustment (pulse parameters before adjustment). , , and These are the adjusted pulse intensity, pulse frequency, pulse duration, and pulse waveform type (adjusted pulse parameters). Calculate the pulse effect score based on the clearance results and the maximum safe offset. : ; in, Rate the pulse effect; This is the maximum safe offset. To clear the results, 1 is assigned to success and 0 to failure; Construct a reinforcement learning optimization framework, define the state space, action space, and reward function, and update the Q-value table: ; in, This represents the updated Q-value entry, indicating the state. Next action A new estimate of the expected cumulative reward to be obtained later; This represents the Q-value entry before the update, indicating the state. Next action Historical cumulative reward estimates; The learning rate is 0.1. This is a discount factor with a value of 0.9; This is the current state. For the current action, The next state; Indicates the next state All possible actions The corresponding maximum Q value; the reward function is Action space = { , , , }; ={Blockage type, patient physiological parameters}; Based on the updated Q-value table, the weight coefficients of the impulse parameter determination model are adjusted: ; in, This indicates the pulse parameter determination model after adjustment. Each weighting coefficient This indicates the pulse parameter determination model before adjustment. Each weighting coefficient Indicates the Q value for the first... Partial derivatives of each weight; The optimal pulse parameter matching strategy is determined by determining the weight coefficients of the model based on the adjusted pulse parameters.

[0040] This invention constructs a closed-loop adaptive optimization system. When pulse intervention is triggered, the system automatically depressurizes through the micro-pressure buffer chamber built into the pulse airway 507, simultaneously calculating a safety offset and dynamically adjusting the pulse parameters accordingly. Subsequently, a pulse effectiveness score is calculated based on the obstruction clearance result and the maximum safety offset, balancing clearance efficiency and operational safety. Building upon this, a reinforcement learning framework is constructed, using the obstruction type and patient physiological parameters as the state space, the pulse parameter combination as the action space, and the effectiveness score as the reward function. The Q-learning algorithm updates the Q-value table, and the weight coefficients of the pulse parameter determination model are adjusted using gradients based on the updated Q-value table, thereby continuously optimizing subsequent pulse strategies.

[0041] For example, in a patient who underwent abdominal surgery, the drainage tube showed signs of blockage, and the flow rate decreased from an initial 60 mL / h to 25 mL / h.

[0042] Algorithm execution flow: Image data acquisition and processing: The fiber optic imaging module acquires 5 frames of images inside the drainage tube within 5 seconds. The CNN segmentation model processes the images, identifies the blockage area, and extracts morphological feature parameters: the blockage coverage area ratio is 52%, the blockage edge clarity index is 0.78, the blockage texture feature vector analysis shows high-contrast particle features, and the blockage color distribution is dominated by red, which is consistent with the characteristics of blood clots.

[0043] Time series analysis and pressure monitoring: A negative pressure of -10 kPa was applied for testing, and the response of the blockage was observed: deformation rate 0.0625; adhesion strength index 0.82. Pressure data collected by the negative pressure sensor was analyzed, and the following results were obtained: pressure difference amplitude 3.8 kPa, pressure fluctuation standard deviation 1.5 kPa, flow resistance coefficient 0.413 kPa·min / mL, and flow resistance coefficient change rate 38%.

[0044] Blockage Classification: The classifier, which incorporates all input parameters, determines type B, adhesive partial blockage.

[0045] Call the type B dedicated model to generate initial pulse parameters: pulse intensity -14kPa, pulse frequency 2.5Hz, duration 2.8 seconds, and waveform square wave.

[0046] The pulse was applied while intra-abdominal pressure was monitored. The pressure threshold was set at 14 mmHg, and the measured peak pressure was 16.2 mmHg. The safety margin was 2.2 mmHg (intervention was required if the threshold was exceeded).

[0047] Triggering automatic pressure relief in the miniature pressure buffer chamber and dynamically adjusting pulse parameters: -11.8 kPa 1.25Hz It was 2.58 seconds. It is a square wave.

[0048] Effect evaluation and reinforcement learning optimization: Blockage removal was successful, maximum safe offset was 2.2 mmHg, and pulse effect score was 0.95. Q-value update is performed. , Model weight adjustment: The initial pulse intensity for type B blockage was found to be too high. The adhesion strength index was adjusted to a key weight of the pulse intensity. After the update, the system will automatically reduce the initial pulse intensity by 10% the next time it encounters a similar situation.

[0049] On the one hand, this invention controls intra-abdominal pressure fluctuations within a safe threshold through real-time pressure relief and adaptive parameter adjustment, significantly reducing the risk of iatrogenic injury. On the other hand, the reinforcement learning mechanism enables the system to accumulate clinical experience from each intervention. As the number of uses increases, the accuracy of pulse parameter matching continuously improves, significantly enhancing its adaptability to different types of obstructions and greatly reducing the operational burden on medical staff.

[0050] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

Claims

1. The abdominal cavity drainage and flushing device based on multi-dimensional parameter monitoring, comprising a device support (1), a multi-way rotary valve (5) and two negative pressure pumps (7), characterized in that, Two negative pressure pumps (7) are installed at the rear end of the device support (1), the multi-way rotary valve (5) is arranged at the front end of the device support (1), and a three-way electromagnetic valve (505) is installed between the multi-way rotary valve (5) and the two negative pressure pumps (7), a flow guide pipe (506) and a backflow pipe (508) are arranged between the two negative pressure pumps (7) and the multi-way rotary valve (5), a pipeline detection module (6) is installed on the flow guide pipe (506) and the backflow pipe (508), and the pipeline detection module (6) comprises a negative pressure sensor, a temperature sensor, a volume sensor, a biochemical marker detection module and an optical fiber imaging module.

2. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 1, wherein, The device support (1) is fixedly provided with a protective cover (101), and a mounting rack (2) is welded in the middle of the device support (1), and the mounting rack (2) is connected with the multi-way rotary valve (5) through fixed screws.

3. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 1, wherein, An electrical box (3) and an alarm module (102) are installed in the device support (1), a wireless transmission module and a processor module are installed in the electrical box (3), and the processor module is electrically connected with the alarm module (102) and the pipeline detection module (6).

4. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 1, wherein, A flushing liquid tank (4) is installed at the bottom of the device support (1), a drainage collection tank (401) is installed at the upper end of the flushing liquid tank (4), a sterilization module is arranged in the drainage collection tank (401), and universal wheels are installed at the bottom of the device support (1).

5. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 4, wherein, The flushing liquid tank (4) is communicated with the negative pressure pump (7) through a pipeline, a drainage discharge pipe (509) is arranged between the drainage collection tank (401) and the multi-way rotary valve (5), the drainage discharge pipe (509) adopts a detachable structure, a drainage pipe quick release head (502) is arranged at the front end of the multi-way rotary valve (5), and an abdominal cavity drainage pipe connector (504) is arranged at the front end of the drainage pipe quick release head (502).

6. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 3, wherein, A pulse generator (501) is installed on one side of the multi-way rotary valve (5), a pulse guide pipe (507) is arranged between the pulse generator (501) and the multi-way rotary valve (5), the pulse guide pipe (507) guides the pulse gas emitted by the pulse generator (501) into the multi-way rotary valve (5) to impact the blockage, the three-way electromagnetic valve (505) is sealingly connected with the flow guide pipe (506) and the backflow pipe (508), and sterile electrical connectors are installed between the flow guide pipe (506) and the backflow pipe (508) and the pipeline detection module (6), and the sterile electrical connectors are electrically connected with the electrical box (3).

7. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 1, wherein, A hydraulic high-precision rotary disc (503) is installed on one side of the multi-way rotary valve (5), a hydraulic pump is installed at one end of the hydraulic high-precision rotary disc (503), two swing oil cylinders are connected with the hydraulic pump, the two swing oil cylinders are connected with the hydraulic high-precision rotary disc (503) through gear transmission, and the hydraulic high-precision rotary disc (503) is fixedly connected with the valve body of the multi-way rotary valve (5) through a flange.

8. The multi-dimensional parameter monitoring based abdominal drainage irrigation device according to claim 1, wherein, The electrical box (3) is provided with a power supply, the device support (1) is externally provided with an operation panel and a display screen, the operation panel and the display screen are connected with a processor module in the electrical box (3), the display screen is used for displaying monitoring data of the pipeline detection module (6), and the operation panel is used for controlling the equipment by medical staff.

9. The negative pressure regulating system for the celiac drainage irrigation device based on multi-dimensional parameter monitoring according to claim 1, characterized in that, Comprise: A bedside device layer for performing real-time monitoring and rapid response to the abdominal drainage medical process, which is realized based on the display pipeline detection module (6) and the electrical box (3); A device layer for integrated intelligent drainage, which adopts a multi-channel rotary valve (5) and two negative pressure pumps (7) to perform abdominal drainage work; An intelligent management layer for intelligent monitoring and adaptive control, which is based on the collected data of the pipeline detection module (6) and is realized through intelligent adjustment of the processor module; The intelligent management layer comprises a conventional stable drainage unit, a pre-warning and automatic intervention unit and a manual intervention unit, the conventional stable drainage unit is based on the monitoring parameters of the pipeline monitoring module (6) and maintains stable negative pressure drainage; The pre-warning and automatic intervention unit is based on the monitoring parameters of the pipeline detection module (6) and automatically triggers the following cascade responses: Primary blockage pre-warning: drainage flow decreases and pipe end pressure fluctuates, and the negative pressure pump (7) temporarily increases the negative pressure; Intermediate blockage alarm: primary dredging is invalid, the pressure difference continuously increases, the pulse dredging program is triggered, the pulse generator (501) is started, the multi-channel rotary valve (5) is quickly switched, and pulse flushing is performed; High-risk infection pre-warning: the turbidity of the drainage liquid continuously rises and the pH value decreases, the treatment flushing multi-channel rotary valve (5) is automatically started, the negative pressure pump (7) extracts the flushing liquid in the flushing liquid tank (4), the pulse generator (501) starts a high-frequency low-amplitude sinusoidal pulse, and the negative pressure pump (7) sucks the drainage pipe; Emergency safety alarm: the intra-abdominal pressure exceeds the safety threshold, and the multi-channel rotary valve (5) is immediately switched to closed, and all negative pressure pumps (7) are stopped; Active bleeding pre-warning: the color of the drainage liquid suddenly changes to bright red, and the flow per unit time suddenly increases, the automatic switching is switched to the safety mode, the negative pressure pump (7) is reduced to extremely low negative pressure to maintain drainage, and any automatic flushing or dredging operation is closed.

10. The negative pressure regulating system for the celiac drainage irrigation device based on multi-dimensional parameter monitoring according to claim 9, characterized in that, The intelligent management layer is configured to: Obtain multiple fiber images through the optical fiber imaging module in the pipeline detection module (6); Segment the fiber images by using a convolutional neural network, extract the blockage shape feature parameters, and the blockage shape feature parameters comprise: a blockage coverage area ratio, a blockage edge sharpness index, a blockage texture feature vector and a blockage color distribution histogram; Perform time series analysis based on the multiple fiber images, calculate the blockage motion characteristics, and the blockage motion characteristics comprise: a blockage deformation rate with negative pressure change and a blockage adhesion strength index with the pipe wall; Collect pressure waveform data through the negative pressure sensor in the pipeline detection module (6); According to the pressure waveform data, extract the pressure fluctuation time-frequency domain features and construct a pressure-flow relationship model; Intelligently classify the blockage situation according to the blockage shape feature parameters, the pressure fluctuation time-frequency domain features and the pressure-flow relationship model, and determine the blockage situation. According to the blockage situation, the best pulse parameters are automatically matched; The pulse generator (501) monitors the intra-abdominal pressure changes in real time during pulse execution according to the best pulse parameters; According to the intra-abdominal pressure changes, pulse intervention is carried out; pulse intervention is a safety mechanism for dynamically adjusting or stopping operation during pulse execution; According to the best pulse parameters and the results of pulse intervention, the subsequent best pulse parameter matching strategy is optimized through the reinforcement learning algorithm; According to the best pulse parameters and the results of pulse intervention, the subsequent best pulse parameter matching strategy is optimized through the reinforcement learning algorithm, including: When the pulse intervention is triggered, the pressure in the pulse airway (507) is released through the built-in micro pressure buffer cavity, and the pulse parameters are adjusted according to the safety offset; According to the clearing results and the maximum safety offset, the pulse effect score is calculated; An reinforcement learning optimization framework is constructed, the state space, action space and reward function are defined, and the Q value table is updated; Based on the updated Q value table, the weight coefficient of the pulse parameter determination model is adjusted; According to the adjusted weight coefficient of the pulse parameter determination model, the subsequent best pulse parameter matching strategy is determined.

Images