Patch multi-probe ultrasound based monitoring system for renal pelvic pressure and overdistension

Abstract

The application discloses a renal pelvis internal pressure and overexpansion monitoring system based on a patch type multi-probe ultrasound, and relates to the technical field of ultrasonic diagnosis and monitoring.The application provides high-quality original ultrasonic signals by taking a patch type multi-probe ultrasonic transducer as a hardware basis, combines accurate renal pelvis shape data and strain distribution data after processing as support, performs multi-data fusion on individual characteristics of a patient, constructs a multi-module deep collaborative monitoring system, inputs reliable internal pressure estimation values and volume data into an early warning module of dynamic threshold adjustment, hierarchical early warning response and data traceability management, and realizes the accurate and personalized dynamic monitoring function of the renal pelvis internal pressure and overexpansion under a non-invasive state.The dynamic monitoring and risk early warning of the renal function related indexes can be completed non-invasively throughout, complete data support is provided for formulating individualized diagnosis and treatment schemes and early intervention of renal damage in the clinic, and the application is suitable for the diversified clinical needs of long-term monitoring and emergency response.

Description

Technical Field

[0001] This invention relates to the field of ultrasound diagnostic monitoring technology, and in particular to a renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound. Background Technology

[0002] Abnormal intrarenal pelvic pressure and excessive distension are key contributing factors to hydronephrosis and renal function impairment. Accurate monitoring of relevant parameters is crucial for the diagnosis and treatment of kidney diseases. For example, Chinese patent application CN118468137A discloses a training method for a renal pelvic pressure prediction model based on a Bayesian neural network. The method includes: acquiring multiple sets of training data through a renal pelvic pressure influencing factor acquisition platform; performing binary splitting on the first sample data included in each of the multiple sets of training data to obtain multiple second sample data; performing principal component decomposition on the multiple second sample data to obtain multiple sample datasets; and training an initial model using the multiple sample datasets and multiple labeled data to obtain a renal pelvic pressure prediction model.

[0003] However, existing Bayesian neural network prediction techniques still have many limitations: 1. The data acquisition process relies on ex vivo samples or intraoperative invasive sensors, which cannot match the real physiological environment of the human body, are prone to iatrogenic damage, and cannot be monitored continuously for a long time, making it difficult to meet the monitoring needs of dynamic fluctuations in intraoperative intraoperative pressure. 2. The prediction model only focuses on mechanical and physical factors and does not integrate core physiological indicators such as the three-dimensional morphology of the renal pelvis and renal cortical blood flow. The prediction dimension is singular and disconnected from clinical physiological logic. Furthermore, it does not include the patient's individual basic data, resulting in poor adaptability to special pathological states such as chronic kidney disease, and limited prediction accuracy and reliability. 3. Conventional single-probe ultrasound can only acquire two-dimensional morphology and rough volume data of the renal pelvis, making it difficult to accurately assess renal pelvis wall strain and internal pressure. In addition, it lacks a positioning and calibration mechanism, and changes in body position can easily lead to data deviation. 4. Existing early warning systems mostly use fixed thresholds and do not take into account individual patient underlying diseases and differences in tissue characteristics, which can easily lead to delayed early warnings or false alarms. Summary of the Invention

[0004] The purpose of this invention is to provide a renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound. This system simultaneously acquires three-dimensional morphology, wall strain, and renal cortical blood flow data of the renal pelvis using a multi-probe ultrasound transducer. After positioning calibration and deviation correction, combined with a multi-dimensional mapping model, it achieves accurate estimation of renal pelvis pressure, adapting to different body positions and patient conditions. This improves the intervention efficiency for kidney-related emergencies and provides a reliable auxiliary tool for clinical kidney diagnosis and treatment, thereby solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A patch-type multi-probe ultrasound-based system for monitoring renal pelvic pressure and over-distension includes patch-type multi-probe ultrasound transducers arranged in a ring or matrix to scan the renal pelvis region from different angles, acquire ultrasound signals from the renal pelvis and renal cortex, and transmit the acquired ultrasound signals to a terminal monitoring platform. The terminal monitoring platform includes: The ultrasonic signal processing module is used to receive ultrasonic signals, perform preprocessing, and output the preprocessed ultrasonic signals. The three-dimensional morphology construction module is used to reconstruct a real-time three-dimensional model of the renal pelvis based on the preprocessed ultrasound signal, and to obtain real-time three-dimensional morphological data and volume data of the renal pelvis based on the real-time three-dimensional model of the renal pelvis. The strain detection module is used to analyze the degree of deformation of different regions of the renal pelvis wall under pressure based on pre-processed ultrasound signals, and to obtain renal pelvis wall strain distribution data. The renal cortical blood flow assessment module is used to extract renal cortical blood flow-related parameters based on preprocessed ultrasound signals and obtain renal cortical blood flow assessment data. The renal pelvis pressure estimation module is used to input real-time three-dimensional morphological data, volume data, renal pelvis wall strain distribution data, and the input patient's basic data into a preset mapping model and output the renal pelvis pressure estimation value in real time. The early warning module is used to compare the estimated intrarenal pressure and renal pelvis volume data with the preset intrarenal pressure threshold and renal pelvis volume threshold, respectively, and trigger an early warning signal based on the comparison result.

[0006] Further, the process of acquiring ultrasound signals from the renal pelvis and renal cortex includes: Acquire positioning marker data of patch-type multi-probe ultrasonic transducers, and track the spatial displacement coordinates of ultrasonic transducers in real time based on the positioning marker data; During initial positioning, a reference position parameter is established. The spatial displacement coordinates are compared with the reference position parameter to obtain the displacement deviation value. When the displacement deviation value exceeds the preset allowable range, the calibration process is automatically triggered. During the calibration process, the offset direction and offset amount of the ultrasonic transducer are calculated based on the displacement deviation value, and the ultrasonic transducer is driven to perform adaptive fine adjustment so that the relative position accuracy between the transducer and the renal pelvis region meets the preset accuracy standard. A replaceable water-based medical coupling agent layer is coated on the surface of a medical silicone substrate. The viscosity and thickness parameters of the coupling agent layer meet the acoustic impedance matching requirements between the ultrasound transducer and the skin. The ultrasound transducer is controlled to perform a fan-shaped scan around the central axis of the renal pelvis within a preset angle range, and ultrasound signals from the renal pelvis and renal cortex are acquired synchronously at a preset frequency. The ultrasonic signal undergoes preliminary noise reduction processing to remove environmental electromagnetic interference signals, and the ultrasonic signal after preliminary noise reduction processing is transmitted to the terminal monitoring platform.

[0007] Furthermore, the process of acquiring real-time three-dimensional morphological and volume data of the renal pelvis includes: The preprocessed ultrasound signal and the real-time positioning data of the patch-type multi-probe ultrasound transducer were acquired to construct an initial three-dimensional model of the renal pelvis. At the same time, the scanning angle deviation of each ultrasound transducer was calculated based on the real-time positioning data. The correction coefficient weights were adjusted according to the signal-to-noise ratio of each transducer to correct the angle deviation of the initial three-dimensional model of the renal pelvis. The corrected initial 3D model of the renal pelvis is segmented into regions. The grayscale threshold and texture complexity features of the renal pelvis wall tissue and the fluid region in the renal pelvis are extracted. The pure renal pelvis region is segmented from the renal cortex and renal medulla, and the pure renal pelvis region is output. The renal pelvis volume is calculated based on the pure renal pelvis region, and the volume calculation results are corrected by taking into account the dynamic changes in the renal pelvis morphology. The key dimensions of the renal pelvis calculated in real time are compared with historical baseline parameters to obtain the parameter deviation value. When the parameter deviation value exceeds the preset accuracy range, the model is reconstructed. If the deviation values ​​of all key dimensional parameters are less than the preset accuracy range, the model data is deemed reliable, and real-time three-dimensional morphological data of the renal pelvis and the corrected renal pelvis volume data are generated.

[0008] Furthermore, the strain detection module acquires strain distribution data of the renal pelvis wall, specifically including: Based on the preprocessed ultrasound signal, the real-time positioning data of the ultrasound transducer is retrieved, the displacement information of different positions of the renal pelvis wall is extracted, and the displacement information is corrected based on the positioning deviation correction parameter. Based on the corrected displacement information, and combined with the pure renal pelvis region, different tissue characteristic regions are distinguished. Based on the ultrasound signal reflection characteristics of each region, the tensile strain value and compressive strain value of the corresponding region are calculated respectively. Spatial mapping of tensile and compressive strain values ​​at different locations is performed and correlated with real-time three-dimensional morphological data of the renal pelvis to generate a strain distribution map of the renal pelvis wall. The strain distribution map of the renal pelvis wall is converted into digital strain distribution data, including strain values, regional coordinates and strain type, and output to the renal pelvis pressure estimation module based on the reliability indicator of the ultrasound signal-to-noise ratio.

[0009] Furthermore, the displacement information is corrected based on the positioning deviation correction parameters, specifically including: The positioning marker data of the ultrasonic transducer at the moment of ultrasonic signal acquisition is extracted in real time to determine the spatial displacement and tilt angle of the ultrasonic transducer, establish a one-to-one correspondence between the positioning deviation at that moment and the renal pelvis wall displacement information acquisition channel, and clarify the degree of influence of the positioning deviation on each acquisition channel. Based on the correspondence between positioning deviation and renal pelvis wall displacement information acquisition channels, and combined with the contour features of the pure renal pelvis region, a differential correction coefficient is assigned to each displacement information acquisition channel. The displacement information is split according to the acquisition channel. The displacement information corresponding to each acquisition channel is initially compensated by the correction coefficient. Based on the characteristic that the displacement change trend of adjacent channels in the same renal pelvis wall region is consistent, the compensated displacement information is corrected. The corrected displacement information is compared with the contour change trend of the real-time three-dimensional morphological data of the renal pelvis to generate a trend consistency deviation value. If the trend consistency deviation value exceeds the preset logic range, the correction coefficient is readjusted until the deviation meets the logic consistency requirements, and the corrected displacement information is output.

[0010] Furthermore, the intrarenal pelvic pressure estimation module specifically includes: Simultaneously receive real-time three-dimensional morphological data and volume data of the renal pelvis, digital strain distribution data, and input patient individual basic data, and perform consistency verification on the received data; Based on the validated data, the multi-dimensional input vector of the preset mapping model is constructed, with renal pelvis volume data as the basic dimension, strain values ​​of different regions of the renal pelvis wall as the sensitive dimension, patient individual basic data as the adaptation dimension, and blood flow velocity change trend output by the renal cortical blood flow assessment module as an auxiliary reference dimension. Differentiated strain and pressure mapping weights were assigned to different regions of the renal pelvis, and the local pressure estimates for each region were calculated by combining volume data and individual adaptation dimensions. The estimated local pressure values ​​for each region are corrected by using auxiliary reference dimensions; Based on the contribution of each region to the conduction of intrarenal pelvis pressure, the weighted average of the corrected local pressure estimates of each region is calculated to generate a global intrarenal pelvis pressure estimate. At the same time, based on the reliability of the multi-dimensional input vector and the correction magnitude, an estimation credibility score is added and output to the early warning module.

[0011] Furthermore, based on the contribution of each region to the conduction of intrarenal pelvic pressure, a weighted average is calculated on the corrected local pressure estimates for each region to generate a global intrarenal pelvic pressure estimate, specifically including: By combining real-time three-dimensional morphological data and strain distribution data of the renal pelvis, a contribution quantification model is constructed. The volume ratio of each region of the renal pelvis is used as the basic weight, and the correlation coefficient between the strain value and pressure of the region is superimposed. At the same time, the patient's individual basic data is introduced to dynamically generate the real-time contribution weight of each region. The revised local pressure estimates for each region were stratified and classified, and the renal pelvis was divided into pressure-dominant areas and pressure-conduction areas according to its anatomical function. The local pressure estimates within the same functional area were weighted and averaged to obtain the average pressure of the functional area. Based on the pressure transmission efficiency model, a global integrated weight is assigned to the pressure dominance zone and the pressure transmission zone, and the weight value is dynamically adjusted with the change of renal pelvis volume; The average pressure of each functional area is weighted twice according to the global integration weight to obtain the estimated value of global intrarenal pelvic pressure; at the same time, the calculation process data of the contribution weight of each area is recorded.

[0012] Furthermore, the early warning module also includes: Based on individual patient data and historical monitoring data, a dynamic threshold update mechanism is constructed, and the threshold trigger sensitivity is dynamically adjusted in conjunction with the rate of change of renal pelvis volume and internal pressure. The warning signals are classified according to the degree of exceedance of the comparison results, and the classified warning signals are output in the form of sound and light, and pushed to the medical terminal through remote communication. The warning trigger time, exceedance data and warning level are recorded in real time, and a warning log is generated.

[0013] Furthermore, before outputting the global intrapelvic pressure estimate, the renal pelvic pressure estimation module performs an acoustic-viscoelastic-thermal coupling dynamic correction step to eliminate estimation errors caused by modulus drift and tissue stress relaxation due to ultrasonic thermal effects. The dynamic correction step specifically includes: Obtain the global intrarenal pressure estimate and real-time renal pelvis volume data output by the preset mapping model; The time-of-flight deviation of ultrasound signals is used to characterize changes in tissue sound velocity. Combined with the obtained global intrarenal pelvic pressure estimate and real-time renal pelvic volume data, the corrected true intrarenal pelvic pressure value is obtained through the following calculation formula. : in, The final output is the corrected true intrarenal pelvis pressure value, in Pascals (Pa). The estimated global intrarenal pressure output by the preset mapping model is expressed in Pascals (Pa). It is a dimensionless thermo-acoustic modulus coupling coefficient, used to correct the nonlinear stiffness drift of tissue caused by temperature changes; This is the preset density constant of the renal pelvis wall tissue, expressed in kilograms per cubic meter (kg / m³). The real-time equivalent sound velocity is calculated based on the flight time of the ultrasound signal within the renal pelvis wall at the current moment, and the unit is meters per second (m / s). The reference tissue sound velocity during initial system calibration is expressed in meters per second (m / s). The term characterizes the dynamic shift in bulk modulus caused by thermal effects and tissue physical hardening; This is the real-time volume data of the renal pelvis at the current moment, in cubic meters (m³). The initial reference volume of the renal pelvis is expressed in cubic meters (m³). Hencky log-true strain characterizing the process of renal pelvis dilation; This is a dimensionless viscous damping correction factor; The effective viscosity coefficient of renal pelvis tissue, expressed in Pascals per second (Pa). ); The first derivative of the renal pelvis volume with respect to time represents the rate of change of volume, expressed in cubic meters per second (m³ / s). Based on the corrected actual intrarenal pelvis pressure value, it is output as the final monitoring data to the early warning module for comparison with the preset intrarenal pelvis pressure threshold and to trigger the corresponding early warning signal.

[0014] Furthermore, the terminal monitoring platform also includes a respiratory shear slip vector decoupling and virtual aperture following module, which is used to solve the problem of relative displacement of the kidney caused by the patch being fixed to the skin and the kidney moving with respiration; The specific execution process of the breathing shear slip vector decoupling and virtual aperture following module is as follows: While acquiring ultrasound signals, the relative motion trajectory between the renal cortex surface and the subcutaneous tissue boundary layer was extracted using ultrasound speckle tracking technology, and the interlaminar shear slip vector field that changes with the respiratory cycle was calculated. Based on the interlayer shear slip vector field, a virtual region of interest coordinate system is constructed that dynamically floats relative to the skin surface. The origin of this coordinate system is always locked at the anatomical center of the renal pelvis. Based on the projection position of the virtual ROI coordinate system in the patch-type multi-probe ultrasound transducer matrix, a dynamic activation strategy is generated in real time: without moving the physical position of the ultrasound transducer, the subset of micro ultrasound transducers in the active state in the matrix is ​​dynamically switched at a millisecond time resolution so that the physical center of the synthesized sound beam is always aligned with the anatomical center of the renal pelvis after the slippage. Based on the modulus and depth components of the interlaminar shear slip vector, the transmission and reception delays of each activated subset of miniature ultrasonic transducers are adjusted, and the focal length of the electron acoustic lens is dynamically adjusted so that the sound field energy focus point falls on the slipped renal pelvis wall. At the same time, dynamic aperture control is performed, following the principle of constant F number. As the focusing depth increases, the aperture range of the activated subset of miniature ultrasonic transducers is automatically expanded to form an electron-suspended aperture that slides relative to the skin surface, ensuring that the renal pelvis imaging is always located at the sound field energy focusing center during the patient's breathing.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention utilizes a patch-type multi-probe ultrasound transducer as its hardware foundation to provide high-quality raw ultrasound signals. Combined with processed, precise renal pelvis morphology and strain distribution data, and integrated with individual patient characteristics, it constructs a multi-module, deeply collaborative monitoring system. Reliable internal pressure estimates and volume data are input into a dynamic threshold adjustment, graded early warning response, and data traceability management module. This enables precise, personalized, and dynamic monitoring of renal pelvis pressure and over-distension in a non-invasive manner. It can perform dynamic monitoring and risk warning of kidney function-related indicators non-invasively throughout the entire process, providing comprehensive data support for developing individualized treatment plans and early intervention for kidney damage, adapting to diverse clinical needs for long-term monitoring and emergency response. Attached Figure Description

[0016] Figure 1 This is a block diagram of the renal pelvis pressure and over-distension monitoring system of the present invention; Figure 2 This is a flowchart of the renal pelvis pressure and over-distension monitoring system of the present invention. Detailed Implementation

[0017] 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] To address the technical challenges of existing intrapelvic pressure monitoring technologies, such as invasive procedures that can easily lead to injury and infection, limited monitoring dimensions resulting in inaccurate data, poor model adaptability, and a lack of real-time tiered early warning systems, please refer to [link to relevant documentation]. Figures 1-2 This embodiment provides the following technical solution: A patch-type multi-probe ultrasound-based system for monitoring renal pelvic pressure and over-distension includes patch-type multi-probe ultrasound transducers arranged in a ring or matrix to scan the renal pelvis region from different angles, acquire ultrasound signals from the renal pelvis and renal cortex, and transmit the acquired ultrasound signals to a terminal monitoring platform. The terminal monitoring platform includes: The ultrasound signal processing module, connected to the patch-type multi-probe ultrasound transducer, is used to receive ultrasound signals and perform preprocessing, including noise reduction based on the noise characteristics of the ultrasound signals to remove random noise from the signals; filtering based on the frequency range of the ultrasound signals from the renal pelvis and renal cortex to retain the effective signal frequency band; amplifying the filtered signals to increase the signal strength; and outputting the preprocessed ultrasound signal. A three-dimensional morphology construction module, connected to the ultrasound signal processing module, is used to reconstruct a real-time three-dimensional model of the renal pelvis based on the preprocessed ultrasound signal using a three-dimensional reconstruction algorithm, including a volume rendering algorithm or a surface rendering algorithm, and to obtain real-time three-dimensional morphological data and volume data of the renal pelvis based on the real-time three-dimensional model of the renal pelvis. The strain detection module is connected to the ultrasonic signal processing module and is used to analyze the degree of deformation of different regions of the renal pelvis wall under pressure based on the pre-processed ultrasonic signal and ultrasonic strain elastography technology, and to obtain the strain distribution data of the renal pelvis wall. The renal cortical blood flow assessment module is connected to the ultrasound signal processing module and is used to extract renal cortical blood flow related parameters, including blood flow velocity and blood flow rate, based on the preprocessed ultrasound signal using Doppler ultrasound technology, and to obtain renal cortical blood flow assessment data. The renal pelvis pressure estimation module is connected to the three-dimensional morphology reconstruction module and the strain detection module, respectively. It is used to input the real-time three-dimensional morphological data, volume data, renal pelvis wall strain distribution data, and the input patient's basic data, including age, weight, and basic kidney disease information, into the preset mapping model. The preset mapping model is trained based on clinical data and outputs the renal pelvis pressure estimation value in real time. The early warning module, connected to the renal pelvis pressure estimation module and the three-dimensional morphological reconstruction module, is used to compare the estimated renal pelvis pressure value and renal pelvis volume data with preset renal pelvis pressure threshold and renal pelvis volume threshold, respectively. Based on the comparison results, when the estimated renal pelvis pressure value exceeds the renal pelvis pressure threshold or the renal pelvis volume data exceeds the renal pelvis threshold, an early warning signal is triggered, including an audible and visual warning or a remote communication warning.

[0019] In this embodiment, the patch-type multi-probe ultrasound transducer includes a medical silicone substrate and 64-128 miniature ultrasound transducers. The miniature ultrasound transducers are arranged in a mixed ring and matrix configuration. The core area corresponding to the renal pelvis projection region is arranged in a 6×6-10×10 matrix, and the outer periphery is arranged in 1-3 rings with 12-20 transducers per ring, covering a 3-8mm range of renal cortex around the renal pelvis. The medical silicone substrate is ≤3mm thick and coated with a replaceable water-based medical coupling agent layer with a dielectric constant εr≥35. 2-4 positioning marks are integrated on the back, either infrared or optical, with a positioning accuracy of ±0.2mm. In this embodiment, a patch-type multi-probe ultrasound transducer is used as the hardware foundation to provide high-quality raw ultrasound signals. Combined with processed and accurate renal pelvis morphology data and strain distribution data, and with the patient's individual characteristics, multi-data fusion is performed to construct a multi-module, deeply collaborative monitoring system. Reliable internal pressure estimates and volume data are input into the early warning module for dynamic threshold adjustment, graded early warning response, and data traceability management. This achieves accurate and personalized dynamic monitoring of renal pelvis pressure and over-distension in a non-invasive state. It can complete the dynamic monitoring and risk warning of kidney function-related indicators non-invasively throughout the entire process, providing complete data support for the clinical development of individualized treatment plans and early intervention for kidney damage, and adapting to the diverse clinical needs of long-term monitoring and emergency response.

[0020] In this embodiment, the process of acquiring ultrasound signals from the renal pelvis and renal cortex includes: Acquire positioning marker data of patch-type multi-probe ultrasonic transducers, and track the spatial displacement coordinates of ultrasonic transducers in real time based on the positioning marker data; During initial positioning, a reference position parameter is established. Through a dynamic threshold judgment mechanism, the spatial displacement coordinates are compared with the reference position parameter to obtain the displacement deviation value. When the displacement deviation value exceeds the preset allowable range, the calibration process is automatically triggered. During the calibration process, the offset direction and offset amount of the ultrasonic transducer are calculated based on the displacement deviation value, and the ultrasonic transducer is driven to perform adaptive fine adjustment. After fine adjustment, the relative position accuracy between the transducer and the renal pelvis region is verified through secondary positioning to ensure that the accuracy of the relative position of the transducer and the renal pelvis region meets the preset accuracy standard. A replaceable water-based medical coupling agent layer is coated on the surface of a medical silicone substrate. The viscosity and thickness parameters of the coupling agent layer meet the acoustic impedance matching requirements between the ultrasound transducer and the skin. The ultrasound transducer is controlled to perform a fan-shaped scan around the central axis of the renal pelvis within a preset angle range, and ultrasound signals from the renal pelvis and renal cortex are acquired synchronously at a preset frequency. The ultrasonic signal undergoes preliminary noise reduction processing to remove environmental electromagnetic interference signals, and the ultrasonic signal after preliminary noise reduction processing is transmitted to the terminal monitoring platform.

[0021] In this embodiment, a positioning array is formed by 2-4 positioning marks on the back of the transducer. The three-dimensional coordinates of the transducer are calculated by triangulation. The sampling frequency is kept synchronized with the ultrasonic scanning frequency to ensure the real-time performance of displacement tracking. In this embodiment, the initial threshold of the preset allowable range is adaptively adjusted based on the patient's position. The threshold is set to ±0.1mm in the supine state and expanded to ±0.15mm in the lateral state. At the same time, the displacement change rate judgment is introduced, and an early warning is given when the displacement change rate is >0.05mm / s. In this embodiment, the process of acquiring real-time three-dimensional morphological data and volume data of the renal pelvis includes: The preprocessed ultrasound signal and the real-time positioning data of the patch-type multi-probe ultrasound transducer were acquired to construct an initial three-dimensional model of the renal pelvis. At the same time, the scanning angle deviation of each ultrasound transducer was calculated based on the real-time positioning data. The correction coefficient weight was adjusted according to the signal-to-noise ratio of each transducer. The transducer with higher signal quality corresponds to a larger correction weight. The angle deviation of the initial three-dimensional model of the renal pelvis was corrected to ensure the stitching accuracy of scanning data at different angles. The corrected initial 3D model of the renal pelvis is segmented into regions. The grayscale threshold and texture complexity features of the renal pelvis wall tissue and the fluid region in the renal pelvis are extracted. The pure renal pelvis region is segmented from the renal cortex and renal medulla, and the pure renal pelvis region is output to exclude interference from non-renal pelvis tissues. The renal pelvis volume is calculated based on the pure renal pelvis region, and the volume calculation results are corrected by taking into account the dynamic changes in the renal pelvis morphology, so as to reduce the calculation error caused by morphological irregularities. The key dimensional parameters of the renal pelvis, such as the maximum diameter and depth of the renal pelvis, are calculated in real time and compared with historical benchmark parameters to obtain the parameter deviation value. When the parameter deviation value exceeds the preset accuracy range, the model is reconstructed to ensure the reliability of the output data. If the deviation values ​​of all key dimensional parameters are less than the preset accuracy range, the model data is deemed reliable, and real-time three-dimensional morphological data of the renal pelvis is generated, including the renal pelvis contour curve, the three-dimensional coordinate matrix of the renal pelvis cavity, key dimensional parameters, and the corrected renal pelvis volume data.

[0022] In this embodiment, an adaptive meshing method combined with a biomechanical deformation model is used to calculate the renal pelvis volume. The mesh unit size is dynamically adjusted according to the curvature of the renal pelvis region, and the mesh is finer where the curvature is greater. In this embodiment, angle deviation correction is performed based on signal-to-noise ratio adjustment of correction coefficient weights, regional segmentation of tissue features is extracted, volume correction is combined with dynamic morphological laws, model reconstruction is performed by comparing historical parameters, and volume calculation is optimized by adaptive mesh partitioning combined with biomechanical deformation model to form accurate real-time three-dimensional morphology and volume data of the renal pelvis. This solves the problems of low accuracy of scanning data stitching, interference from non-renal pelvis tissues, and large volume calculation errors caused by irregular morphology in traditional three-dimensional modeling. By dynamically allocating correction weights based on signal-to-noise ratio and combining adaptive mesh partitioning with biomechanical model, the individual differences and irregular features of renal pelvis morphology are adapted to improve data accuracy, making the internal pressure estimation more consistent with the actual physiological state of the kidney, and the volume warning of the early warning module is more accurate, adapting to the monitoring needs of renal pelvis with different morphological characteristics.

[0023] In this embodiment, the strain detection module acquires strain distribution data of the renal pelvis wall, specifically including: Based on the real-time positioning data of the ultrasonic transducer retrieved from the preprocessed ultrasonic signal, ultrasonic strain elastography was used to extract displacement information at different positions of the renal pelvis wall, and the displacement information was corrected based on the positioning deviation correction parameter. Based on the corrected displacement information, and combined with the pure renal pelvis region, we distinguish different tissue characteristics such as the weak area and the thickened area of ​​the renal pelvis wall. Based on the ultrasound signal reflection characteristics of each area, we calculate the tensile strain value and compressive strain value of the corresponding area to avoid strain calculation deviation caused by tissue density differences. Spatial mapping of tensile strain and compressive strain values ​​at different locations is performed and associated with real-time three-dimensional morphological data of the renal pelvis to generate a strain distribution map of the renal pelvis wall. The map contains strain gradient annotations to intuitively present the differences in strain intensity in different regions. The strain distribution map of the renal pelvis wall is converted into digital strain distribution data, including strain values, regional coordinates and strain type, and output to the renal pelvis pressure estimation module based on the reliability indicator of the ultrasound signal-to-noise ratio.

[0024] In this embodiment, the displacement information is corrected based on the positioning deviation correction parameters, specifically including: Positioning deviation correlation: Real-time extraction of positioning marker data of the ultrasonic transducer at the moment of ultrasonic signal acquisition, determination of the spatial displacement and tilt angle of the ultrasonic transducer, establishment of a one-to-one correspondence between the positioning deviation at that moment and the renal pelvis wall displacement information acquisition channel, and clarification of the degree of influence of positioning deviation on each acquisition channel; Dynamic allocation of correction coefficients: Based on the correspondence between positioning deviation and renal pelvis wall displacement information acquisition channels, combined with the contour features of the pure renal pelvis region, such as renal pelvis wall curvature and regional location, differentiated correction coefficients are assigned to each displacement information acquisition channel. For acquisition channels that are close to the transducer displacement direction or have a large renal pelvis wall curvature, the weight of the correction coefficient is increased to enhance the deviation compensation. Layered correction execution: The displacement information is split according to the acquisition channel, and the displacement information corresponding to each acquisition channel is initially compensated by the correction coefficient. Based on the characteristic that the displacement change trend of adjacent channels in the same renal pelvis wall region is consistent, the compensated displacement information is corrected to exclude isolated abnormal correction values. Correction effect verification: The corrected displacement information is compared with the contour change trend of the real-time three-dimensional morphological data of the renal pelvis to generate a trend consistency deviation value, including the difference between the displacement change amplitude and the shape size change amplitude, and the degree of matching between the displacement change direction and the shape contour extension direction. If the trend consistency deviation value exceeds the preset logic range, that is, the displacement change and the shape size change do not match, the correction coefficient is readjusted until the deviation meets the logic consistency requirements, and the corrected high-precision displacement information is output.

[0025] In this embodiment, by dynamically and precisely correcting the positioning deviation and deeply correlating strain with three-dimensional morphology, the system adapts to individual differences and anatomical characteristics of the renal pelvis wall tissue. Through dynamic correction of the positioning deviation, the accuracy of strain data is improved. The renal pelvis pressure estimation module outputs digital strain data with reliability indicators, providing a core basis for distinguishing the deformation sensitivity of different regions of the renal pelvis wall and allocating differentiated mapping weights during internal pressure estimation. This avoids internal pressure estimation deviations caused by inaccurate strain data, strengthens the data source reliability of the internal pressure estimation module, and enables internal pressure estimation to better match the actual deformation characteristics of the renal pelvis wall. Furthermore, it connects with the risk judgment logic of the early warning module and adapts to the strain monitoring needs of renal pelvis with different tissue characteristics.

[0026] In this embodiment, the intrarenal pelvic pressure estimation module specifically includes: Simultaneously receive real-time three-dimensional morphological and volume data of the renal pelvis, digital strain distribution data, and input patient individual basic data. Perform consistency verification on the received data, including data timestamp synchronization to ensure that the data is from the same monitoring time, data reliability identification verification, and remove invalid or low reliability data. Based on the validated data, the renal pelvis volume data is used as the basic dimension, the strain value of different regions of the renal pelvis wall is used as the sensitive dimension to distinguish between weak and thickened areas, the patient's individual basic data is used as the adaptation dimension, and the blood flow velocity change trend output by the renal cortical blood flow assessment module is used as an auxiliary reference dimension to construct a multi-dimensional input vector for the preset mapping model. Because different regions have different sensitivities to pressure deformation, differentiated strain and pressure mapping weights are assigned to different regions of the renal pelvis, such as the renal pelvis body and the neck of the renal pelvis calyces. Combined with volume data and individual adaptation dimensions, the estimated local pressure values ​​of each region are calculated. By using auxiliary reference dimensions, such as the trend of renal cortical blood flow changes, the estimated local pressure values ​​of each region are corrected. If the blood flow velocity in a certain region decreases significantly and the corresponding estimated local pressure value is high, the mapping weight of that region is further adjusted to avoid the influence of strain deviation caused by tissue ischemia on the accuracy of pressure estimation. For example, when the blood flow velocity in a certain region of the renal cortex decreases by more than a preset proportion compared to the previous monitoring period, and the estimated local pressure value of the corresponding region is higher than the average level, the mapping weight of that region is reduced by a preset proportion. The corrected local pressure estimate is more in line with the actual physiological state of the kidney. Based on the contribution of each region to the conduction of intrarenal pelvis pressure, the weighted average of the corrected local pressure estimates of each region is calculated to generate a global intrarenal pelvis pressure estimate. At the same time, based on the reliability of the multi-dimensional input vector and the correction magnitude, an estimation credibility score is added. The estimation credibility score is divided into three levels: high, medium and low. Low credibility score data will trigger a data review prompt from the early warning module and be output to the early warning module.

[0027] In this embodiment, the strain and pressure mapping weights of the renal pelvis body are higher than those of the renal pelvis calyx neck, because the renal pelvis body has a larger volume and is more sensitive to pressure deformation; for patients whose underlying kidney disease is chronic kidney disease, the mapping weights of the thickened area of ​​the renal pelvis wall are lowered compared to healthy individuals, to match the low deformation characteristics of the thickened tissue. In this embodiment, based on the contribution of each region to the conduction of renal pelvis pressure, a weighted average is calculated on the corrected local pressure estimates of each region to generate a global renal pelvis pressure estimate, specifically including: Contribution quantification: Combining the volume ratio of each region of the renal pelvis in real-time three-dimensional morphological data of the renal pelvis, the contour curvature and the strain sensitivity of the region in the strain distribution data, a contribution quantification model is constructed. The volume ratio of each region of the renal pelvis is used as the basic weight, and the correlation coefficient between the strain value and the pressure of the region is superimposed. The higher the strain sensitivity, the larger the correlation coefficient. At the same time, the patient's individual basic data is introduced, such as the conduction efficiency correction coefficient of the thickened area of ​​the renal pelvis wall in patients with chronic kidney disease, to dynamically generate the real-time contribution weight of each region. Stratified weighted preprocessing: The corrected local pressure estimates for each region are stratified and classified according to the anatomical function of the renal pelvis into pressure-dominant regions, such as the renal pelvis body, which has a high volume proportion and strong strain sensitivity, and pressure transmission regions, such as the renal pelvis calyx neck, which has a low volume proportion and weak strain sensitivity. The contribution weight of each sub-region within the same functional region is used to calculate the weighted average of the local pressure estimates in the same functional region, so as to obtain the average pressure of the functional region and avoid the estimation bias caused by the direct superposition of data from different functional regions. Global weight adaptation: The pressure transmission efficiency model trained based on clinical data assigns global integrated weights to the pressure dominance zone and the pressure transmission zone. The global weight of the pressure dominance zone is higher than that of the pressure transmission zone, and the weight value is dynamically adjusted with the change of renal pelvis volume. For example, when the renal pelvis is dilated, the global weight of the pressure dominance zone is further increased to adapt to the concentration of pressure transmission in the dilated state. Global estimate generation: The average pressure of each functional area is weighted twice according to the global integration weight to obtain the global intrarenal pelvic pressure estimate; at the same time, the calculation process data of the contribution weight of each area is recorded to assist in the judgment of the rationality of the weight in the subsequent estimation credibility score. If the contribution weight of a certain area deviates abnormally from the average, it indicates that the data of that area may be interfered with.

[0028] In this embodiment, by utilizing regional deformation sensitivity, the regional volume ratio and anatomical function division of three-dimensional morphological data, and the physiological state feedback of blood flow data, the problems of neglecting regional differences, insufficient individual adaptability, and lack of physiological indicator correction in traditional intrarenal pressure estimation are solved. It adapts to different pathological and morphological states, outputs global intrarenal pressure estimation values ​​and reliability scores, provides core basis for the early warning module to judge the risk level, and ensures the reliability of early warning decisions. The accuracy of intrarenal pressure estimation is improved through multi-module data collaboration, and the reliability of data is strengthened by the reliability score. It connects with the accurate data results of the preceding modules and supports the scientific decision-making of the early warning module, adapting to the intrarenal pressure monitoring needs of patients with different pathological types and morphological characteristics.

[0029] In this embodiment, the early warning module further includes: Based on individual patient data and historical monitoring data, a dynamic threshold update mechanism is constructed. For special populations such as those with chronic kidney disease and renal pelvis wall thickening, the renal pelvis volume threshold is automatically lowered, for example, by 15%-20% compared to healthy individuals, while the tolerance for fluctuations in renal pelvis pressure threshold is increased, for example, from ±5% to ±8%, to adapt to the kidney's tolerance characteristics under pathological conditions. At the same time, combined with the rate of change of renal pelvis volume and internal pressure, if the volume increases by more than a preset value per hour, the threshold trigger sensitivity is dynamically adjusted. The faster the rate of change, the more sensitive the threshold trigger, thus avoiding the risk of acute dilation in advance. Based on the degree of exceedance of the comparison results, the warning signals are classified as mild, moderate, and severe exceedances. The classified warning signals are then output in the form of sound and light and pushed to the medical terminal via remote communication. The warning trigger time, exceedance data, and warning level are recorded in real time, and a warning log is generated. In this embodiment, a three-level warning system is established: a level 1 warning is triggered by a slight exceedance of the threshold, corresponding to an internal pressure / volume exceeding the threshold by less than 10%, no significant decrease in blood flow, and high reliability; a level 2 warning is triggered by a moderate exceedance of the threshold, corresponding to an internal pressure / volume exceeding the threshold by 10%-20% or a decrease in blood flow by 10%-15%, and medium reliability; and a level 3 warning is triggered by a severe exceedance of the threshold, corresponding to an internal pressure / volume exceeding the threshold by more than 20% or a decrease in blood flow by more than 15%, and low reliability. Different warning methods are matched for different levels. The level 1 warning only triggers local audio-visual prompts, a low-frequency buzzer, and a green indicator light; the level 2 warning is superimposed with a remote communication warning, pushing a text alert to the attending physician; and the level 3 warning further triggers an emergency contact mechanism, simultaneously calling the patient's family and pushing emergency suggestions containing monitoring data. This optimizes the warning judgment, avoids the risk of misjudgment from traditional single-indicator warnings, and adapts to the risk characteristics of patients with different pathological states. It ensures rapid intervention for acute risks while reducing unnecessary warning interference for chronic patients, further improving the clinical usability and risk management capabilities of the system.

[0030] In this embodiment, before outputting the global intrapelvic pressure estimate, the renal pelvic pressure estimation module performs an acoustic-viscoelastic-thermal coupling dynamic correction step to eliminate estimation errors caused by modulus drift and tissue stress relaxation due to the ultrasonic thermal effect. The dynamic correction step specifically includes: Obtain the global intrarenal pressure estimate and real-time renal pelvis volume data output by the preset mapping model; The time-of-flight deviation of ultrasound signals is used to characterize changes in tissue sound velocity. Combined with the obtained global intrarenal pelvic pressure estimate and real-time renal pelvic volume data, the corrected true intrarenal pelvic pressure value is obtained through the following calculation formula. : in, The final output is the corrected true intrarenal pelvis pressure value, in Pascals (Pa). The estimated global intrarenal pressure output by the preset mapping model is expressed in Pascals (Pa). It is a dimensionless thermo-acoustic modulus coupling coefficient, used to correct the nonlinear stiffness drift of tissue caused by temperature changes; This is the preset density constant of the renal pelvis wall tissue, expressed in kilograms per cubic meter (kg / m³). The real-time equivalent sound velocity is calculated based on the flight time of the ultrasound signal within the renal pelvis wall at the current moment, and the unit is meters per second (m / s). The reference tissue sound velocity during initial system calibration is expressed in meters per second (m / s). The term characterizes the dynamic shift in bulk modulus caused by thermal effects and tissue physical hardening; This is the real-time volume data of the renal pelvis at the current moment, in cubic meters (m³). The initial reference volume of the renal pelvis is expressed in cubic meters (m³). Hencky log-true strain characterizing the process of renal pelvis dilation; This is a dimensionless viscous damping correction factor; The effective viscosity coefficient of renal pelvis tissue, expressed in Pascals per second (Pa). ); The first derivative of the renal pelvis volume with respect to time represents the rate of change of volume, expressed in cubic meters per second (m³ / s). Based on the corrected actual intrarenal pelvis pressure value, it is output as the final monitoring data to the early warning module for comparison with the preset intrarenal pelvis pressure threshold and to trigger the corresponding early warning signal.

[0031] In this embodiment, to address the nonlinear errors introduced by the thermal effects of continuous operation of the ultrasonic transducer and the inherent viscoelastic characteristics of biological tissues during long-term dynamic monitoring, the renal pelvis pressure estimation module executes a dynamic correction logic based on deep coupling of physical fields. This logic is not a simple numerical compensation, but rather a multi-dimensional real-time calibration closed loop constructed based on the underlying mechanisms of acoustic physics, thermodynamics, and biorheology. In the actual operation of the system, when the patch-type multi-probe ultrasonic transducer is in continuous operation, the piezoelectric crystals inside the array generate mechanical vibrations due to high-frequency electrical excitation. The accompanying energy loss is inevitably converted into heat energy, causing a slight but continuous accumulation of temperature on the probe contact surface and in the superficial subcutaneous tissue. Although this temperature change is within the medical safety range, it is a significant source of interference for high-precision elastic imaging and pressure inversion. According to the thermophysical characteristics of biological soft tissues, there is a complex nonlinear mapping relationship between the tissue's bulk modulus and temperature. As the temperature increases, the molecular conformation of the collagen fiber network relaxes, causing a drift in the macroscopic stiffness of the tissue. This drift is acoustically manifested as a change in the propagation speed of ultrasonic waves. If the system ignores this physical phenomenon and only calculates based on the baseline parameters under constant temperature conditions, it will misjudge the change in thermal stiffness as a stress change caused by the change in intrarenal pelvis pressure, thus causing baseline drift of the monitoring data.

[0032] To eliminate this thermal interference, the system first performs a thermo-acoustic modulus coupling correction step based on the deviation of the sound velocity time of flight. This step begins with the precise measurement of the ultrasound propagation time within the tissue. During the initial baseline calibration phase of the system initialization—when the patient is wearing the device and the probe is cold—the system controls the transducer to emit a series of baseline pulses. Using an internal high-precision timing unit, the system records the entire flight time of the ultrasound waves from the transmitting element, through the skin, subcutaneous fat, and back muscle layer, to the anterior and posterior edges of the renal pelvis wall, and back to the receiving element. Simultaneously, combined with the three-dimensional morphological data at this time, the baseline sound velocity parameters are locked. In the continuous monitoring phase, the system acquires the current ultrasound echo signal in real time during each internal pressure estimation cycle and extracts the current time of flight data using a cross-correlation algorithm. Since the anatomical distance between the transducer and the renal pelvis wall is tracked in real time through three-dimensional modeling, any minute deviation in the time of flight is interpreted by the system as a dynamic change in the medium's sound velocity. The system calculates the equivalent sound velocity of the ultrasound signal within the renal pelvis wall at the current moment and compares it with the baseline tissue sound velocity.

[0033] After acquiring the change in sound velocity, the system, based on the fundamental principles of the acoustic wave equation, uses the squared difference of sound velocity as the core physical quantity characterizing the bulk modulus drift. Specifically, the system calculates the difference between the square of the real-time equivalent sound velocity and the square of the reference tissue sound velocity. Assuming the change in biological tissue density is negligible and can be considered constant, this squared difference is directly proportional to the change in tissue bulk modulus. To convert this abstract acoustic parameter into a specific pressure correction value, the system calls upon a preset thermo-acoustic modulus coupling coefficient. This coefficient is a crucial physical conversion parameter, calibrated based on extensive thermoacoustic experimental data of ex vivo kidney tissue, accurately describing the degree of internal thermal stress accumulation in the renal pelvis wall tissue under a unit change in sound velocity for specific pathological types. The system performs a continuous convolution operation on the preset renal pelvis wall tissue density constant, the calculated squared difference of sound velocity, and this coupling coefficient. To further improve the physical fidelity of the correction, this calculation process also deeply integrates the current deformation state, because the tissue's sensitivity to thermal effects varies under different stretching conditions. The system introduces the concept of Hencky logarithmic true strain, which quantifies the true cumulative deformation of the renal pelvis by calculating the natural logarithm of the ratio of the current real-time volume to the initial reference volume. The system multiplies the aforementioned thermal modulus drift with the Hencky logarithmic true strain value to calculate the first-level pressure correction component caused by the single physical factor of temperature change. The physical meaning of this component is that it represents the difference between the stress that the tissue should exhibit at the current strain and the measured stress if the thermal effect were eliminated.

[0034] Meanwhile, to address the dynamic errors caused by the viscoelastic rheological characteristics of biological soft tissues, the system performs a viscoelastic damping correction step in parallel. The renal pelvis wall, as a biopolymer material filled with interstitial fluid and blood, exhibits a significant time-dependent mechanical response. In clinical pathological processes, such as hydronephrosis caused by acute ureteral stone obstruction, the renal pelvis volume increases rapidly within a short period. At this time, the flow of matrix fluid and the slippage of polymer chains within the tissue generate significant internal frictional resistance, i.e., viscous damping. This damping effect causes the renal pelvis wall to appear "hardened," resulting in an instantaneously measured internal pressure far exceeding the static pressure of the same volume. Conversely, in chronic dilation or during periods of stable disease, the tissue has sufficient time for stress relaxation, resulting in minimal viscous resistance, and the internal pressure is determined more by elastic deformation. If the monitoring system cannot distinguish between these two states, it may underestimate the risk in the acute phase or misjudge the pressure in the chronic phase.

[0035] To accurately reproduce this rheological process, the system performs first-order differential processing on the continuously acquired real-time renal pelvis volume data in the time dimension. The system calculates the increase in renal pelvis volume at the current sampling moment relative to the previous moment and divides it by the time interval to obtain the rate of change of renal pelvis volume over time. This physical quantity directly reflects the instantaneous velocity of renal pelvis expansion or contraction. To eliminate the influence of individual organ size differences on the calculation results and achieve normalization, the system divides this rate of change of volume by the current real-time volume to obtain the relative volume change rate, which is mechanically equivalent to the strain rate of the tissue. Subsequently, the system introduces the effective viscosity coefficient of the renal pelvis tissue. This coefficient reflects the average viscous characteristics of the renal pelvis smooth muscle layer and connective tissue, and is usually adapted during system initialization based on the patient's age and underlying medical history (such as the presence of renal fibrosis). The system multiplies the effective viscosity coefficient, the normalized relative volume change rate, and a dimensionless viscosity damping correction factor. The result of this calculation is the second pressure correction component generated by the tissue viscosity effect. When rapid dilation of the renal pelvis is detected, this component is positive and significant, indicating that the system recognizes a large dynamic resistance at this time; when the volume change is gradual, this component approaches zero.

[0036] Finally, the renal pelvis pressure estimation module performs a comprehensive correction calculation. The system receives the initial global renal pelvis pressure estimate output by the preset mapping model based on the static elasticity assumption, and uses it as the baseline. Subsequently, based on the principles of energy superposition and stress decomposition, the system loads the calculated thermo-acoustic modulus coupling correction value (used to eliminate thermal drift) and viscoelastic damping correction value (used to compensate for dynamic rheological resistance) onto the baseline in the form of an algebraic sum. Specifically, the system subtracts the spurious high-pressure component caused by thermohardening (or replenishes the low-pressure component caused by thermosoftening) from the total estimate, and separates or labels the viscous component according to the clinical monitoring purpose (such as obtaining pure hydrostatic pressure or assessing total wall tension). After this deep cleaning by dual physical mechanisms, the system outputs the corrected true renal pelvis pressure value. This value is no longer a fuzzy estimate affected by environmental and process interference, but a precise indicator that truly reflects the physical state of the fluid within the renal pelvis. Finally, this true value is transmitted to the early warning module as the sole reliable basis for triggering graded early warning signals, ensuring that the system can still provide standard monitoring data in complex and ever-changing clinical environments.

[0037] In this embodiment, the terminal monitoring platform further includes a respiratory shear slip vector decoupling and virtual aperture following module, which is used to solve the problem of relative displacement of the kidney caused by the movement of the kidney with respiration while the patch is fixed to the skin. The specific execution process of the breathing shear slip vector decoupling and virtual aperture following module is as follows: While acquiring ultrasound signals, the relative motion trajectory between the renal cortex surface and the subcutaneous tissue boundary layer was extracted using ultrasound speckle tracking technology, and the interlaminar shear slip vector field that changes with the respiratory cycle was calculated. Based on the interlayer shear slip vector field, a virtual region of interest coordinate system is constructed that dynamically floats relative to the skin surface. The origin of this coordinate system is always locked at the anatomical center of the renal pelvis. Based on the projection position of the virtual ROI coordinate system in the patch-type multi-probe ultrasound transducer matrix, a dynamic activation strategy is generated in real time: without moving the physical position of the ultrasound transducer, the subset of micro ultrasound transducers in the active state in the matrix is ​​dynamically switched at a millisecond time resolution so that the physical center of the synthesized sound beam is always aligned with the anatomical center of the renal pelvis after the slippage. Based on the modulus and depth components of the interlaminar shear slip vector, the transmission and reception delays of each activated subset of miniature ultrasonic transducers are adjusted, and the focal length of the electron acoustic lens is dynamically adjusted so that the sound field energy focus point falls on the slipped renal pelvis wall. At the same time, dynamic aperture control is performed, following the principle of constant F number. As the focusing depth increases, the aperture range of the activated subset of miniature ultrasonic transducers is automatically expanded to form an electron-suspended aperture that slides relative to the skin surface, ensuring that the renal pelvis imaging is always located at the sound field energy focusing center during the patient's breathing.

[0038] In this embodiment, the terminal monitoring platform is equipped with a respiratory shear slip vector decoupling and virtual aperture following module. Its core logic is that although the physical probe is fixed to the patient's back skin without moving, through algorithm control, the ultrasonic wave transmission and reception window (i.e., the effective aperture) can slide freely on the probe surface, and the sliding trajectory is completely consistent with the respiratory movement trajectory of the kidney in the body, thereby achieving relatively static "magnetic" locking monitoring.

[0039] The specific workflow of this module begins with the real-time perception and decoupling of the movement field of tissues within the body. In conventional ultrasound imaging, the system typically assumes the object being measured is stationary. However, in the application scenario of this invention, the kidney, as a retroperitoneal organ, is directly driven by the contraction and relaxation of the diaphragm, resulting in significant periodic reciprocating motion along the longitudinal axis of the body (head-to-tail direction), with displacement amplitudes reaching several centimeters. The patch probe, however, is relatively fixed in position due to the constraints of the ribs and muscles on the surface. This leads to a shear slip layer between the probe and the kidney that varies with the respiratory rhythm, primarily located between the perirenal fat capsule and the deep fascia of the back. To capture this motion, the system intermittently transmits high-frame-rate plane wave signals for motion detection between the transmission of conventional imaging pulses. The system utilizes ultrasound speckle tracking technology to focus on analyzing the echo signal differences between the strongly echogenic capsule on the surface of the renal cortex and the back muscle fascia layer. Because biological tissues contain numerous tiny scatterers (such as cell clusters), they form unique speckle textures in ultrasound images, and these textures undergo overall translation as the tissue moves physically. The system employs a cross-correlation search algorithm to perform inter-frame comparisons of continuously acquired ultrasound radio frequency signals, accurately calculating the spatial displacement vector of speckle texture within the region of interest. Through statistical filtering of massive local vectors, the system eliminates random noise and high-frequency micro-motions caused by heartbeats, resolving the interlaminar shear slip vector field representing the overall movement trend of the kidney. This vector field is a dynamic vector containing three-dimensional spatial information, precisely describing the kidney's offset direction (e.g., towards the foot or head), offset distance, and movement velocity relative to the probe at the current moment.

[0040] Based on the real-time acquired interlayer shear slip vector field, the system enters the coordinate system reconstruction stage at the logical operation level. Traditional ultrasound systems use a fixed Eulerian coordinate system with the center of the probe's physical array as the origin, which leads to continuous drift of the observation section when the target moves. This system innovatively constructs a virtual region of interest (ROI) coordinate system that dynamically floats relative to the skin surface. This is a coordinate system based on the Lagrange perspective. The origin of this coordinate system is forcibly anchored by the algorithm logic to the anatomical center of the renal pelvis after slippage. Specifically, the system uses the real-time calculated slip vector to perform an inverse transformation of the physical coordinate system. For example, when the kidney is detected to have slipped N millimeters towards the foot, the system in the logical model also translates the origin of the virtual ROI coordinate system towards the foot by N millimeters on the physical array plane. This means that regardless of whether the patient's inhalation causes the kidney to move downward or exhalation causes the kidney to move upward, this virtual coordinate system remains mathematically static relative to the kidney.

[0041] After establishing this dynamic coordinate system, the system performs dynamic tracking and activation of the virtual aperture. The patch-type multi-probe ultrasound transducer used in this invention employs a high-density matrix arrangement (e.g., 10×10 or more elements) in its hardware design. Its physical coverage is significantly larger than the actual projected area of ​​the kidney, providing a physical redundancy basis for the electronic movement of the beam. At any given monitoring moment, the system does not activate all transducer elements, as full-element transmission would not only consume enormous power but also reduce the frame rate. Based on the projection position of the virtual ROI coordinate system onto the physical array, the system intelligently selects and activates a subset of elements in optimal observation positions. These activated elements constitute the current effective aperture or electronic observation window. For example, when the patient takes a deep breath and the kidney moves downwards (towards the feet), the interlaminar shear slip vector points towards the feet. Within a millisecond time window, the system controller, through the underlying multiplexed switching circuit, cuts off the excitation signal of a row of elements above the matrix (head side) while simultaneously activating a row of elements below the matrix (foot side) that was previously in a dormant state. The physical effect of this operation is that, although the physical patch itself does not move, the ultrasonic transmitting and receiving areas in operation slide downwards a corresponding distance on the patch plane. Because the speed of electronic switching is much faster than the speed of human respiration, this sliding appears as a continuous and smooth following in imaging. Through this mechanism, the system ensures that the physical central axis of the synthesized sound beam remains highly aligned with the anatomical central axis of the renal pelvis after the slide, thus preventing energy attenuation and image resolution degradation caused by the target deviating from the acoustic axis.

[0042] In addition to planar positioning, the system must also address depth-direction focusing and angular deflection. This is because renal respiratory movements are often three-dimensional, involving not only vertical translation but also changes in depth (i.e., alterations in the distance between the kidney and the body surface) and slight rotation. The system adjusts the beamsynthesizer parameters in real-time based on the modulus and depth component of the interlaminar shear slip vector. For depth variations, the system recalculates the transmit and receive delays of each active element and dynamically adjusts the focal length of the electroacoustic lens to ensure that the strongest acoustic energy focal point always falls on the moved renal pelvis wall, rather than on the perirenal fat or muscle layer. Simultaneously, to maintain consistent lateral resolution, the system implements dynamic aperture control (constant F-number), automatically expanding the aperture range of the active elements as the focusing depth increases, introducing more elements to participate in imaging and maintaining beam fineness. Regarding angular deviation, the system utilizes the phased array principle to control the main lobe direction of the synthesized beam by adjusting the emission phase difference between the array elements on both sides of the aperture, thereby achieving lateral scanning or tilt detection of the kidney and ensuring that the sound beam always enters the target area at a vertical or optimal angle.

[0043] Furthermore, this module utilizes a slip vector field to perform motion compensation on Doppler blood flow monitoring data. In conventional blood flow detection, the relative motion between the probe and the blood vessel generates spurious frequency shift signals (motion artifacts). Since this system knows the relative velocity between the probe and the tissue (i.e., the time derivative of the slip vector), it can subtract this velocity component from the measured Doppler frequency shift during signal processing, thereby outputting a pure, true blood flow velocity that reflects only the actual blood flow.

[0044] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound, characterized in that, It includes patch-type multi-probe ultrasound transducers, arranged in a ring or matrix, used to scan the renal pelvis region from different angles, acquire ultrasound signals from the renal pelvis and renal cortex, and transmit the acquired ultrasound signals to a terminal monitoring platform. The terminal monitoring platform includes: The ultrasonic signal processing module is used to receive ultrasonic signals, perform preprocessing, and output preprocessed ultrasonic signals. The three-dimensional morphology construction module is used to reconstruct a real-time three-dimensional model of the renal pelvis based on the preprocessed ultrasound signal, and to obtain real-time three-dimensional morphological data and volume data of the renal pelvis based on the real-time three-dimensional model of the renal pelvis. The strain detection module is used to analyze the degree of deformation of different regions of the renal pelvis wall under pressure based on pre-processed ultrasound signals, and to obtain renal pelvis wall strain distribution data. The renal cortical blood flow assessment module is used to extract renal cortical blood flow-related parameters based on preprocessed ultrasound signals and obtain renal cortical blood flow assessment data. The renal pelvis pressure estimation module is used to input real-time three-dimensional morphological data, volume data, renal pelvis wall strain distribution data, and the input patient's basic data into a preset mapping model and output the renal pelvis pressure estimation value in real time. The early warning module is used to compare the estimated intrarenal pressure and renal pelvis volume data with the preset intrarenal pressure threshold and renal pelvis volume threshold, respectively, and trigger an early warning signal based on the comparison results; The respiratory shear slip vector decoupling and virtual aperture following module is used to solve the problem of relative displacement of the kidneys due to respiratory movement while the patch is fixed to the skin. The specific execution process is as follows: While acquiring ultrasound signals, the relative motion trajectory between the renal cortex surface and the subcutaneous tissue boundary layer was extracted using ultrasound speckle tracking technology, and the interlaminar shear slip vector field that changes with the respiratory cycle was calculated. Based on the interlayer shear slip vector field, a virtual region of interest coordinate system is constructed that dynamically floats relative to the skin surface. The origin of this coordinate system is always locked at the anatomical center of the renal pelvis. Based on the projection position of the virtual ROI coordinate system in the patch-type multi-probe ultrasound transducer matrix, a dynamic activation strategy is generated in real time: without moving the physical position of the ultrasound transducer, the subset of micro ultrasound transducers in the active state in the matrix is ​​dynamically switched at a millisecond time resolution so that the physical center of the synthesized sound beam is always aligned with the anatomical center of the renal pelvis after the slippage. Based on the modulus and depth components of the interlaminar shear slip vector, the transmission and reception delays of each activated subset of miniature ultrasonic transducers are adjusted, and the focal length of the electron acoustic lens is dynamically adjusted so that the sound field energy focus point falls on the slipped renal pelvis wall. At the same time, dynamic aperture control is performed, following the principle of constant F number. As the focusing depth increases, the aperture range of the activated subset of miniature ultrasonic transducers is automatically expanded to form an electron-suspended aperture that slides relative to the skin surface, ensuring that the renal pelvis imaging is always located at the sound field energy focusing center during the patient's breathing.

2. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 1, characterized in that, The process of acquiring ultrasound signals from the renal pelvis and renal cortex includes: Acquire positioning marker data of patch-type multi-probe ultrasonic transducers, and track the spatial displacement coordinates of ultrasonic transducers in real time based on the positioning marker data; During initial positioning, a reference position parameter is established. The spatial displacement coordinates are compared with the reference position parameter to obtain the displacement deviation value. When the displacement deviation value exceeds the preset allowable range, the calibration process is automatically triggered. During the calibration process, the offset direction and offset amount of the ultrasonic transducer are calculated based on the displacement deviation value, and the ultrasonic transducer is driven to perform adaptive fine adjustment so that the relative position accuracy between the transducer and the renal pelvis region meets the preset accuracy standard. A replaceable water-based medical coupling agent layer is coated on the surface of a medical silicone substrate. The viscosity and thickness parameters of the coupling agent layer meet the acoustic impedance matching requirements between the ultrasound transducer and the skin. The ultrasound transducer is controlled to perform a fan-shaped scan around the central axis of the renal pelvis within a preset angle range, and ultrasound signals from the renal pelvis and renal cortex are acquired synchronously at a preset frequency. The ultrasonic signal undergoes preliminary noise reduction processing to remove environmental electromagnetic interference signals, and the ultrasonic signal after preliminary noise reduction processing is transmitted to the terminal monitoring platform.

3. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 2, characterized in that, The process of acquiring real-time three-dimensional morphological and volume data of the renal pelvis includes: The preprocessed ultrasound signal and the real-time positioning data of the patch-type multi-probe ultrasound transducer were acquired to construct an initial three-dimensional model of the renal pelvis. At the same time, the scanning angle deviation of each ultrasound transducer was calculated based on the real-time positioning data. The correction coefficient weights were adjusted according to the signal-to-noise ratio of each transducer to correct the angle deviation of the initial three-dimensional model of the renal pelvis. The corrected initial 3D model of the renal pelvis is segmented into regions. The grayscale threshold and texture complexity features of the renal pelvis wall tissue and the fluid region in the renal pelvis are extracted. The pure renal pelvis region is segmented from the renal cortex and renal medulla, and the pure renal pelvis region is output. The renal pelvis volume is calculated based on the pure renal pelvis region, and the volume calculation results are corrected by taking into account the dynamic changes in the renal pelvis morphology. The key dimensions of the renal pelvis calculated in real time are compared with historical baseline parameters to obtain the parameter deviation value. When the parameter deviation value exceeds the preset accuracy range, the model is reconstructed. If the deviation values ​​of all key dimensional parameters are less than the preset accuracy range, the model data is deemed reliable, and real-time three-dimensional morphological data of the renal pelvis and the corrected renal pelvis volume data are generated.

4. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 3, characterized in that, The strain detection module acquires strain distribution data of the renal pelvis wall, specifically including: Based on the preprocessed ultrasound signal, the real-time positioning data of the ultrasound transducer is retrieved, the displacement information of different positions of the renal pelvis wall is extracted, and the displacement information is corrected based on the positioning deviation correction parameter. Based on the corrected displacement information, and combined with the pure renal pelvis region, different tissue characteristic regions are distinguished. Based on the ultrasound signal reflection characteristics of each region, the tensile strain value and compressive strain value of the corresponding region are calculated respectively. Spatial mapping of tensile and compressive strain values ​​at different locations is performed and correlated with real-time three-dimensional morphological data of the renal pelvis to generate a strain distribution map of the renal pelvis wall. The strain distribution map of the renal pelvis wall is converted into digital strain distribution data, including strain values, regional coordinates and strain type, and output to the renal pelvis pressure estimation module based on the reliability indicator of the ultrasound signal-to-noise ratio.

5. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 4, characterized in that, The displacement information is corrected based on the positioning deviation correction parameters, specifically including: The positioning marker data of the ultrasonic transducer at the moment of ultrasonic signal acquisition is extracted in real time to determine the spatial displacement and tilt angle of the ultrasonic transducer, establish a one-to-one correspondence between the positioning deviation at that moment and the renal pelvis wall displacement information acquisition channel, and clarify the degree of influence of the positioning deviation on each acquisition channel. Based on the correspondence between positioning deviation and renal pelvis wall displacement information acquisition channels, and combined with the contour features of the pure renal pelvis region, a differential correction coefficient is assigned to each displacement information acquisition channel. The displacement information is split according to the acquisition channel. The displacement information corresponding to each acquisition channel is initially compensated by the correction coefficient. Based on the characteristic that the displacement change trend of adjacent channels in the same renal pelvis wall region is consistent, the compensated displacement information is corrected. The corrected displacement information is compared with the contour change trend of the real-time three-dimensional morphological data of the renal pelvis to generate a trend consistency deviation value. If the trend consistency deviation value exceeds the preset logic range, the correction coefficient is readjusted until the deviation meets the logic consistency requirements, and the corrected displacement information is output.

6. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 1, characterized in that, The renal pelvis pressure estimation module specifically includes: Simultaneously receive real-time three-dimensional morphological data and volume data of the renal pelvis, digital strain distribution data, and input patient individual basic data, and perform consistency verification on the received data; Based on the validated data, the multi-dimensional input vector of the preset mapping model is constructed, with renal pelvis volume data as the basic dimension, strain values ​​of different regions of the renal pelvis wall as the sensitive dimension, patient individual basic data as the adaptation dimension, and blood flow velocity change trend output by the renal cortical blood flow assessment module as an auxiliary reference dimension. Differentiated strain and pressure mapping weights were assigned to different regions of the renal pelvis, and the local pressure estimates for each region were calculated by combining volume data and individual adaptation dimensions. The estimated local pressure values ​​for each region are corrected by using auxiliary reference dimensions; Based on the contribution of each region to the conduction of intrarenal pelvis pressure, the weighted average of the corrected local pressure estimates of each region is calculated to generate a global intrarenal pelvis pressure estimate. At the same time, based on the reliability of the multi-dimensional input vector and the correction magnitude, an estimation credibility score is added and output to the early warning module.

7. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 6, characterized in that, Based on the contribution of each region to intrapelvic pressure conduction, a weighted average is calculated on the corrected local pressure estimates for each region to generate a global intrapelvic pressure estimate, which includes: By combining real-time three-dimensional morphological data and strain distribution data of the renal pelvis, a contribution quantification model is constructed. The volume ratio of each region of the renal pelvis is used as the basic weight, and the correlation coefficient between the strain value and pressure of the region is superimposed. At the same time, the patient's individual basic data is introduced to dynamically generate the real-time contribution weight of each region. The revised local pressure estimates for each region were stratified and classified, and the renal pelvis was divided into pressure-dominant areas and pressure-conduction areas according to its anatomical function. The local pressure estimates within the same functional area were weighted and averaged to obtain the average pressure of the functional area. Based on the pressure transmission efficiency model, a global integrated weight is assigned to the pressure dominance zone and the pressure transmission zone, and the weight value is dynamically adjusted with the change of renal pelvis volume. The average pressure of each functional area is weighted twice according to the global integration weight to obtain the global intrarenal pelvic pressure estimate. At the same time, the calculation process data of the contribution weight of each area is recorded.

8. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 1, characterized in that, The early warning module also includes: Based on individual patient data and historical monitoring data, a dynamic threshold update mechanism is constructed, and the threshold trigger sensitivity is dynamically adjusted in conjunction with the rate of change of renal pelvis volume and internal pressure. The warning signals are classified according to the degree of exceedance of the comparison results, and the classified warning signals are output in the form of sound and light, and pushed to the medical terminal through remote communication. The warning trigger time, exceedance data and warning level are recorded in real time, and a warning log is generated.

9. The renal pelvis pressure and over-distension monitoring system based on patch-type multi-probe ultrasound as described in claim 6, characterized in that, Before outputting the global intrapelvic pressure estimate, the renal pelvis pressure estimation module performs an acoustic-viscoelastic-thermal coupling dynamic correction step to eliminate estimation errors caused by modulus drift and tissue stress relaxation due to the ultrasonic thermal effect. The dynamic correction step specifically includes: Obtain the global intrarenal pressure estimate and real-time renal pelvis volume data output by the preset mapping model; The time-of-flight deviation of ultrasound signals is used to characterize changes in tissue sound velocity. Combined with the obtained global intrarenal pelvic pressure estimate and real-time renal pelvic volume data, the corrected true intrarenal pelvic pressure value is obtained through the following calculation formula. : in, This is the final corrected true intrarenal pelvic pressure value; The estimated global intrarenal pressure output by the preset mapping model; This is the thermo-acoustic modulus coupling coefficient, used to correct the nonlinear stiffness drift of the tissue caused by temperature changes; This is a preset density constant for the renal pelvis wall tissue; The real-time equivalent speed of sound is calculated based on the flight time of the ultrasound signal within the renal pelvis wall at the current moment; The reference tissue sound velocity during the initial system calibration. This term characterizes the dynamic shift in bulk modulus caused by thermal effects and tissue physical hardening; This provides the real-time volume data of the renal pelvis at the current moment. The initial reference volume of the renal pelvis. Hencky log-true strain characterizing the process of renal pelvis dilation; This is the viscous damping correction factor; The effective viscosity coefficient of renal pelvis tissue; This is the first derivative of the renal pelvis volume with respect to time, representing the rate of change of volume; Based on the corrected actual intrarenal pelvis pressure value, it is output as the final monitoring data to the early warning module for comparison with the preset intrarenal pelvis pressure threshold and to trigger the corresponding early warning signal.

Images