Bladder form measuring method and device, electronic equipment and storage medium

Abstract

The invention discloses a bladder morphology measuring method and device, electronic equipment and a storage medium, and relates to the technical field of ultrasonic noninvasive detection and medical image processing. The method comprises the following steps: extracting an effective echo sequence based on an ultrasonic signal, fitting a bladder form according to spatial arrangement characteristics of three-dimensional discrete point cloud coordinate information, and optimizing fitting parameters by adopting a least square method and a Newton iteration method; performing breathing follow-up compensation processing on the initial geometric model, and dispersing the initial geometric model after breathing compensation obtained after processing into an elastic network model of a topological structure so as to perform local elastic deformation optimization; determining the geometric volume of the bladder according to the elastic correction curved surface by adopting a voxel division method; and performing weighted correction according to the individual correction coefficient and the geometric volume of the bladder to obtain the bladder capacity so as to realize measurement of the bladder form. The method can be applied to bladder capacity calculation of the flexible ultrasonic patch, and the accuracy of capacity calculation is improved.

Description

Technical Field

[0001] This application relates to the field of ultrasound non-invasive detection and medical image processing technology, and in particular to a method and apparatus for measuring bladder morphology, electronic equipment, and storage medium. Background Technology

[0002] Lower urinary tract dysfunction (LUTD) is a general term encompassing various diseases related to the bladder and urethra. Bladder volume measurement is crucial for the early diagnosis and management of LUTD. Currently, bladder catheterization is considered the gold standard for bladder volume measurement. Although this method has high accuracy, it is generally not recommended as a routine examination due to the potentially serious adverse consequences of its invasive procedure. Therefore, bladder ultrasound imaging is widely used clinically to estimate bladder volume. This method is not only fast and non-invasive but also effectively provides relevant information such as residual urine volume.

[0003] Currently, the main devices used for bladder capacity assessment are desktop and handheld bladder scanners. These devices generally have strong performance and image processing capabilities, and relatively complete functional configurations. Under standardized operating conditions, they exhibit good measurement stability and repeatability. However, these desktop and handheld bladder scanners are mostly intermittent measurement devices, making it difficult to continuously monitor bladder capacity during patients' daily activities. Furthermore, measurement results are prone to deviation when there are changes in body position, respiratory follow-up, and motion artifacts, making it difficult to meet the actual needs of wearable, long-term bladder function management.

[0004] Several commercially available wearable devices for bladder function monitoring have been launched, quantifying bladder fullness by acquiring ultrasound signals from the anterior and posterior walls of the bladder. While these devices can alert users when bladder fullness reaches a certain threshold, there is generally a significant discrepancy between the calculated bladder capacity and the actual measured value. Existing bladder capacity calculation methods struggle to meet the real-time monitoring accuracy and stability requirements of wearable devices and often neglect the influence of factors such as bladder positional shifts caused by respiratory movements. Currently, wearable bladder capacity monitoring instruments mostly employ ellipsoidal volume algorithms and empirical formulas for capacity calculation. For example, Li Zaiyuan proposed an improved ellipsoidal volume formula in his master's thesis, "Research and Design of Portable B-mode Ultrasonic Bladder Volume Measurement Instrument" (Chang'an University, 2015). ,in This represents the maximum area of ​​a cross-section of the bladder. This represents the maximum area of ​​a longitudinal section of the bladder. This indicates the depth of ultrasound detection during bladder probing; for example, in the article "Experience in Ultrasound Measurement of Bladder Urine Volume and Catheterization Treatment in Patients with Urinary Retention" (Minimally Invasive Medicine, 2015, 10(2):3), the lengths of the three mutually perpendicular maximum diameters on the maximum transverse and longitudinal sections of the bladder were obtained by ultrasound. , , And use empirical formulas Calculating bladder volume is crucial. However, in actual measurements, significant differences exist in bladder anatomy between individuals, and the bladder contour of the same subject can vary considerably under different urine volumes. Simply using an ellipsoidal volume model or a single empirical formula will inevitably lead to large systematic errors, which are more pronounced with irregular bladder shapes. Furthermore, existing methods for calculating bladder capacity in wearable devices generally do not adequately consider motion artifacts caused by changes in body position and respiratory movements, which significantly increase measurement noise and further reduce the accuracy of bladder capacity calculations. Summary of the Invention

[0005] The purpose of this application is to provide a bladder morphology measurement method and device, electronic equipment, and storage medium, which can be applied to bladder capacity calculation using flexible ultrasound patches to improve the accuracy of capacity calculation.

[0006] To achieve the above objectives, this application provides the following solution: In a first aspect, this application provides a method for measuring bladder morphology, including: Acquiring ultrasound signals; the ultrasound signals are obtained by attaching a flexible ultrasound patch to the abdominal bladder region and sequentially transmitting and receiving ultrasound pulses based on each piezoelectric transducer. Based on the ultrasound signal, the effective echo sequence is extracted, and the three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder is determined. The bladder morphology is fitted based on the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and the fitting parameters are optimized using the least squares method and Newton's iteration method to obtain an initial geometric model that matches the bladder boundary. The initial geometric model is subjected to breathing follow-up compensation processing, and the initial geometric model after breathing compensation is discretized into an elastic network model of topological structure for local elastic deformation optimization to obtain an elastic correction surface. The geometric volume of the bladder is determined based on the elastically modified surface using the voxel division method. Bladder capacity is obtained by weighting and correcting the individual correction coefficient with the geometric volume of the bladder to achieve bladder morphology measurement. The individual correction coefficient is determined based on the individual physiological characteristic parameters of the subject and according to the mapping relationship between the pre-calibrated physiological characteristic parameters and the correction coefficient. The individual correction coefficient is used to compensate for measurement deviations caused by differences in body size and the sound speed propagation characteristics of soft tissue.

[0007] Secondly, this application provides a bladder morphology measuring device, comprising: An ultrasound signal acquisition module is used to acquire ultrasound signals; the ultrasound signals are obtained by attaching a flexible ultrasound patch to the abdominal bladder region and sequentially transmitting and receiving ultrasound pulses based on each piezoelectric transducer. The extraction module is used to extract the effective echo sequence based on the ultrasound signal and determine the three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder. The fitting optimization module is used to fit the bladder morphology based on the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and to optimize the fitting parameters using the least squares method and Newton's iteration method to obtain an initial geometric model that matches the bladder boundary. The deformation optimization module is used to perform breathing follow-up compensation processing on the initial geometric model, and to discretize the breathing-compensated initial geometric model into an elastic network model of topological structure for local elastic deformation optimization to obtain an elastic correction surface. A geometric volume determination module is used to determine the geometric volume of the bladder based on the elastic modified surface using a voxel division method; The correction module is used to obtain bladder capacity by weighting and correcting the individual correction coefficient and the geometric volume of the bladder, so as to realize the measurement of bladder morphology. The individual correction coefficient is determined based on the individual physiological characteristic parameters of the examinee and according to the pre-calibrated mapping relationship between the physiological characteristic parameters and the correction coefficient. The individual correction coefficient is used to compensate for the measurement deviation caused by differences in body size and the sound speed propagation characteristics of soft tissue.

[0008] Thirdly, this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the bladder morphology measurement method described above.

[0009] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the bladder morphology measurement method described above.

[0010] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application provides a method, device, electronic equipment, and storage medium for measuring bladder morphology. It can be applied to bladder capacity calculation using flexible ultrasound patches. Based on the extraction of effective echo sequences from ultrasound signals, the method optimizes the fitting parameters for bladder morphology fitting using the least squares method and Newton's iteration method, based on the spatial arrangement characteristics of three-dimensional discrete point cloud coordinates. Respiratory follow-up compensation is introduced, discretizing the initial geometric model after respiratory compensation into a topological elastic network model for local elastic deformation optimization. The geometric volume of the bladder is determined based on the elastic correction surface using a voxel partitioning method. Then, the bladder capacity is obtained by weighted correction based on individual correction coefficients, thus achieving bladder morphology measurement. This application improves the accuracy of bladder capacity calculation through bladder morphology fitting, fitting parameter optimization, respiratory follow-up compensation, and local elastic deformation optimization, and by weighted correction based on individual correction coefficients. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart of the bladder morphology measurement method; Figure 2 A schematic diagram of the application of a flexible ultrasound patch; Figure 3 This is a schematic diagram illustrating the reflection path of ultrasound signals on the anterior and posterior walls of the bladder and the principle of effective echo acquisition. Figure 4 A schematic diagram illustrating the evolution of bladder morphology under different filling states; Figure 5 This is a schematic diagram illustrating the fitting principle and residual determination based on a spherical model. Figure 6 This is a schematic diagram illustrating the fitting principle and axial parameters based on the ellipsoidal model. Figure 7 This is a schematic diagram illustrating the principle of adaptive selection of a spherical fitting model based on the coefficient of variation. Figure 8 This is a schematic diagram illustrating the principle of adaptive selection of an ellipsoidal fitting model based on the coefficient of variation. Figure 9 A schematic diagram illustrating the effect of respiratory follow-up on bladder position and the principle of respiratory compensation. Figure 10 This is a schematic diagram of the local deformation optimization and voxel mesh volume calculation process based on elastically modified surfaces. Figure 11 This is a schematic diagram of the overall process; Figure 12 This is a structural diagram of a bladder morphology measuring device; Figure 13 This is a schematic diagram of the operational steps for measuring bladder morphology. Detailed Implementation

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0015] In one exemplary embodiment, such as Figure 1 As shown, a method for measuring bladder morphology is provided, including: Step 100: Acquiring ultrasound signals. Ultrasound signals are obtained by attaching a flexible ultrasound patch to the abdominal bladder region and sequentially transmitting and receiving ultrasound pulses based on each piezoelectric transducer.

[0016] Step 200: Extract the effective echo sequence based on the ultrasound signal and determine the three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder.

[0017] This includes extracting the effective echo sequence based on ultrasound signals and determining the three-dimensional discrete point cloud coordinates of the anterior and posterior walls of the bladder, specifically: Based on ultrasound signals, windows are set on the time axis according to the preset anatomical depth range of the anterior and posterior walls of the bladder. Peaks whose timestamps fall within the corresponding window range are marked as candidate wall reflection signals, and noise echoes that exceed the window range are discarded. The windows include the anterior wall time window and the posterior wall time window.

[0018] Peak detection is performed on the waveform envelope within the window, and isolated noise peaks are removed based on neighboring peaks to obtain a valid echo sequence. For each detected local peak, the corresponding timestamp is recorded, and a time neighborhood is set within the range of the timestamp. The existence of another local peak within the time neighborhood is counted. If at least one accompanying peak exists, it is marked as a valid echo peak and retained as a candidate point. If it is an isolated peak within the time neighborhood, it is considered an isolated noise peak caused by random noise, electronic interference, or transient artifacts and is deleted.

[0019] Each candidate point is converted into a candidate reflection point in spatial coordinate form, and the one-way propagation distance is determined based on the timestamp and the preset sound speed to determine the three-dimensional coordinates of each candidate reflection point, thereby obtaining the three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder.

[0020] Step 300: Fit the bladder morphology based on the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and optimize the fitting parameters using the least squares method and Newton's iteration method to obtain an initial geometric model that matches the bladder boundary.

[0021] In one embodiment, the bladder morphology is fitted based on the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and the fitting parameters are optimized using the least squares method and Newton's iteration method to obtain an initial geometric model that matches the bladder boundary, specifically including: Based on the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, a sphere or ellipsoid is adaptively selected for bladder morphology fitting; among which, the depth coordinate difference between the anterior and posterior wall points of the bladder in the normal direction is used based on the detection direction of each piezoelectric transducer. Construct a set of depth differences and calculate the arithmetic mean of the set of depth differences. and standard deviation To construct the coefficient of variation As an indicator of dispersion; ;when If the bladder shape is below the preset threshold, it is determined that the bladder shape is close to a regular sphere, and a sphere is selected as the fitting model for bladder shape fitting; otherwise, an ellipsoid is selected as the fitting model for bladder shape fitting.

[0022] The geometric equations are used to optimize the fitting parameters, with the center coordinates and radius of the sphere, or the center and principal axis length of the ellipsoid. The objective function is to minimize the sum of squared radial distance residuals from the coordinate information of the three-dimensional discrete point cloud to the surface of the fitting model.

[0023] Based on the geometric equations and the objective function, the initial estimates of the fitting parameters are obtained using the least squares method. The radial residuals of the coordinate information of each three-dimensional discrete point cloud are calculated and outliers exceeding the preset geometric threshold are removed. The remaining effective point cloud is updated nonlinearly using the Newton-Raphson iteration method until the preset convergence condition is met, thus obtaining the initial geometric model that matches the bladder boundary.

[0024] Step 400: Perform breathing motion compensation on the initial geometric model, and discretize the initial geometric model after breathing compensation into an elastic network model of topological structure for local elastic deformation optimization to obtain an elastic correction surface.

[0025] In one embodiment, the initial geometric model is subjected to breathing-based compensation processing, and the resulting breathing-compensated initial geometric model is discretized into an elastic network model of topological structure for local elastic deformation optimization to obtain an elastically corrected surface, specifically including: Under continuous sampling conditions, the geometric center coordinates of the initial geometric model at each sampling time are taken as the time series; the average respiratory position of the time series is determined by the time window averaging method or trend extraction method, and the instantaneous displacement vector of the geometric center coordinates at the current time relative to the average respiratory position is determined.

[0026] The point cloud of the anterior and posterior walls of the bladder is compensated by reverse translation using instantaneous displacement vectors. The calculation reference is corrected to the average respiratory position to obtain the initial geometric model after respiratory compensation. The initial geometric model after respiratory compensation is discretized into an elastic mesh model with topological structure. The elastic mesh model is given the physical property of finite stiffness. The radial normal residual from the point cloud of the anterior and posterior walls of the bladder to the surface of the elastic mesh model is determined and a deformation judgment threshold is set. Local elastic deformation optimization is performed based on the radial normal residual to obtain the elastic correction surface.

[0027] Specifically, local elastic deformation optimization is performed based on the radial normal residual to obtain an elastically corrected surface, which includes: When the point cloud of the anterior and posterior walls of the bladder is located outside the elastic mesh model and the radial normal residual is greater than the deformation judgment threshold, the mesh nodes in the corresponding region are driven to expand outward for fitting. When the point cloud of the anterior and posterior walls of the bladder is located inside the elastic mesh model and the absolute value of the radial normal residual is greater than the deformation judgment threshold, the mesh nodes in the corresponding region are driven to contract inward for fitting. When the absolute value of the radial normal residual is less than or equal to the deformation judgment threshold, or in a region not covered by the point cloud of the anterior and posterior walls of the bladder, the physical property of finite stiffness is used to maintain the continuity and smoothness of the mesh surface of the elastic mesh model, thereby generating a closed elastic modified surface.

[0028] Step 500: Determine the geometric volume of the bladder based on the elastically modified surface using the voxel division method.

[0029] In one embodiment, the geometric volume of the bladder is determined based on an elastically modified surface using a voxel partitioning method, specifically including: A three-dimensional coordinate system is established within the space enclosed by an elastically modified surface, and the space is discretized into multiple units with constant unit volume according to a preset voxel size. A voxel grid composed of regular voxels; the voxel size is determined by the voxel spatial resolution.

[0030] Traverse all voxel meshes, extract the geometric center coordinates of each voxel, and use a ray casting algorithm or a signed distance field algorithm to determine the internal and external topological relationships of the geometric center coordinates of each voxel relative to the elastic modified surface, and count the number of voxels determined to be located inside the elastic modified surface. ,according to The geometric volume of the bladder was calculated. .

[0031] Step 600: The bladder capacity is obtained by weighting and correcting the individual correction coefficients based on the geometric volume of the bladder, thereby achieving the measurement of bladder morphology. The individual correction coefficients are determined based on the individual physiological characteristic parameters of the subject, according to the pre-calibrated mapping relationship between the physiological characteristic parameters and the correction coefficients; the individual correction coefficients are used to compensate for measurement deviations caused by differences in body size and the sound propagation characteristics of soft tissue.

[0032] The expression corresponding to bladder capacity is: .

[0033] in, Let be the geometric volume of the bladder; Individual correction coefficient; This refers to bladder capacity.

[0034] This application acquires effective echoes using a flexible ultrasound patch and calculates the Z-axis one-way distance to generate three-dimensional point clouds of the anterior and posterior walls. Based on the spatial characteristics of the point cloud, it adaptively selects a spherical or ellipsoidal fitting model, and uses the least squares method and Newton's method to determine the model parameters. Respiratory follow-up compensation is introduced to eliminate motion artifacts, and the fitted model is treated as an elastic surface for local deformation correction. Finally, voxel partitioning technology is used to accumulate and calculate the geometric volume, and combined with individualized coefficients, the final bladder capacity is obtained. This application effectively solves the problems of poor adaptability of single models and dynamic interference, improving the measurement accuracy.

[0035] A flexible ultrasound patch is attached to the abdominal bladder region. Each piezoelectric transducer sequentially emits and receives ultrasound pulses to acquire effective echoes, determining three-dimensional discrete cloud point data for the anterior and posterior walls. Adaptive spherical or ellipsoidal fitting is performed, and respiratory tracking error compensation is applied. The fitted model surface is used as an elastic correction surface for local deformation optimization. The bladder's geometric volume is calculated through voxel division, and individual correction coefficients are introduced to calculate bladder capacity. Specific operational steps are as follows: Figure 13 As shown.

[0036] S1: Parameter Calibration: Pre-calibrate the parameters related to bladder capacity calculation. These parameters include the equivalent speed of sound for time-distance conversion, the dispersion threshold for sphere / ellipsoid model selection, the residual threshold for elastic surface correction, and individualized correction coefficients. wait.

[0037] S2: Bladder capacity calculation: Continuous monitoring of the bladder using a flexible ultrasound patch, specifically including: Effective echoes from the anterior and posterior walls of the bladder were acquired to determine three-dimensional discrete point cloud data. Based on the spatial characteristics of the point cloud, a spherical or ellipsoidal shape was selected for fitting, and the fitting model parameters were determined using the least squares method and Newton's iteration method. Respiratory follow-up analysis and compensation were performed to obtain the initial spherical or ellipsoidal surface after respiratory compensation. Local deformation optimization was performed to obtain an elastic correction surface, and voxel partitioning was used to calculate the geometric volume of the bladder. The bladder capacity is obtained by converting units using an individual correction factor.

[0038] The flexible ultrasound patch comprises a flexible PCB board, piezoelectric ceramics, an epoxy matching layer, a hydrogel capable of long-term adhesion, an epoxy backing plate, and a microcontroller circuit system. It can be worn for extended periods to acquire and monitor bladder capacity data in real time. The piezoelectric ceramics employ a flanged electrode structure and are soldered to the flexible PCB board. The flexible PCB board has through-holes with a diameter slightly smaller than the piezoelectric ceramics at their corresponding locations. Its back side is adhered to the epoxy backing plate using double-sided adhesive of a certain thickness, suspending the back of the piezoelectric ceramics and forming an air backing layer. This makes it difficult for back-facing acoustic energy to be conducted to the backing plate, which helps improve the transducer's forward radiation efficiency and receiving sensitivity.

[0039] The raw ultrasound echo signal acquired by the flexible ultrasound patch is preprocessed, including: setting an anterior wall time window and a posterior wall time window on the time axis according to the preset anatomical depth range of the anterior and posterior walls of the bladder, respectively; performing peak detection on the echo envelope; identifying and removing isolated noise peaks by judging adjacent peaks to obtain effective echoes; calculating the one-way distance according to the preset sound velocity; and determining the three-dimensional coordinates of each reflection point to obtain the three-dimensional discrete point cloud coordinates of the anterior and posterior walls of the bladder.

[0040] Based on the spatial arrangement characteristics of the point clouds of the anterior and posterior walls of the bladder, an adaptive selection of spherical or ellipsoidal fitting steps is performed. The depth coordinate difference between the anterior and posterior wall points in the normal direction along the detection direction of each piezoelectric transducer is calculated. Construct a set of depth differences. Calculate the arithmetic mean of the depth difference set. and standard deviation Construct the coefficient of variation As a measure of dispersion, among which This is used to characterize the degree of normalized deformation of the bladder morphology relative to an equivalent ideal sphere. When If the bladder shape is below the preset threshold, it is determined that the bladder shape is close to a regular sphere, and the spherical equation is selected as the fitting model; otherwise, the ellipsoidal equation is selected as the fitting model.

[0041] Construct a geometric equation with the coordinates and radius of the sphere's center, or the center and principal axis length of the ellipsoid, as the parameters to be optimized; establish the sum of squared radial distance residuals from the bladder point cloud to the model surface as the objective function to be minimized; use the least squares method to obtain the initial estimates of the parameters, calculate the radial residuals of each point cloud and remove outliers that exceed the preset geometric threshold, and perform nonlinear optimization updates on the remaining effective point cloud using the Newton-Raphson iteration method until the preset convergence condition is met, thereby obtaining the initial geometric model that best matches the bladder boundary.

[0042] To reduce the calculation error caused by respiratory motion, respiratory follow-up compensation is introduced. Under continuous sampling conditions, the geometric center coordinates of the initial geometric model at each sampling time are used as a time series. The average respiratory position of the time series is calculated using the time window averaging method or trend extraction method, and the instantaneous displacement vector of the geometric center at the current time relative to the average respiratory position is calculated. The displacement vector is used to perform reverse translation compensation on the point cloud of the anterior and posterior walls of the bladder, and the calculation benchmark is corrected to the average respiratory position.

[0043] The steps for local deformation optimization based on elastic modified surfaces specifically include: discretizing the initial geometric model after breathing compensation into an elastic mesh model with a topological structure, and assigning the mesh model the physical property of finite stiffness; calculating the radial normal residual from the bladder point cloud to the surface of the elastic mesh model, and setting a deformation judgment threshold; performing local elastic deformation according to the residual: when the point cloud is located outside the mesh model and the residual is greater than the threshold, driving the mesh nodes in the corresponding region to expand outward to fit the point; when the point cloud is located inside the mesh model and the absolute value of the residual is greater than the threshold, driving the mesh nodes in the corresponding region to contract inward to fit the point; when the absolute value of the residual is less than or equal to the threshold, or in a region without point cloud coverage, using the physical property of finite stiffness to maintain the continuity and smoothness of the mesh surface, thereby generating a closed elastic modified surface.

[0044] Calculation of bladder geometric volume based on elastic modified surface The steps specifically include: constructing a minimum bounding box that completely encloses the closed elastic modified surface, and setting the voxel spatial resolution; based on the spatial resolution, discretizing the three-dimensional bounding box into several components with constant unit volumes. A voxel mesh composed of regular voxels is formed; the voxel mesh is traversed, and the geometric center coordinates of each voxel are extracted. The topological relationship between the geometric center coordinates and the closed elastic modified surface is determined using a ray casting algorithm or a signed distance field algorithm; the number of effective voxels located inside the elastic modified surface is counted. And according to the formula The geometric volume of the bladder is calculated.

[0045] The steps for calculating bladder capacity using individualized correction factors specifically include: obtaining the subject's individual physiological characteristics, including at least one of height, weight, body mass index (BMI), and abdominal wall thickness; determining an individualized correction factor that matches the subject based on a pre-calibrated parameter-coefficient mapping relationship or empirical formula to compensate for systematic measurement biases caused by differences in body size and the sound propagation characteristics of soft tissue; and using the formula... For geometric volume After weighted adjustment, the final bladder capacity is obtained. .

[0046] In practical applications, the flexible ultrasound patch in this application includes a flexible PCB board, piezoelectric ceramics, an epoxy matching layer, a hydrogel that can adhere for a long time, an epoxy backing plate, and a circuit system of a microcontroller. The patch can be attached to the body surface for a long time and worn on the body, thereby realizing real-time online acquisition of bladder capacity.

[0047] The flexible ultrasonic patch contains multiple piezoelectric ceramic elements, which are arranged at corresponding positions in the patch's planar coordinate system. Each element acts as an independent transducer for transmitting and receiving ultrasonic signals, and records the flight time of the corresponding echo, with the flight time measured in microseconds.

[0048] The bladder capacity calculation method includes data preprocessing of the raw ultrasound signals acquired by the flexible ultrasound patch. Specifically, four piezoelectric transducers on the flexible ultrasound patch emit ultrasound pulses in groups according to a preset emission cycle, and simultaneously acquire the echo signals of each channel, forming a multi-channel raw echo data sequence with timestamps. Based on the propagation speed of ultrasound in soft tissue and the anatomical distance range between the anterior and posterior walls of the bladder and the patch, time windows for anterior and posterior wall echoes are preset, dividing the time axis into anterior wall candidate time windows and posterior wall candidate time windows, and searching for effective echoes from the bladder wall only within the corresponding time windows.

[0049] The round-trip flight time of the ultrasound echoes was recorded, and the acquired raw data was imported into a data table. The first column of the table contained the piezoelectric transducer number, and the second column contained the corresponding timestamp. The timestamps grouped by piezoelectric transducer number were then extracted into an index structure to ensure that the original peak sequences of each X-ray were independent. Based on the anatomical geometry of the human body, two time windows were set for the anterior and posterior walls of the bladder, with the anterior wall time window being... The time window for the back wall is Only peaks whose timestamps fall within the corresponding window range are marked as candidate wall reflection signals, while noise echoes and premature or late reflections that exceed the window range are directly rejected.

[0050] After determining the candidate time windows for the front and rear walls, peak detection is performed on the waveform envelope within each window. For each detected local peak, its corresponding timestamp is recorded, and a narrow time neighborhood is set around that timestamp, for example, centered on that timestamp and extending outwards and inwards. Within a given time range, we statistically analyze whether another local peak exists within that time neighborhood. When a peak exists within its... If at least one accompanying peak exists within the time range, that peak is marked as a valid echo peak and retained as a candidate point for subsequent time-of-flight calculations. If the peak is an isolated peak within the aforementioned time neighborhood (i.e., there is no second peak), it is considered an isolated noise peak caused by random noise, electronic interference, or transient artifacts and is deleted. The entire process relies solely on constant windows, proximity judgment, and averaging operations, without interpolation or complex filtering, thus preserving the original geometric information to the maximum extent while eliminating significant noise.

[0051] In this embodiment, after extracting the effective echo peaks of the front and rear walls, it is necessary to convert the flight time of each effective echo into candidate reflection points in spatial coordinates. Specifically, a three-dimensional Cartesian coordinate system is pre-established locally on the patch, with the patch plane as the reference plane. Plane, normal outer side In the axial direction, the spatial position of each piezoelectric transducer in this coordinate system is determined by the patch design dimensions and arrangement, and is denoted as... Among them, the most commonly used are .

[0052] Timestamp Using microseconds ( Equivalent speed of sound (in units of ) The sound velocity of water / urine after calibration Then it can be used directly. Converting flight time into one-way propagation distance in millimeters, where, For the first Timestamp of each valid echo peak (unit) ),coefficient Used to complete unit conversion ( Dividing by 2 represents converting the round-trip path length into the one-way propagation distance, resulting in the final value. For from the first The spatial distance from each transducer unit to the corresponding bladder wall reflex point is expressed in millimeters.

[0053] Its planar position is the same as the first Each transducer has the same coordinates in the patch plane, that is... The depth coordinates are taken as the distance calculated from the flight time. ,Right now Through this time-distance-coordinate conversion process, each received echo peak can be mapped to a three-dimensional point in the patch coordinate system. This results in discrete point cloud data representing the bladder boundary.

[0054] Based on spatial arrangement characteristics, an adaptive selection of spherical or ellipsoidal fitting is made. Specifically, taking the plane where the piezoelectric ceramic transducer is located as the reference coordinate system, the current shape is determined to be near-spherical or ellipsoidal based on the three-dimensional coordinate relationship of the front and rear wall point clouds obtained through preprocessing.

[0055] In one implementation, this application employs adaptive decision-making logic based on ray depth difference. Specifically, instead of pre-estimating the sphere's center position, it directly calculates the depth difference (i.e., the bladder's straight-line thickness in each detection direction) between the anterior and posterior wall points of the bladder based on the detection paths of each piezoelectric transducer, constructing a set of depth differences. The arithmetic mean and standard deviation of this set are calculated, and the coefficient of variation is further determined. This index serves as a normalized measure of the relative dispersion of bladder morphology. It effectively eliminates the interference of absolute bladder capacity on morphological assessment. When the calculated coefficient of variation is below a preset sphericity threshold... When the thickness distribution of the bladder is relatively uniform in all detection directions, and its morphological characteristics conform to a near-spherical shape, the system selects a spherical model as the fitting model for the bladder boundary; otherwise, it is determined to be an irregular shape, and an ellipsoid is selected. Figure 5 This is a schematic diagram illustrating the fitting principle and residual determination based on a spherical model. Figure 6 This is a schematic diagram illustrating the fitting principle and axial parameters based on the ellipsoidal model.

[0056] The fitting model using a sphere as the bladder boundary is explained below. The calculation process is basically the same when using an ellipsoid. Unless otherwise explicitly stated, it can be applied by analogy. Any differences will be explained later.

[0057] Once the current bladder shape is determined to be near-spherical based on the spatial arrangement characteristics of the point clouds of the anterior and posterior walls, and a sphere is selected as the fitting model for the bladder boundary, the bladder capacity can be calculated based on spherical fitting, elastic surface correction, etc.

[0058] After obtaining preliminary estimates of the model parameters, a secondary denoising process based on spatial geometric features is further performed. Specifically, the normal radial residuals of all 3D discrete point clouds to the surface of the fitted model are calculated. (That is, the shortest Euclidean distance from each sampling point to the fitted surface). Simultaneously, a geometric decision threshold is set. The system iterates through all point clouds to determine: if the radial residual of a certain point... If a point is found to be out of place, it is determined to be a noisy point cloud or an outlier that does not conform to the continuous boundary characteristics of the bladder. The system removes these outliers from the dataset so that they no longer participate in subsequent parameter iterations, thereby effectively eliminating shape distortions caused by local artifacts of the bladder wall or reflections from non-target tissues.

[0059] A spherical equation is constructed from the bladder boundary point cloud, using the coordinates of the center of the sphere to be determined and the radius as parameters. The geometric distance from each boundary point to the candidate sphere is used as the error. A least squares criterion is introduced to minimize the sum of squared radial errors from all boundary points to the sphere, thereby optimizing the solution for the center of the sphere. and radius Simultaneously, physical constraints are imposed on the center position and radius of the ellipsoid based on human anatomical dimensions to ensure the rationality and stability of the fitting results. When selecting ellipsoidal fitting, a "center + principal axis" parameter form is adopted. Let the center of the ellipsoid be... The lengths of the three main shafts are respectively , , .

[0060] By linearizing the equations of the sphere, the solution process can be transformed into a system of linear algebraic equations, allowing the initial sphere fitting parameters for the center and radius to be obtained using conventional numerical methods. Specifically, the mean of the point cloud coordinates can be used as the initial value for the center of the sphere, and the initial value for the radius can be set to... The fitted radius It can be directly used for preliminary estimation of bladder volume; the volume calculation formula is as follows: Here, when choosing an ellipsoidal fit, the volume calculation formula is: (in The unit is , The unit is mL), and the output sphere center coordinates can be used as the positioning result of the bladder center in the sensor coordinate system.

[0061] The coordinates of the sphere's center are obtained using the least squares method. and radius After initial estimation, Newton's iteration method is used to perform nonlinear optimization of the spherical parameters. Based on the obtained spherical model, the spherical parameters are denoted as vectors. The first point in the cloud on the anterior and posterior walls of the bladder. Points Calculate the radial residual from this point to the sphere. and residual sum of squares In the current parameter estimation At each point, calculate the first derivative of each residual with respect to the parameter to form the Jacobian matrix. and will As an approximation of the Hessian matrix of the objective function, a Newton-style iterative update formula is constructed. The parameter vector is then corrected to obtain new parameter estimates. Through multiple iterations, the iteration terminates when the change in the objective function or the parameter increment between two adjacent iterations falls below a preset convergence threshold. The resulting coordinates of the sphere's center and radius can then be used as an initial spherical geometric model that more closely matches the point cloud of the anterior and posterior walls of the bladder.

[0062] To address the overall displacement and deformation of the bladder under conditions of free breathing and slight body movement, and to reduce the impact of respiratory motion on bladder capacity calculations, respiratory motion analysis and compensation are introduced after initial fitting of the bladder boundary. Specifically, under continuous sampling conditions, the system performs spherical fitting of the bladder boundary point cloud over multiple adjacent sampling periods to obtain the bladder geometric center coordinates at each sampling time. Considering that bladder filling is a slow process, while bladder displacement and volume fluctuations caused by respiration exhibit periodic characteristics, the system analyzes the time series... Perform frequency domain or time domain decomposition to extract the slowly changing trend term and the fast oscillation term.

[0063] Time series of bladder geometric center coordinates Perform trend extraction or moving average processing based on time windows to obtain a smooth center trajectory after removing the respiratory frequency band. This is considered as the "average respiratory position" of the bladder in the patch coordinate system; the difference between the original center coordinates and the smoothed trajectory is... This is considered as the instantaneous displacement caused by respiratory motion. For the boundary point cloud at any sampling time, the three-dimensional coordinates of all points can be adjusted before fitting. The direction is translated and transformed to perform spherical or ellipsoidal fitting and elastic surface correction in the coordinate system after respiratory compensation, so that the fitting process corresponds as closely as possible to the average position state of the bladder.

[0064] The method of introducing respiratory follow-up compensation can identify and reduce the impact of respiratory follow-up on bladder geometry fitting and capacity estimation by relying solely on continuously acquired bladder ultrasound point clouds without adding additional hardware sensors, thereby maintaining high measurement stability and repeatability under free breathing conditions.

[0065] To improve adaptability to local deformation and irregular contours, an elastic surface correction mechanism is introduced. The sphere is considered an elastic surface with finite stiffness. With the spherical normal as the deformation direction, several control nodes or sampling points are arranged in the spherical parameter space. A radial displacement field is used to induce minor deformations in the reference sphere. An objective function is constructed, comprising a data fitting term and a smoothness constraint term. The data fitting term measures the distance error between the corrected surface and the actual boundary points, while the smoothness constraint term limits the displacement and curvature changes of adjacent control nodes, ensuring the continuity and smoothness of the corrected surface. By solving for the minimum value of this objective function, the radial correction amount of each control node is obtained and mapped onto the entire sphere, forming an elastic corrected surface that more closely conforms to the actual bladder boundary.

[0066] A three-dimensional coordinate system is established within the space enclosed by an elastically modified surface. The space is discretized into a regular voxel mesh according to a preset voxel size, resulting in several voxel units with defined volumes. For each voxel center point, it is determined whether it is located inside the bladder. This can be achieved through surface function notation or by projecting the center point onto a local normal and comparing it with the corresponding local radius. The volumetric volume of the voxel units located inside the bladder is then summed to obtain the geometric volume of the bladder. For voxels located near the bladder boundary, the volume fraction can be weighted and corrected based on their relative position to the boundary surface, or a finer local voxel resolution can be used in the boundary region to reduce the volume approximation error caused by voxel discretization.

[0067] Considering the differences in body shape characteristics (e.g., height, weight, body mass index, abdominal wall thickness) among different subjects, the volume calculated based on the spherical geometric model is prone to systematic bias. Therefore, this application introduces an individual correction factor based on the initial estimated volume. The corrected bladder volume was obtained. Among them, the spherical shape represents the initial fitted model, and the corrected bladder volume represents the final shape. When the ellipsoid is the initial fitted model, it is... .

[0068] Individual correction coefficient It can be determined based on the subject's body shape parameters. For example, during the system initialization phase, body surface measurement data such as the subject's height, weight, waist circumference or abdominal wall thickness are collected, the body mass index (BMI) is calculated, and the subject is divided into a preset body shape classification interval. A set or range of correction coefficient values ​​is pre-calibrated for each body shape classification.

[0069] Clinically calibrated The value of is usually located in The scope is within the range, but this application is not limited to that specific scope.

[0070] The bladder's geometric volume is calculated by fitting an adaptive sphere or ellipsoid, compensating for respiratory motion errors, and using the fitted model surface as an elastic correction surface for local deformation optimization. And introduce individual correction coefficients The calculation method, under the conditions of limited number of sensors and relatively sparse point cloud, can better characterize the three-dimensional contour of near-spherical or ellipsoidal bladders under different filling states, and achieve high-precision estimation of bladder capacity.

[0071] This application uses a wearable flexible ultrasound patch as the data acquisition front end. For example... Figure 2 As shown, the ultrasonic patch employs a multi-layered composite structure, comprising, from the side in contact with the human body outwards: a hydrogel layer, an acoustic impedance matching layer, a piezoelectric ceramic array, a flexible PCB board, and an epoxy-based backing plate. The hydrogel layer is used to achieve acoustic coupling and long-term adhesion between the patch and the abdominal skin; the acoustic impedance matching layer is attached to the forward radiating surface of the piezoelectric ceramic, and its thickness is configured to be one-quarter wavelength corresponding to the center frequency to improve transmission efficiency.

[0072] In terms of structural connection, the piezoelectric ceramics are welded to the flexible PCB board via a flanged electrode structure. To optimize signal quality, the flexible PCB board has through holes at the locations corresponding to each piezoelectric ceramic. An epoxy-based backing plate is bonded to the back side of the flexible PCB board, leaving the back of the piezoelectric ceramics suspended, thus forming an air backing layer between the piezoelectric ceramics and the epoxy-based backing plate. This air backing structure effectively prevents sound energy from being conducted backward by utilizing the high acoustic impedance characteristics of air, significantly improving the forward radiation efficiency and receiving sensitivity of the transducer.

[0073] like Figure 2 and Figure 3 As shown in the working principle, when in use, the patch is attached to the abdominal area above the pubic symphysis. Each piezoelectric ceramic acts as an independent transducer to emit ultrasonic pulses in sequence and receive effective echo signals reflected from the anterior and posterior walls of the bladder for subsequent capacity measurement.

[0074] like Figure 11 As shown, the method specifically includes: parameter calibration, signal acquisition and preprocessing, adaptive model fitting, respiratory compensation, elastic surface correction, and volume calculation. In this embodiment, the system controls multiple piezoelectric ceramic transducers on the flexible ultrasonic patch to emit ultrasonic pulses at a preset period, such as... Figure 3 As shown, ultrasound waves penetrate the abdominal wall and enter the bladder, where they are reflected at the anterior and posterior walls. To extract effective 3D point cloud data from the raw radiofrequency signal, preprocessing specifically includes: Time window gating. Based on human anatomical characteristics, anterior wall time windows and posterior wall time windows are preset on the time axis. Preferably, the anterior wall time window is set to... The back wall time window is set to The method only retains echo signals that fall within the corresponding window, directly eliminating premature reverberation or late noise outside the window.

[0075] Peak detection and noise suppression. Peak detection is performed on the echo envelope within the time window. A neighboring peak determination strategy is adopted; if a detected peak is within the neighboring peak's range... If an isolated peak without an accompanying peak is found in the time neighborhood, it is considered random noise and is removed to ensure that the actual effective echo from the wall is obtained.

[0076] Spatial coordinate mapping. Recording the flight time of each valid echo. (Unit is) According to the formula Calculate the one-way distance from the transducer to the reflection point, where, The calibrated equivalent speed of sound (e.g.) (Based on the known positions of each transducer in the patch coordinate system) The one-dimensional time series is transformed into three-dimensional discrete point cloud coordinates representing the positions of the anterior and posterior walls of the bladder. .

[0077] like Figure 4 As shown, the morphological characteristics of the bladder during filling process evolve from an irregular shape to a near-ellipsoidal and spherical shape. This embodiment abandons a single fixed model and adopts an adaptive spherical and ellipsoidal optimization strategy. Figure 4 A to Figure 4 G in the diagram illustrates the change from an irregular shape during the contraction phase to a near-spherical / ellipsoidal shape during the filling phase. For example... Figure 7 and Figure 8 As shown, an adaptive decision-making strategy based on the coefficient of variation.

[0078] like Figure 7 and Figure 8 As shown, this application establishes a spatiotemporal index mapping mechanism at the signal acquisition source. First, a unique mapping relationship is established between each piezoelectric transducer on the flexible ultrasonic patch and its spatial position. The spatial positions of the transducers on the patch are pre-defined. Each piezoelectric transducer is physically coded and sorted, and the local coordinates of the patch corresponding to each number are stored. .

[0079] The control unit sequentially excites each piezoelectric transducer according to a preset hardware scan sequence. For the first... Each transducer receives an echo signal, and the system directly assigns it a unique data label. This label remains unchanged in all subsequent processing steps (preprocessing, fitting, compensation), thus ensuring that each discrete point cloud data can be accurately traced back to its physical emission source.

[0080] Establish a reference frame. Based on the acquired point clouds of the anterior and posterior walls, calculate the normal linear distance between the incident point of the anterior wall of the bladder and the reflection point of the posterior wall along each ray direction. (Right now Figure 7 and Figure 8 The Chinese logo to (etc.). Calculate the arithmetic mean of all ray distances. Construct a diameter of The equivalent ideal sphere is used as the reference benchmark. The formula for calculating the arithmetic mean is: .

[0081] Calculate the dispersion. Calculate the distance of each ray. Relative to the mean Standard deviation The standard deviation is used to measure the absolute variation in bladder shape relative to an ideal sphere. The formula for calculating the standard deviation is: .

[0082] Normalization judgment. To eliminate the interference of the bladder's own size on morphological judgment, a coefficient of variation is introduced. As a normalized indicator for measuring "relative deformation": .

[0083] In this embodiment, a preset threshold is set as follows: When the calculation yields When the bladder shape is below the preset threshold, it is determined that the current bladder shape is close to a regular sphere, and the system adaptively selects a spherical fitting model; when When the value exceeds a preset threshold, the system determines that the bladder has significant flattened, elongated, or irregular deformation, and the system adaptively selects an ellipsoidal fitting model.

[0084] Geometric model construction and outlier removal. Based on the adaptively determined model type (spherical or ellipsoidal) from the above examples, the system performs initial fitting and secondary geometric denoising.

[0085] Model building and initial estimation. Figure 11 Steps for building and removing noise from point clouds in the model Figure 7 and Figure 8 As shown, if the above example is determined to be spherical, then a structure is constructed with the center of the sphere as the center. and radius For a spherical property with parameters; if the above example is determined to be ellipsoidal, then construct a structure centered on... spindle length The equation of the ellipsoid with parameters is given. Finally, the least squares method is used to perform preliminary calculations on all current point cloud data to obtain initial estimates of the model parameters.

[0086] Outlier removal based on spatial domain (geometric denoising). The system calculates the normal radial residuals (i.e., the shortest distance from a point to a surface) of all 3D discrete point clouds to the initial estimation model. Basis and threshold are set. If the radial residual at a certain point (Right now Figure 7 , Figure 8 If a hollow point is located outside the envelope of the dashed line, it is determined to be a noisy point cloud or an outlier that does not conform to the geometric characteristics of the bladder. The system removes these outliers from the dataset and they are no longer included in subsequent parameter iterations, thereby eliminating shape distortions caused by local artifacts of the bladder wall or reflections from non-target tissues.

[0087] Nonlinear precise optimization. Using the effective point cloud after outlier removal, the Jacobian and Hessian matrices are constructed using Newton's iteration method. The model parameters are then updated nonlinearly until the objective function converges, thus obtaining an initial geometric model that best fits the true bladder boundary.

[0088] Respiratory follow-up compensation. This addresses the periodic shift in bladder position caused by respiratory movements, such as... Figure 9 As shown.

[0089] In this embodiment, based on the above-mentioned signal-level denoising and geometric-level denoising, a lightweight time window trend extraction mechanism is used for compensation.

[0090] Respiratory movement trends were acquired. Under continuous sampling conditions, the system recorded each sampling moment. Corresponding initial model geometric center coordinates Based on the removal of outliers in the examples, the obtained center coordinate sequence Random jitter caused by local noise has been eliminated, and only the trajectory characteristics reflecting the overall movement of the bladder have been retained.

[0091] Mean calculation. Without relying on complex frequency domain filtering, the system sets a time window (e.g., 3–5 seconds) covering the entire respiratory cycle and calculates the optimized geometric center coordinates of all components within this window. The arithmetic mean, such as Figure 9 As shown, it is defined as the bladder. "Average breathing position".

[0092] Displacement vector calculation and reverse compensation. Calculation of the geometric center at the current moment. Relative to the average respiratory position instantaneous displacement components Before performing subsequent refined calculations and voxel partitioning, displacement vectors are used. The effective point cloud coordinates after denoising at that moment are shifted back as a whole, and the calculation benchmark is uniformly corrected to a stable average breathing position, thereby effectively eliminating the interference of breathing motion artifacts on volume calculation under low computing power conditions.

[0093] Elastic surface correction and voxelized volume calculation. To address the difficulty of accurately describing local irregular boundaries (such as the bladder trigone or compression deformation zone) using rigid geometric models, this embodiment introduces a closed elastic correction surface for local deformation optimization. For example... Figure 10 As shown.

[0094] Elastic deformation. The fitted surface after the above steps is considered as an elastic mesh with finite stiffness. Calculate the normal residual from each measured point cloud to the surface. For residuals exceeding a set threshold... In the region where data is not covered, the mesh nodes are driven to undergo elastic deformation along the normal: if the point is on the outside of the surface, the mesh bulges outward; if it is on the inside, the mesh is concave inward; in the region where no data is covered, the surface remains smooth, thus constructing an elastically modified surface that closely matches the actual bladder boundary.

[0095] Voxelization calculation. A three-dimensional mesh is created within the space enclosed by the elastically modified surface and discretized into regular voxels with small side lengths. All voxel elements are traversed, and the center point of each voxel is determined using ray casting or a signed distance field to determine whether it lies inside the closed elastically modified surface.

[0096] Volume accumulation. The volumes of all voxel units determined to be "internal" are accumulated to obtain a high-precision bladder geometric volume. The volume calculation formula is as follows: .

[0097] in, It is the volume of each voxel. Volume unit representing a voxel The binary state (1 represents inside the bladder, 0 represents outside); , , These represent the number of volume elements in the horizontal, vertical, and longitudinal directions, respectively. This method effectively avoids the fundamental errors caused by shape assumptions in traditional formula methods.

[0098] Personalized modifications are introduced. In this embodiment, such as... Figure 11 As shown, considering the systematic bias that may be introduced by differences in abdominal wall thickness, BMI index, and soft tissue sound velocity among subjects, an individual correction factor is introduced. The correction coefficient The optimal range of values ​​is determined beforehand by consulting a table or calculation based on the examinee's height, weight, and other body type parameters. The final formula for calculating bladder capacity is: .

[0099] In one exemplary embodiment, such as Figure 12 As shown, a bladder morphology measuring device is provided, comprising: An ultrasound signal acquisition module is used to acquire ultrasound signals; the ultrasound signals are obtained by attaching a flexible ultrasound patch to the abdominal bladder region and sequentially transmitting and receiving ultrasound pulses based on each piezoelectric transducer.

[0100] The extraction module is used to extract the effective echo sequence based on the ultrasound signal and determine the three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder.

[0101] The fitting optimization module is used to fit the bladder morphology based on the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and to optimize the fitting parameters using the least squares method and Newton's iteration method to obtain an initial geometric model that matches the bladder boundary.

[0102] The deformation optimization module is used to perform breathing follow-up compensation processing on the initial geometric model, and to discretize the initial geometric model after breathing compensation into an elastic network model of topological structure for local elastic deformation optimization to obtain an elastic correction surface.

[0103] The geometric volume determination module is used to determine the geometric volume of the bladder based on the elastic modified surface using a voxel division method.

[0104] The correction module is used to obtain bladder capacity by weighting and correcting the individual correction coefficient and the geometric volume of the bladder, so as to realize the measurement of bladder morphology. The individual correction coefficient is determined based on the individual physiological characteristic parameters of the examinee and according to the pre-calibrated mapping relationship between the physiological characteristic parameters and the correction coefficient. The individual correction coefficient is used to compensate for the measurement deviation caused by differences in body size and the sound speed propagation characteristics of soft tissue.

[0105] In one exemplary embodiment, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0106] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0107] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0108] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0109] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method of measuring the shape of a urinary bladder, characterized by, The method comprises the following steps: acquiring an ultrasonic signal; the ultrasonic signal is obtained by attaching a flexible ultrasonic patch to the abdominal bladder area and sequentially transmitting and receiving ultrasonic pulses based on each piezoelectric transducer; extracting an effective echo sequence based on the ultrasonic signal and determining three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder; fitting the bladder shape according to the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and optimizing the fitting parameters by using the least square method and the Newton iteration method to obtain an initial geometric model matched with the bladder boundary; performing respiratory follow-up compensation processing on the initial geometric model, and discretizing the initial geometric model after the respiratory compensation into a topological structure elastic network model to perform local elastic deformation optimization to obtain an elastic correction surface; determining the geometric volume of the bladder according to the elastic correction surface by using the voxel division method; performing weighted correction on the geometric volume of the bladder according to an individual correction coefficient to obtain the bladder capacity, so as to realize the measurement of the bladder shape; the individual correction coefficient is determined based on the individual physiological characteristic parameters of the subject according to the mapping relationship between the pre-calibrated physiological characteristic parameters and the correction coefficient; and the individual correction coefficient is used to compensate the measurement deviation caused by the body type difference and the soft tissue sound speed propagation characteristics.

2. The bladder shape measurement method according to claim 1, characterized by, extracting an effective echo sequence based on the ultrasonic signal and determining three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder, specifically comprising: setting a window on the time axis based on the preset anatomical depth range of the anterior and posterior walls of the bladder according to the ultrasonic signal, and marking the wave peaks whose time stamps fall within the corresponding window range as candidate wall reflection signals, and removing the noise echoes outside the window range; the window comprises an anterior wall time window and a posterior wall time window; performing peak value detection on the waveform envelope in the window, and removing isolated noise peaks based on adjacent peak values to obtain an effective echo sequence; wherein for each local peak value detected, the corresponding time stamp is recorded, a time neighborhood is set within the adjacent range of the time stamp, and whether there is another local peak value in the time neighborhood is counted; if there is at least one accompanying peak value, it is marked as an effective echo peak and reserved as a candidate point; if it is an isolated peak in the time neighborhood, it is considered as an isolated noise peak generated by random noise, electronic interference or transient artifact and is deleted; converting each candidate point into a candidate reflection point in the form of spatial coordinates, and determining the three-dimensional coordinates of each candidate reflection point based on the time stamp and the preset sound speed to obtain the three-dimensional discrete point cloud coordinate information of the anterior and posterior walls of the bladder.

3. The method of claim 1, wherein, fitting the bladder shape according to the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, and optimizing the fitting parameters by using the least square method and the Newton iteration method to obtain an initial geometric model matched with the bladder boundary, specifically comprising: According to the spatial arrangement characteristics of the three-dimensional discrete point cloud coordinate information, a spherical or ellipsoidal shape is adaptively selected for bladder shape fitting; wherein, based on the depth coordinate difference of the bladder anterior wall point and the bladder posterior wall point in the normal direction in the detection direction of each piezoelectric transducer , a depth difference set is constructed, and the arithmetic mean and the standard deviation of the depth difference set are calculated and the standard deviation to construct the coefficient of variation as a dispersion index ; when is lower than a preset determination threshold, it is determined that the bladder shape is close to a regular sphere, and a spherical shape is selected as a fitting model for bladder shape fitting; otherwise, an ellipsoidal shape is selected as a fitting model for bladder shape fitting taking the spherical center coordinates and the radius, or the ellipsoid center and the major axis length as the geometric equation of the to-be-optimized fitting parameters, and establishing an objective function for minimizing the square sum of the radial distance residuals from the three-dimensional discrete point cloud coordinate information to the surface of the fitting model; Based on the geometric equation and the objective function, an initial estimated value of a fitting parameter is obtained by using a least square method, a radial residual of each three-dimensional discrete point cloud coordinate information is calculated, and an outlier point exceeding a preset geometric threshold is removed, and a remaining effective point cloud is updated by using a nonlinear optimization method of Newton iteration until a preset convergence condition is met, so that an initial geometric model matched with a bladder boundary is obtained.

4. The method of claim 1, wherein, The initial geometric model is subjected to a respiratory following compensation processing, and the initial geometric model after the respiratory compensation is discretized into an elastic network model with a topological structure to perform a local elastic deformation optimization, so that an elastic correction surface is obtained, and the local elastic deformation optimization specifically includes: Under a continuous sampling condition, a geometric center coordinate of the initial geometric model at each sampling time is taken as a time sequence; A respiratory average position of the time sequence is determined by using a time window average method or a trend extraction method, and an instantaneous displacement vector of the geometric center coordinate at a current time relative to the respiratory average position is determined; The bladder anterior and posterior wall point clouds are subjected to an inverse translation compensation by using the instantaneous displacement vector, a calculation reference is corrected to the respiratory average position, and the initial geometric model after the respiratory compensation is obtained; The initial geometric model after the respiratory compensation is discretized into an elastic grid model with a topological structure, the elastic grid model is given a physical attribute of limited rigidity, a radial normal residual of the bladder anterior and posterior wall point clouds to a surface of the elastic grid model is determined, and a deformation determination threshold is set; A local elastic deformation optimization is performed according to the radial normal residual, and an elastic correction surface is obtained.

5. The method of claim 4, wherein, A local elastic deformation optimization is performed according to the radial normal residual, and an elastic correction surface is obtained, and the local elastic deformation optimization specifically includes: When the bladder anterior and posterior wall point clouds are located outside the elastic grid model and the radial normal residual is greater than the deformation determination threshold, a grid node of a corresponding region is driven to expand outward to perform fitting; when the bladder anterior and posterior wall point clouds are located inside the elastic grid model and an absolute value of the radial normal residual is greater than the deformation determination threshold, the grid node of the corresponding region is driven to contract inward to perform fitting; when the absolute value of the radial normal residual is less than or equal to the deformation determination threshold, or in a region without the bladder anterior and posterior wall point cloud coverage, the grid surface of the elastic grid model is kept continuous and smooth by using the physical attribute of limited rigidity, so that a closed elastic correction surface is generated.

6. The bladder shape measurement method according to claim 1, wherein A geometric volume of the bladder is determined according to the elastic correction surface by using a voxel division method, and the geometric volume of the bladder specifically includes: A three-dimensional coordinate system is established within the space enclosed by the elastically modified surface, and the space is discretized into multiple units with constant unit volume according to a preset voxel size. A voxel grid composed of regular voxels; the voxel size is determined by the voxel spatial resolution; traversing all the voxel grids, extracting the geometric center coordinates of each voxel, and determining the inside and outside topological relationship of the geometric center coordinates of each voxel relative to the elastic correction surface by using a ray casting algorithm or a signed distance field algorithm, and counting the number of voxels located inside the elastic correction surface , according to calculating the geometric volume of the bladder .

7. The bladder shape measurement method according to claim 1, wherein An expression corresponding to the bladder capacity is: ； wherein, is the geometric volume of the bladder; is the individual correction factor; is the bladder volume.

8. A bladder form measuring device characterized by, including: An ultrasonic signal acquisition module is configured to acquire an ultrasonic signal, the ultrasonic signal being obtained by attaching a flexible ultrasonic patch to an abdominal bladder region and sequentially emitting and receiving ultrasonic pulses based on each piezoelectric transducer; An extraction module is configured to extract an effective echo sequence based on the ultrasonic signal and determine three-dimensional discrete point cloud coordinate information of a bladder anterior and posterior wall; A fitting optimization module is configured to fit a bladder shape according to a spatial arrangement feature of the three-dimensional discrete point cloud coordinate information, and optimize a fitting parameter by using a least square method and a Newton iteration method, so as to obtain an initial geometric model matched with a bladder boundary. a deformation optimization module, configured to perform a respiratory compensation processing on the initial geometric model, and discretize the initial geometric model after the processing to obtain a respiratory-compensated initial geometric model into a topological elastic network model to perform a local elastic deformation optimization to obtain an elastic corrected surface; a geometric volume determination module, configured to determine a geometric volume of the bladder according to the elastic corrected surface by using a voxel division method; a correction module, configured to perform a weighted correction on the geometric volume of the bladder according to an individual correction coefficient to obtain a bladder capacity, so as to measure a bladder shape; the individual correction coefficient is determined according to a mapping relationship between a physiological characteristic parameter of a subject and a correction coefficient based on individual physiological characteristic parameters of the subject, and is used to compensate for a measurement deviation caused by a body type difference and a soft tissue sound velocity propagation characteristic.

9. An electronic device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the bladder shape measurement method in any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the bladder shape measurement method in any one of claims 1-7.

Images