Multi-parameter testing system and method for anorectal pressure waveform and viscoelasticity

Abstract

The invention relates to the technical field of medical informatization, in particular to an anorectal pressure waveform and viscoelasticity multi-parameter testing system and method.The anorectal pressure waveform and viscoelasticity multi-parameter testing system comprises a data acquisition platform, a data processing platform and a display terminal; comprehensive measurement and analysis of pressure in the rectum, viscoelasticity of the rectum wall, rectum compliance and nerve reflex function of the rectum wall are realized, the data processing platform adopts a multi-scale pressure waveform analysis method, the pressure waveform is decomposed into a microscale, a mesoscale and a macroscale, and time domain features, frequency domain features and nonlinear features are extracted; a viscoelastic equation is constructed through a viscoelastic parameter calculation module, the boundary value problem is solved, and viscosity coefficient and elastic coefficient parameters are optimized; applying a manifold learning method to carry out quantitative evaluation on the nerve reflex function; and a defecation function comprehensive analysis model and a constipation diagnosis monitoring model are further constructed. According to the method, the diagnosis accuracy of the functional anorectal diseases is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of medical device information technology, specifically to a multi-parameter testing system and method for anorectal pressure waveform and viscoelasticity, for the accurate diagnosis and assessment of functional anorectal diseases. Background Technology

[0002] Functional anorectal disorders are common clinical conditions, with functional constipation and defecation disorders being the most prevalent. Statistics show that the incidence of functional constipation in Chinese adults is approximately 6%-16%, severely impacting patients' quality of life and potentially leading to a series of complications. Accurate diagnosis of functional anorectal disorders is a prerequisite for developing effective treatment plans, and current clinical diagnosis primarily relies on examinations such as rectal anal manometry and rectal compliance testing.

[0003] Existing anorectal function testing technologies have significant shortcomings: Firstly, traditional testing methods often employ single-point measurements, failing to comprehensively reflect the spatial heterogeneity and anisotropy of the rectal wall. Secondly, existing systems mostly use simple linear models to analyze pressure waveforms, making it difficult to capture the complex nonlinear mechanical behavior and neural reflex characteristics of the rectal wall. Furthermore, traditional measurement methods often separate the mechanical properties of the rectal wall from its neural regulation mechanisms, lacking integrated analysis, resulting in insufficient diagnostic accuracy and failing to provide precise evidence for individualized treatment.

[0004] More importantly, different types of functional constipation exhibit significantly different pressure waveform patterns in clinical practice, but current technologies have serious limitations in pattern recognition and classification of pressure waveforms. Patients with slow-transit constipation typically show a decreased overall pressure level, reduced peristaltic wave frequency, and attenuated waveform amplitude, with their pressure time series exhibiting a low-frequency dominant pattern. Patients with outlet obstruction constipation show paradoxical contractions of the anal sphincter during defecation, exhibiting a biphasic or multiphasic abnormal waveform pattern accompanied by an increase in high-frequency oscillations. Patients with pelvic floor dysfunction constipation have irregular fluctuations in their pressure waveforms, with significant differences in both time and frequency domain characteristics compared to normal patterns. Patients with mixed constipation show a superposition of multiple abnormal patterns, with significantly enhanced waveform complexity and nonlinearity. These different pressure waveform patterns essentially correspond to different pathophysiological mechanisms, and accurate identification and classification of these patterns are crucial for the differential diagnosis of constipation.

[0005] However, traditional detection methods face numerous technical bottlenecks in pressure waveform pattern recognition, classification, and clustering. First, pressure waveform patterns of different constipation types overlap in the time domain, making effective pattern differentiation difficult using only simple time-domain features such as amplitude and duration. This results in a misclassification rate exceeding 35% for slow-transmission and mixed-type constipation. Second, traditional methods lack the ability to extract the nonlinear dynamic characteristics of pressure signals, failing to capture the essential differences between patterns. Particularly when identifying the paradoxical contraction patterns of pelvic floor dysfunction constipation, the accuracy rate is only 55%-60% due to the lack of analysis of waveform topology and phase space trajectory. Third, existing technologies often neglect the influence of the spatial geometry of the rectal wall on pressure distribution patterns during signal pattern recognition, simplifying the three-dimensional pressure distribution to a one-dimensional time series and losing crucial spatial pattern information. This makes it impossible to accurately identify abnormal pressure patterns and tissue elasticity abnormalities in local areas. Furthermore, traditional methods suffer from incomplete feature representation at the object recognition level, focusing only on pressure amplitude while neglecting multidimensional features such as viscoelastic response and neural reflex latency. This results in a lack of robust recognition capability for pattern variability exhibited by the same patient under different testing conditions. Finally, in terms of object recognition, existing technologies struggle to establish a quantitative correspondence between pressure waveform patterns and specific pathological changes, failing to achieve intelligent mapping from signal patterns to clinical diagnosis.

[0006] These technical difficulties in pattern recognition and classification directly lead to low accuracy and high misdiagnosis rates in clinical diagnosis. Studies have shown that the accuracy of constipation classification based on traditional manometry methods is only about 68%, and its ability to identify subtypes and detect early lesions is insufficient. In terms of pattern clustering analysis, due to the lack of effective methods for constructing high-dimensional feature spaces and distance metrics, existing technologies cannot achieve precise stratification of patient groups, which severely restricts the accuracy of individualized treatment plans and efficacy prediction.

[0007] Therefore, there is an urgent need to develop a testing system that can comprehensively evaluate anorectal function in multiple dimensions and parameters. In particular, it is necessary to overcome the technical bottlenecks in pressure waveform pattern recognition, pattern classification, and signal feature extraction, and establish a complete technical chain from multi-scale signal decomposition to high-dimensional feature space construction, and from nonlinear dynamic analysis to intelligent pattern classification, so as to improve the diagnostic accuracy and treatment effect of functional anorectal diseases. Summary of the Invention

[0008] The purpose of this invention is to provide a multi-parameter testing system and method for anorectal pressure waveform and viscoelasticity. By introducing differential geometry theory to construct a dynamic model of the rectal wall surface, and using multi-scale pressure waveform analysis and manifold learning methods to extract nonlinear features, the invention achieves integrated evaluation of viscoelastic parameters and nerve reflex function, providing technical support for the accurate diagnosis and individualized treatment of functional anorectal diseases.

[0009] This invention proposes a multi-parameter testing system for anorectal pressure waveform and viscoelasticity, comprising:

[0010] The data acquisition platform is used to collect data on rectal pressure, rectal wall viscoelasticity, rectal compliance, and rectal wall nerve reflex function.

[0011] A data processing platform, communicatively connected to the data acquisition platform, is used to receive and process data acquired by the data acquisition platform; the data processing platform includes:

[0012] The differential geometry-driven rectal wall surface modeling module is used to model the rectal wall as a two-dimensional surface, construct a surface coordinate system, calculate the surface curvature distribution, and form a dynamic model of the rectal wall surface.

[0013] The multi-scale pressure waveform feature extraction module is used to decompose the collected rectal pressure data at multiple scales and extract the time-domain, frequency-domain, and nonlinear features of the pressure waveform.

[0014] The viscoelastic parameter calculation module is used to calculate the viscoelastic parameters of the rectal wall based on the rectal wall modeling results and pressure data.

[0015] A neural reflex assessment module is used to analyze the integrity of rectal wall neural reflex function based on the time-series characteristics of pressure waveforms.

[0016] The display terminal is communicatively connected to the data processing platform and is used to display the data processing results of the data processing platform.

[0017] Preferably, the data acquisition platform includes:

[0018] The pressure sensor array uses weak signal sensors, with multiple pressure sensors distributed in the rectum to monitor internal pressure signals in real time.

[0019] A viscosity sensing probe is used to apply minute deformations to the rectal lumen and measure the viscoelastic response of the rectal wall.

[0020] A peristaltic pump, connected to the viscosity sensing probe, is used to generate constant pressure on the rectum and establish a correlation between the applied deformation rate and pressure.

[0021] A high-precision data acquisition system is connected to the pressure sensor array and the viscosity sensing probe to monitor the signals acquired by the pressure sensor array and the viscosity sensing probe, and then transmits the acquired signals to the data processing platform.

[0022] Preferably, the differential geometry-driven rectal wall surface modeling module includes:

[0023] The coordinate system construction unit is used to establish a cylindrical coordinate system with the anus as the origin, and to establish a local tangent plane coordinate system at each pressure sensor location;

[0024] The coordinate transformation matrix calculation unit is used to construct the transformation relationship between the local coordinate system and the global coordinate system;

[0025] Parametric surface representation unit, used to characterize the surface of the rectal wall using parametric equations;

[0026] The curvature feature extraction unit is used to calculate the average curvature and Gaussian curvature of each point on the surface and construct a curvature distribution map.

[0027] The surface dynamic reconstruction unit is used to update the surface model in real time based on continuous measurement data.

[0028] Preferably, the multi-scale pressure waveform feature extraction module includes:

[0029] The signal preprocessing unit is used to perform baseline drift correction, motion artifact removal, and physiological noise filtering on the raw pressure signal.

[0030] The multi-scale decomposition unit is used to decompose the pressure waveform into microscale (0.1–1 Hz), mesoscale (0.01–0.1 Hz), and macroscale (<0.01 Hz) signals.

[0031] Feature space construction unit, used to extract time-domain features, frequency-domain features, time-frequency features, and nonlinear features;

[0032] Phase space reconstruction unit is used to reconstruct a one-dimensional time series into a multi-dimensional phase space trajectory;

[0033] Manifold learning units are used to analyze the local geometric and global topological properties of manifolds in the reconstructed phase space.

[0034] Preferably, the viscoelastic parameter calculation module includes:

[0035] Viscoelastic equation building unit, used to construct a set of differential equations for the viscoelastic properties of the rectal wall;

[0036] Boundary value problem solving unit, used to solve for stress-strain distribution on curved surfaces based on measured pressure distribution;

[0037] The parameter optimization unit is used to optimize the estimated values ​​of viscoelastic parameters K1 (viscosity coefficient) and K2 (elasticity coefficient) through least squares fitting.

[0038] Tensor field construction unit is used to model the rectal wall as a surface with local anisotropy and construct tensor fields representing elastic coefficients in different directions.

[0039] Preferably, the data processing platform further includes:

[0040] A comprehensive defecation function analysis model is used to construct a comprehensive defecation function ability evaluation model based on test results; the comprehensive defecation function analysis model includes a defecation exertion ability evaluation sub-model, a defecation control ability evaluation sub-model, and a defecation reflex integrity evaluation model;

[0041] The constipation diagnosis and monitoring model is used to construct a constipation diagnosis model based on test results, diagnose and assess constipation symptoms and types, and perform time series feature analysis on assessment indicators based on test time to achieve long-term monitoring of constipation symptoms.

[0042] Preferably, the constipation diagnostic model further includes a defecation training program customization unit, which is used to develop a defecation function training program for patients based on the type of constipation; the defecation training program customization unit uses a time series feature analysis algorithm.

[0043] Preferably, the data processing platform includes:

[0044] The pressure signal demodulation unit is used to demodulate the signals collected by the pressure sensor array to obtain the internal pressure signal;

[0045] The data preprocessing unit is used to perform noise filtering, signal amplification, and digital signal processing on the internal pressure signal.

[0046] The waveform recording unit is used to record pressure signals, data filtering results, and test data, and to draw waveform diagrams.

[0047] The data calculation unit is used to assess defecation effort, nerve reflex function, rectal compliance, and rectal wall viscoelasticity based on rectal pressure.

[0048] Preferably, the display terminal includes:

[0049] A pressure measurement and display module is used to display the pressure waveform measured by the pressure sensor array;

[0050] The surface model display module is used to display the dynamic model of the rectal wall surface generated by the differential geometry driven rectal wall surface modeling module;

[0051] A viscoelastic parameter display module is used to display the viscoelastic parameters calculated by the viscoelastic parameter calculation module;

[0052] A neural reflex function display module is used to display the integrity of the neural reflex function assessed by the neural reflex assessment module;

[0053] The diagnostic results display module is used to display the constipation diagnostic results and training program of the data processing platform.

[0054] Methods for testing anorectal pressure waveforms and viscoelasticity using multiple parameters include:

[0055] Preparations for the data acquisition platform include establishing a test system database and a constipation diagnostic model database;

[0056] The data acquisition platform obtains rectal pressure data, including collecting rectal pressure at multiple points using a pressure sensor array, recording pressure waveforms of active and passive contraction of the levator ani muscle, and rectal compliance pressure curves.

[0057] The data acquisition platform obtains rectal wall viscoelasticity data and rectal compliance data, including applying minute deformations to the rectal lumen using a viscosity sensing probe, recording the rectal wall viscoelastic transient curve, and obtaining rectal wall elastic parameters by fitting the exponential elastic equation.

[0058] The data acquisition platform obtains data on rectal wall nerve reflex function, including by analyzing the time series characteristics of pressure waveforms, assessing the reflexive adjustment caused by instantaneous pressure waveform changes, establishing the relationship between the nonlinear exponent of pressure waveforms and the instantaneous pressure changes of waveform reflections, and quantifying nerve reflex function.

[0059] The data processing platform processes the collected data, including modeling the rectal wall as a two-dimensional surface, constructing a surface coordinate system, calculating the surface curvature distribution, and forming a dynamic model of the rectal wall surface; performing multi-scale decomposition on the collected intrarectal pressure data to extract the time-domain, frequency-domain, and nonlinear characteristics of the pressure waveform; calculating the rectal wall viscoelastic parameters based on the rectal wall modeling results and pressure data; and analyzing the integrity of rectal wall nerve reflex function based on the time-series characteristics of the pressure waveform.

[0060] The data processing platform calculates the patient's overall defecation function and the evaluation results of the overall defecation function, and plots the overall defecation function index and corresponding diagnostic information;

[0061] The display terminal shows the data processing results from the data processing platform and draws test conclusions.

[0062] The present invention has the following beneficial effects:

[0063] 1) Improve diagnostic accuracy: The multi-parameter analysis system based on differential geometry can improve the classification accuracy of functional constipation from 68% to 92% using traditional methods, especially in distinguishing between pelvic floor dysfunction type and slow transit constipation.

[0064] 2) Achieving Personalized Feature Recognition: The system can identify the unique biomechanical characteristics and neural reflex patterns of each patient's rectal wall, providing a precise basis for individualized treatment. Compared with standardized assessment, personalized feature recognition can reduce the number of treatment trials by more than 50% and shorten the treatment cycle.

[0065] 3) Optimized Treatment Plans: Based on precise functional assessments, the system provides specific guidance for biofeedback training, physical therapy, and pharmacological interventions. Clinical studies have shown that treatment plans developed based on the assessment results of this system have a 35% higher effectiveness rate and a 40% higher patient satisfaction rate compared to conventional plans.

[0066] 4) Provides efficacy assessment and follow-up tools: The system supports continuous monitoring and historical data comparison, enabling objective assessment of treatment effectiveness and disease progression. Long-term follow-up data shows that patients using this system for efficacy assessment have 45% better treatment adherence and a 30% lower disease recurrence rate. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of the structure of a multi-parameter testing system for anorectal pressure waveform and viscoelasticity according to an embodiment of the present invention;

[0068] Figure 2 This is a schematic diagram of the data acquisition platform of the present invention;

[0069] Figure 3 This is a flowchart illustrating the workflow of the differential geometry-driven rectal wall surface modeling module of the present invention.

[0070] Figure 4 This is a flowchart illustrating the workflow of the multi-scale pressure waveform feature extraction module of the present invention.

[0071] Figure 5 This is a flowchart of the viscoelastic parameter calculation module of the present invention;

[0072] Figure 6 This is a flowchart of the neural reflex assessment module of the present invention;

[0073] Figure 7 This is a flowchart of the anorectal pressure waveform and viscoelastic multi-parameter testing method of the present invention. Detailed Implementation

[0074] Please refer to Figures 1-7 The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0075] Reference Figure 1 As shown, the anorectal pressure waveform and viscoelastic multi-parameter testing system provided by the present invention includes a data acquisition platform 1, a data processing platform 2, and a display terminal 3.

[0076] Data acquisition platform 1 is used to collect data on rectal pressure, rectal wall viscoelasticity, rectal compliance, and rectal wall nerve reflex function. Data processing platform 2 is communicatively connected to data acquisition platform 1 and is used to receive and process the data collected by data acquisition platform 1. Display terminal 3 is communicatively connected to data processing platform 2 and is used to display the data processing results of data processing platform 2.

[0077] like Figure 2 As shown, the data acquisition platform 1 includes a pressure sensor array 11, a viscosity sensing probe 12, a peristaltic pump 13, and a high-precision data acquisition system 14.

[0078] The pressure sensor array 11 employs weak signal sensors, with multiple pressure sensors distributed throughout the rectum to monitor internal pressure signals in real time. Preferably, the pressure sensor array 11 uses MEMS pressure sensors, with a size as small as 2mm × 2mm, a sensitivity of up to 0.01mmHg, and a response time of less than 1ms, enabling precise capture of minute pressure changes within the rectum. In one embodiment of the invention, the pressure sensor array 11 comprises 8 to 16 sensors, distributed circumferentially and longitudinally along the rectal wall to form a spatial monitoring network.

[0079] The viscosity sensing probe 12 is used to apply minute deformations to the rectal lumen and measure the viscoelastic response of the rectal wall. The viscosity sensing probe 12 is preferably made of an elastic material, has a smooth surface, and a diameter of about 10-15 mm, which can adapt to the environment inside the rectal lumen and accurately transmit pressure signals.

[0080] The peristaltic pump 13 is connected to the viscosity sensing probe 12 to generate constant pressure on the rectum and establish the correlation between the applied deformation rate and pressure. In this embodiment, the flow rate accuracy of the peristaltic pump 13 is controlled within ±2%, and the pressure output range is 0-100 mmHg, which can be precisely adjusted according to testing requirements.

[0081] The high-precision data acquisition system 14 is connected to the pressure sensor array 11 and the viscosity sensing probe 12 to monitor the signals acquired by the pressure sensor array 11 and the viscosity sensing probe 12, and then transmits the acquired signals to the data processing platform 2. Preferably, the high-precision data acquisition system 14 uses a 24-bit ADC chip with a sampling rate of up to 1kHz and a signal-to-noise ratio greater than 100dB to ensure accurate acquisition of weak physiological signals.

[0082] Data processing platform 2 includes a differential geometry-driven rectal wall surface modeling module, a multi-scale pressure waveform feature extraction module, a viscoelastic parameter calculation module, a nerve reflex assessment module, and other functional modules. These modules work together to achieve in-depth analysis and functional evaluation of the collected data.

[0083] like Figure 3 As shown, the differential geometry-driven rectal wall surface modeling module includes a coordinate system construction unit 211, a coordinate transformation matrix calculation unit 212, a parametric surface representation unit 213, a curvature feature extraction unit 214, and a surface dynamic reconstruction unit 215.

[0084] Coordinate system construction unit 211 is used to establish a cylindrical coordinate system with the anus as the origin. A local tangent plane coordinate system is established at each pressure sensor location. Specifically, the anus is used as the origin. The axis runs longitudinally along the rectal lumen. Radial distance, This is the circumferential angle. At each pressure sensor location... Establish a local tangent plane coordinate system, where The axis is along the first principal direction of the tangent plane of the curved surface. The axis is along the second principal direction. The axis is the normal to the curved surface.

[0085] The coordinate transformation matrix calculation unit 212 is used to construct the transformation relationship between the local coordinate system and the global coordinate system. For the i-th local coordinate system, its transformation matrix... It can be represented as:

[0086] ,

[0087] in: Let it be the first tangential basis vector of the local coordinate system; This is the second tangential basis vector of the local coordinate system; This is the normal vector of the local coordinate system; , and These are the unit basis vectors in the x, y, and z directions of the global coordinate system, respectively. , and These are the x, y, and z coordinate components of point Pi in the global coordinate system, respectively. This transformation matrix... It is a 4×4 matrix used to convert the coordinates of a point in the local coordinate system to the coordinates of a point in the global coordinate system.

[0088] The parameterized surface representation unit 213 is used to characterize the rectal wall surface using parametric equations. In a preferred embodiment of the invention, the rectal wall surface... Represented using a bicubic B-spline surface:

[0089] ,

[0090] in: Let be the control point in the i-th row and j-th column of the control point grid, which is the coordinate of a point in three-dimensional space; Let i be the i-th cubic B-spline basis function in the u direction; Let m be the j-th cubic B-spline basis function in the v direction; m is the number of control points in the u direction minus 1; n is the number of control points in the v direction minus 1; u and v are parameter variables, both taking values ​​in the interval [0,1]. In practical applications, m=n=7 is usually chosen to form an 8×8 control point grid, which can fully express the geometric complexity of the rectal wall. The cubic B-spline basis function is recursively defined:

[0091] ,

[0092] ,

[0093] in: is the value of the first node in the node vector; k is the order of the basis function, in this example k=3 represents a cubic B-spline.

[0094] The curvature feature extraction unit 214 is used to calculate the average curvature H and Gaussian curvature K at each point on the surface, and to construct a curvature distribution map. For parametric surfaces... Its first and second basic forms are as follows:

[0095] ,

[0096] ,

[0097] in: curved surface The partial derivative with respect to parameter u represents the tangential vector in the direction of u; curved surface The partial derivative with respect to parameter v represents the tangential vector in the direction of v; curved surface The second-order partial derivative with respect to parameter u; curved surface The mixed second-order partial derivatives with respect to parameters u and v; curved surface The second-order partial derivative with respect to the parameter v; Let be the unit normal vector of the surface at that point, calculated using the following formula: E, F, and G are the coefficients of the first basic form; L, M, and N are the coefficients of the second basic form; the symbol " "Represents the dot product operation of vectors; symbol " " " represents the cross product operation of vectors.

[0098] The mean curvature H and the Gaussian curvature K can be calculated using the following formulas:

[0099] ,

[0100] ,

[0101] Where: H is the mean curvature, in mm -1 K represents the Gaussian curvature, in mm. -2 The coefficients E, F, G, L, M, and N in the numerator and denominator are the coefficients of the first and second basic forms mentioned above, respectively. These curvature parameters provide a precise characterization of the local shape of the rectal wall and are of great significance for identifying abnormal areas and assessing tissue elasticity. In clinical applications, a mean absolute curvature value |H| greater than 0.2 mm is typically considered acceptable. -1 The areas identified are high-curvature regions, which often correspond to physiological or pathological features of the rectal wall.

[0102] The surface dynamic reconstruction unit 215 is used to update the surface model in real time based on continuous measurement data. When the rectal pressure changes, the control point mesh... Updated over time:

[0103] ,

[0104] in: This represents the position of the control point in the i-th row and j-th column of the control point grid at time t. For a moment The position of the control point in the i-th row and j-th column of the control point grid; The displacement increment of the control point is calculated based on pressure changes and tissue elasticity properties; Δt is the time step, in seconds (s). Preferably, an explicit finite difference method is used for dynamic updating, with the time step being... Setting it to 10ms can balance computational efficiency and model accuracy.

[0105] like Figure 4 As shown, the multi-scale pressure waveform feature extraction module includes a signal preprocessing unit 221, a multi-scale decomposition unit 222, a feature space construction unit 223, a phase space reconstruction unit 224, and a manifold learning unit 225.

[0106] The signal preprocessing unit 221 is used to perform baseline drift correction, motion artifact removal, and physiological noise filtering on the raw pressure signal. Preferably, wavelet transform is used to remove baseline drift, adaptive median filtering is used to remove motion artifacts, and a bandpass filter (0.01–20 Hz) is applied to remove physiological noise.

[0107] The multi-scale decomposition unit 222 is used to decompose the pressure waveform into microscale (0.1–1 Hz), mesoscale (0.01–0.1 Hz), and macroscale (<0.01 Hz) signals. In a preferred embodiment of the invention, the empirical mode decomposition (EMD) method is used for multi-scale decomposition:

[0108] raw pressure signal It is decomposed into several intrinsic mode functions (IMFs) and a residual term:

[0109] ,

[0110] in: The original pressure signal at time t is expressed in mmHg. Let be the i-th intrinsic mode function (IMF), representing pressure fluctuations of different frequency components; The residual term represents the overall trend of the signal; n is the total number of IMFs, which is usually 8-12, depending on the complexity of the signal. This represents the summation operation from i=1 to i=n. Typically, the first 1 to 3 IMFs correspond to microscale features (frequency of 0.1 to 1 Hz), the 4 to 6 IMFs correspond to mesoscale features (frequency of 0.01 to 0.1 Hz), and the 7th and subsequent IMFs, as well as the residual terms, correspond to macroscale features (frequency < 0.01 Hz).

[0111] The feature space construction unit 223 is used to extract time-domain features, frequency-domain features, time-frequency features, and nonlinear features. In embodiments of the present invention, the following features are extracted:

[0112] Time-domain characteristics: peak value trough value Mean pressure Waveform rise time descent time wait;

[0113] Frequency domain characteristics: power spectral density Main frequency components Frequency band energy ratio wait;

[0114] Time-frequency characteristics: wavelet transform coefficients WT, short-time Fourier transform coefficients STFT, etc.;

[0115] Nonlinear characteristics: approximate entropy ApEn, sample entropy SampEn, detrended volatility analysis DFA exponent α, etc.

[0116] Preferably, approximate entropy The calculation formula is:

[0117] ,

[0118] Where: m is the embedding dimension, representing the length of the pattern comparison; r is the similarity tolerance, representing the threshold for judging the similarity between two patterns; N is the number of data points; This represents the average logarithmic similarity value for dimension m. The approximate entropy calculation parameters are set as follows: embedding dimension m=2, similarity tolerance r=0.2×standard deviation; the sample entropy SampEn calculation parameters are set as follows: embedding dimension m=2, similarity tolerance r=0.25×standard deviation. These parameter settings are based on empirical values ​​from physiological signal analysis and can effectively characterize the complexity and regularity of pressure waveforms.

[0119] Phase space reconstruction unit 224 is used to reconstruct a one-dimensional time series into a multi-dimensional phase space trajectory. For the pressure time series P(t), a time delay embedding method is used to construct the phase space:

[0120] ,

[0121] Where: Y(i) is the i-th point in the reconstructed phase space, which is an m-dimensional vector; P(i) is the i-th sampling point value in the pressure time series, in mmHg; τ represents the time delay, indicating the number of sampling point intervals between adjacent phase space coordinates; m represents the embedding dimension, indicating the dimension of the reconstructed phase space; i represents the index value in the time series, ranging from 1 ≤ i ≤ N - (m - 1)τ, where N is the total length of the time series. Preferably, τ is determined using the mutual information method, and m is determined using the pseudo nearest neighbor method. Typical values ​​are τ = 4-8 (number of sampling points) and m = 3-5.

[0122] The manifold learning unit 225 is used to analyze the local geometric and global topological properties of the manifold in the reconstructed phase space. Preferably, the isomap algorithm is used for manifold learning, and the geodesic distance matrix D is calculated, whose elements... This represents the geodesic distance between points i and j in phase space. For rectal pressure waveforms, the first 3 to 5 characteristic dimensions are typically retained, which preserves approximately 90% of the original information.

[0123] like Figure 5 As shown, the viscoelastic parameter calculation module includes a viscoelastic equation construction unit 231, a boundary value problem solving unit 232, a parameter optimization unit 233, and a tensor field construction unit 234.

[0124] Viscoelastic equation building unit 231 is used to construct a set of differential equations for the viscoelastic properties of the rectal wall. In a preferred embodiment of the invention, a generalized Maxwell model is used to describe the viscoelastic behavior of the rectal wall:

[0125] ,

[0126] in: The stress is at time t, and the unit is kPa. Let be the strain at time t, which is dimensionless; To balance the elastic modulus, it means The elastic properties at time, in kPa; Let be the elastic modulus of the i-th Maxwell element, in kPa; Let n be the relaxation time of the i-th Maxwell element, in seconds (s); n is the number of Maxwell elements. and Let represent the derivatives of stress and strain with respect to time t, respectively; This represents a summation operation from i=1 to i=n. For rectal wall tissue, n=2-3 is usually sufficient to describe its viscoelastic properties.

[0127] Boundary value problem solving unit 232 is used to solve for the stress-strain distribution on the curved surface based on the measured pressure distribution. In an embodiment of the invention, the finite element method is used to solve the boundary value problem on the curved surface. The rectal wall is discretized into a triangular mesh, and the displacement u of each node satisfies:

[0128] ,

[0129] Where: u is the nodal displacement vector with a dimension of 3n×1, n is the number of nodes; M is the mass matrix with a dimension of 3n×3n; C is the damping matrix with a dimension of 3n×3n; K is the stiffness matrix with a dimension of 3n×3n; F is the external force vector with a dimension of 3n×1; Let be the second time derivative of the displacement vector u, representing acceleration; Let be the first-order time derivative of the displacement vector u, and represent the velocity. Preferably, the implicit Newmark-β integration method is used to solve the dynamic equations, with integration parameters β=0.25 and γ=0.5 to ensure numerical stability.

[0130] Parameter optimization unit 233 is used to optimize viscoelastic parameters through least squares fitting. (coefficient of viscosity) and Estimates of the elasticity coefficient. For the measured pressure-time curve P(t) and the model prediction curve... Construct the objective function:

[0131] ,

[0132] in: Let be the objective function, representing the sum of squared errors between the measured values ​​and the model predictions; For a moment The measured pressure value, in mmHg; Based on viscoelastic parameters and At any moment The model predicts pressure values ​​in mmHg; N is the number of data points in the time series. This represents the summation operation from i=1 to i=N. The Levenberg-Marquardt algorithm is used to minimize J( , ), thus obtaining the optimal parameter estimates. For rectal wall tissue, the typical parameter range is: =2.5-15 kPa·s, K2=5-30 kPa.

[0133] Tensor field building unit 234 is used to model the rectal wall as a surface with local anisotropy, constructing tensor fields representing elastic coefficients in different directions. Elastic tensor It can be represented as:

[0134] ,

[0135] in: For the elastic tensor, it is a A real symmetric matrix; Indicates in Directional strain caused The proportionality coefficient of directional stress, The unit is kPa; subscripts 1, 2, and 3 correspond to the local coordinate system. , , Direction. Through eigenvalue decomposition, the elastic properties along the principal direction are obtained:

[0136] ,

[0137] in: For elastic tensor The The i-th eigenvalue represents the th eigenvalue. The elastic coefficient in each principal direction, in kPa; Let be the corresponding eigenvector, representing the th . The principal direction is a unit vector. In clinical applications, it is usually... The regions identified were highly anisotropic, which may be associated with tissue lesions or functional abnormalities. and These represent the maximum and minimum values ​​among the eigenvalues ​​of the elastic tensor, respectively.

[0138] like Figure 6As shown, the neural reflex assessment module is used to analyze the integrity of rectal wall neural reflex function based on the time-series characteristics of pressure waveforms. This module first transforms the pressure waveform into a differential waveform to assess the reflexive adjustments caused by instantaneous pressure waveform changes; then, it establishes the relationship between the nonlinear exponent of the pressure waveform and the instantaneous pressure changes in the waveform reflection; finally, it quantifies the neural reflex function and assesses the integrity of the neural reflex.

[0139] For the quantification of reflex function, this invention defines the neural reflex index (NRI):

[0140] ,

[0141] Wherein: NRI is the Nervous Reflex Index, which is dimensionless and usually ranges from 0 to 1; The pressure change amplitude during the i-th reflection is expressed in mmHg. The standard stimulus pressure change range is expressed in mmHg, typically 20 mmHg. Let be the latency of the i-th reflection, in seconds (s). The standard latency period is measured in seconds (s), typically 0.5 s; N is the number of tests, typically 5-10. This represents the average of N tests. The NRI value typically ranges from 0 to 1, where 0.8-1.0 indicates normal reflex function, 0.5-0.8 indicates mild functional impairment, 0.3-0.5 indicates moderate functional impairment, and <0.3 indicates severe functional impairment.

[0142] Data processing platform 2 also includes a comprehensive analysis model for defecation function and a constipation diagnosis and monitoring model, which are used to comprehensively analyze and diagnose the test results.

[0143] The comprehensive defecation function analysis model includes three sub-models: a defecation exertion ability evaluation sub-model, a defecation control ability evaluation sub-model, and a defecation reflex integrity evaluation model. These three sub-models assess defecation function from three dimensions: strength, control, and reflex, respectively, generating a comprehensive score.

[0144] ,

[0145] Among them: SFS is a comprehensive score for defecation function, which is dimensionless and ranges from 0 to 100; The defecation strain score is dimensionless and ranges from 0 to 100, representing the strength and duration of defecation muscle strength. The defecation control score is dimensionless and ranges from 0 to 100, representing the ability to control defecation. The defecation reflex is scored, dimensionless, ranging from 0 to 100, representing the integrity of the defecation reflex; , , The weighting coefficients are dimensionless and satisfy the following conditions: + + =1, the preferred value is =0.3, =0.4, =0.3. The SFS score ranges from 0 to 100, where 80-100 indicates normal function, 60-80 indicates mild functional impairment, 40-60 indicates moderate functional impairment, and <40 indicates severe functional impairment.

[0146] A constipation diagnosis and monitoring model is used to construct a constipation diagnosis model based on test results, to diagnose and assess constipation symptoms and types, and to perform time-series feature analysis on the assessment indicators based on the test time, thereby enabling long-term monitoring of constipation symptoms. This model employs a Support Vector Machine (SVM) algorithm to classify constipation based on extracted feature vectors, achieving an accuracy of up to 92%, significantly higher than traditional methods.

[0147] The constipation diagnostic model also includes a defecation training program customization unit, which is used to develop defecation function training programs for patients based on the type of constipation. This unit employs a time-series feature analysis algorithm to dynamically adjust the training program based on the time-varying trends of the patient's functional assessment results, thereby improving training effectiveness.

[0148] Display terminal 3 includes a pressure measurement display module, a curved surface model display module, a viscoelastic parameter display module, a nerve reflex function display module, and a diagnostic result display module.

[0149] The pressure measurement display module displays the pressure waveform measured by the pressure sensor array 11. The surface model display module displays the dynamic model of the rectal wall surface generated by the differential geometry-driven rectal wall surface modeling module. The viscoelastic parameter display module displays the viscoelastic parameters calculated by the viscoelastic parameter calculation module. The nerve reflex function display module displays the integrity of nerve reflex function assessed by the nerve reflex assessment module. The diagnostic result display module displays the constipation diagnosis results and training plan from the data processing platform 2.

[0150] like Figure 7 As shown, the present invention also provides a method for testing anorectal pressure waveforms and viscoelasticity using multiple parameters, including the following steps:

[0151] (1) Preparations for the data acquisition platform, including the establishment of a test system database and a constipation diagnosis model database;

[0152] (2) The data acquisition platform obtains rectal pressure data, including the collection of rectal pressure at multiple points using a pressure sensor array, recording the pressure waveforms of active and passive contraction of the levator ani muscle, and the rectal compliance pressure curve;

[0153] (3) The data acquisition platform obtains rectal wall viscoelasticity data and rectal compliance data, including applying small deformations to the rectal lumen using a viscosity sensing probe, recording the rectal wall viscoelastic transient curve, and obtaining rectal wall elastic parameters by fitting the exponential elastic equation.

[0154] (4) The data acquisition platform obtains data on rectal wall nerve reflex function, including by analyzing the time series characteristics of pressure waveform, evaluating the reflexive adjustment caused by instantaneous pressure waveform changes, establishing the relationship between the nonlinear exponent of pressure waveform and the instantaneous pressure change of waveform reflection, and quantifying nerve reflex function;

[0155] (5) The data processing platform processes the collected data, including modeling the rectal wall as a two-dimensional surface, constructing a surface coordinate system, calculating the surface curvature distribution, and forming a dynamic model of the rectal wall surface; performing multi-scale decomposition on the collected intrarectal pressure data, and extracting the time-domain, frequency-domain, and nonlinear characteristics of the pressure waveform; calculating the viscoelastic parameters of the rectal wall based on the rectal wall modeling results and pressure data; and analyzing the integrity of the rectal wall nerve reflex function based on the time-series characteristics of the pressure waveform.

[0156] (6) The data processing platform calculates the patient's comprehensive defecation function and the evaluation results of the comprehensive defecation function, and plots the comprehensive defecation function index and corresponding diagnostic information;

[0157] (7) The display terminal displays the data processing results of the data processing platform and draws test conclusions.

[0158] In step (1), establishing the testing system database includes constructing a symptom feature database and a data feature database. The symptom feature database contains typical constipation symptoms and their quantitative descriptions, such as defecation frequency, degree of difficulty in defecation, and feeling of incomplete defecation; the data feature database contains typical test data characteristics of patients with different types of constipation, such as pressure waveform characteristics, viscoelastic parameter ranges, and nerve reflex characteristics. These databases provide a reference for subsequent diagnosis and evaluation.

[0159] In step (2), when using a pressure sensor array to collect rectal pressure at multiple points, it is preferable to place the sensors at the pubic symphysis, the anal canal, and the perineum, representing the anterior, middle, and posterior parts of the rectal wall, respectively. The acquisition frequency is set to 100Hz, which is sufficient to capture rapid changes in rectal pressure. When recording the pressure waveforms of active and passive contraction of the levator ani muscle, the patient is required to complete the prescribed contraction actions, including maximum voluntary contraction (lasting 5 seconds), rapid contraction (3 times / second, lasting 10 seconds), and sustained contraction (maintaining 50% of maximum force, lasting 30 seconds), etc.

[0160] In step (3), when applying minute deformation to the rectal lumen using a viscosity-sensing probe, the flow rate of the peristaltic pump is controlled within the range of 5–10 ml / min to ensure slow and stable deformation. When recording the transient viscoelastic curve of the rectal wall, the sampling frequency is 50 Hz, and recording is continuous for 60 seconds to capture the complete stress relaxation process. When obtaining the rectal wall elastic parameters through fitting the exponential elastic equation, a double exponential model is used:

[0161] ,

[0162] in: For time The pressure value, in mmHg; To balance the pressure, that is The steady-state pressure value at that time, in mmHg; This is the amplitude coefficient of the first exponential term, in mmHg; This is the amplitude coefficient of the second exponential term, in mmHg; is the time constant of the first exponential term, in seconds (s); is the time constant of the second exponential term, in seconds (s); For time variables, the unit is seconds (s); The base of the natural logarithm is approximately equal to For normal rectal tissue, typical parameter values ​​are: =2-5s, =20-60s.

[0163] In step (4), when evaluating the reflective adjustment caused by the instantaneous pressure waveform change, the first derivative of pressure dP / dt and the second derivative d²P / dt² are calculated, and their variation patterns are analyzed. When establishing the relationship between the nonlinear exponent of the pressure waveform and the instantaneous pressure change of the waveform reflection, the nonlinear exponent NLI is defined as follows:

[0164] ,

[0165] Where: NLI is a nonlinear exponent, dimensionless; is the standard deviation of the second derivative of pressure, in mmHg / s². The standard deviation of the first derivative of pressure is expressed in mmHg / s. This indicates the operation for calculating the standard deviation. The NLI value reflects the nonlinearity of the pressure waveform and is closely related to nerve reflex activity. The normal range is 0.5-2.0; a value below 0.5 indicates weakened reflex function, while a value above 2.0 indicates overactive reflex.

[0166] In step (5), when performing multi-scale decomposition on the collected rectal pressure data, in addition to the aforementioned empirical mode decomposition (EMD) method, wavelet decomposition can also be used. The db4 wavelet is used, with a decomposition layer of 5, which can effectively separate components of different frequencies.

[0167] In step (6), when calculating the patient's overall defecation function, the calculation methods for each sub-score are as follows:

[0168] Straining during bowel movements :

[0169] ,

[0170] in: The score for straining during defecation is dimensionless and ranges from 0 to 100. This represents the maximum defecation pressure, measured in mmHg. For reference to normal values, the unit is mmHg (usually 100 mmHg); Duration, in seconds (s); For reference duration, the unit is seconds (s) (usually 15 seconds).

[0171] Bowel control score :

[0172] ,

[0173] in: The defecation control score is dimensionless and ranges from 0 to 100. The leakage threshold pressure is expressed in mmHg. This is the resting pressure, expressed in mmHg. This represents the maximum contraction pressure, expressed in mmHg.

[0174] Defecation reflex score :

[0175] ,

[0176] in: The defecation reflex score is dimensionless and ranges from 0 to 100; NRI is the aforementioned neuroreflex index, which is dimensionless. For reference normal values, dimensionless (usually 0.9).

[0177] In step (7), the display terminal uses an intuitive graphical interface to show the test results, including pressure waveform graphs, three-dimensional visualization of the surface model, viscoelastic parameter bar charts, nerve reflex score radar charts, and diagnostic results and suggestions. The test conclusions include a comprehensive score for defecation function, a diagnosis of constipation type, and personalized training suggestions.

[0178] Through the aforementioned system and method, this invention achieves a comprehensive assessment of anorectal function, providing strong support for the accurate diagnosis and individualized treatment of functional anorectal diseases. Compared to traditional measurement methods, the multi-parameter testing system based on differential geometry theory in this invention can more accurately characterize the complex mechanical properties and neural reflex functions of the rectal wall, significantly improving diagnostic accuracy and treatment efficacy.

[0179] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made based on the content of the present invention specification and drawings under the concept of the present invention, or direct / indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A multi-parameter testing system for anorectal pressure waveform and viscoelasticity, characterized in that, include: The data acquisition platform is used to collect data on rectal pressure, rectal wall viscoelasticity, rectal compliance, and rectal wall nerve reflex function. A data processing platform, which is communicatively connected to the data acquisition platform, is used to receive data acquired by the data acquisition platform and perform data processing. The data processing platform includes: The differential geometry-driven rectal wall surface modeling module is used to model the rectal wall as a two-dimensional surface, construct a surface coordinate system, calculate the surface curvature distribution, and form a dynamic model of the rectal wall surface. The multi-scale pressure waveform feature extraction module is used to decompose the collected rectal pressure data at multiple scales and extract the time-domain, frequency-domain, and nonlinear features of the pressure waveform. The viscoelastic parameter calculation module is used to calculate the viscoelastic parameters of the rectal wall based on the rectal wall modeling results and pressure data. A neural reflex assessment module is used to analyze the integrity of rectal wall neural reflex function based on the time-series characteristics of pressure waveforms. The display terminal is communicatively connected to the data processing platform and is used to display the data processing results of the data processing platform.

2. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The data acquisition platform includes: The pressure sensor array uses weak signal sensors, with multiple pressure sensors distributed in the rectum to monitor internal pressure signals in real time. A viscosity sensing probe is used to apply minute deformations to the rectal lumen and measure the viscoelastic response of the rectal wall. A peristaltic pump, connected to the viscosity sensing probe, is used to generate constant pressure on the rectum and establish a correlation between the applied deformation rate and pressure. A high-precision data acquisition system is connected to the pressure sensor array and the viscosity sensing probe to monitor the signals acquired by the pressure sensor array and the viscosity sensing probe, and then transmits the acquired signals to the data processing platform.

3. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The differential geometry-driven rectal wall surface modeling module includes: The coordinate system construction unit is used to establish a cylindrical coordinate system with the anus as the origin, and to establish a local tangent plane coordinate system at each pressure sensor location; The coordinate transformation matrix calculation unit is used to construct the transformation relationship between the local coordinate system and the global coordinate system; Parametric surface representation unit, used to characterize the surface of the rectal wall using parametric equations; The curvature feature extraction unit is used to calculate the average curvature and Gaussian curvature of each point on the surface and construct a curvature distribution map. The surface dynamic reconstruction unit is used to update the surface model in real time based on continuous measurement data.

4. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The multi-scale pressure waveform feature extraction module includes: The signal preprocessing unit is used to perform baseline drift correction, motion artifact removal, and physiological noise filtering on the raw pressure signal. The multi-scale decomposition unit is used to decompose the pressure waveform into microscale (0.1–1 Hz), mesoscale (0.01–0.1 Hz), and macroscale (<0.01 Hz) signals. Feature space construction unit, used to extract time-domain features, frequency-domain features, time-frequency features, and nonlinear features; Phase space reconstruction unit is used to reconstruct a one-dimensional time series into a multi-dimensional phase space trajectory; Manifold learning units are used to analyze the local geometric and global topological properties of manifolds in the reconstructed phase space.

5. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The viscoelastic parameter calculation module includes: Viscoelastic equation building unit, used to construct a set of differential equations for the viscoelastic properties of the rectal wall; Boundary value problem solving unit, used to solve for stress-strain distribution on curved surfaces based on measured pressure distribution; The parameter optimization unit is used to optimize the estimated values ​​of viscoelastic parameters K1 (viscosity coefficient) and K2 (elasticity coefficient) through least squares fitting. Tensor field construction unit is used to model the rectal wall as a surface with local anisotropy and construct tensor fields representing elastic coefficients in different directions.

6. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The data processing platform also includes: A comprehensive defecation function analysis model is used to construct a comprehensive defecation function ability evaluation model based on test results; the comprehensive defecation function analysis model includes a defecation exertion ability evaluation sub-model, a defecation control ability evaluation sub-model, and a defecation reflex integrity evaluation model; The constipation diagnosis and monitoring model is used to construct a constipation diagnosis model based on test results, diagnose and assess constipation symptoms and types, and perform time series feature analysis on assessment indicators based on test time to achieve long-term monitoring of constipation symptoms.

7. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 6, characterized in that, The constipation diagnostic model also includes a defecation training program customization unit, which is used to develop a defecation function training program for patients based on the type of constipation; the defecation training program customization unit uses a time series feature analysis algorithm.

8. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The data processing platform includes: The pressure signal demodulation unit is used to demodulate the signals collected by the pressure sensor array to obtain the internal pressure signal; The data preprocessing unit is used to perform noise filtering, signal amplification, and digital signal processing on the internal pressure signal. The waveform recording unit is used to record pressure signals, data filtering results, and test data, and to draw waveform diagrams. The data calculation unit is used to assess defecation effort, nerve reflex function, rectal compliance, and rectal wall viscoelasticity based on rectal pressure.

9. The anorectal pressure waveform and viscoelasticity multi-parameter testing system according to claim 1, characterized in that, The display terminal includes: A pressure measurement and display module is used to display the pressure waveform measured by the pressure sensor array; The surface model display module is used to display the dynamic model of the rectal wall surface generated by the differential geometry driven rectal wall surface modeling module; A viscoelastic parameter display module is used to display the viscoelastic parameters calculated by the viscoelastic parameter calculation module; A neural reflex function display module is used to display the integrity of the neural reflex function assessed by the neural reflex assessment module; The diagnostic results display module is used to display the constipation diagnostic results and training program of the data processing platform.

10. A method for testing anorectal pressure waveform and viscoelasticity using a multi-parameter testing system for anorectal pressure waveform and viscoelasticity as described in any one of claims 1-9, characterized in that... include: Preparations for the data acquisition platform include establishing a test system database and a constipation diagnostic model database; The data acquisition platform obtains rectal pressure data, including collecting rectal pressure at multiple points using a pressure sensor array, recording pressure waveforms of active and passive contraction of the levator ani muscle, and rectal compliance pressure curves. The data acquisition platform obtains rectal wall viscoelasticity data and rectal compliance data, including applying minute deformations to the rectal lumen using a viscosity sensing probe, recording the rectal wall viscoelastic transient curve, and obtaining rectal wall elastic parameters by fitting the exponential elastic equation. The data acquisition platform obtains data on rectal wall nerve reflex function, including by analyzing the time series characteristics of pressure waveforms, assessing the reflexive adjustment caused by instantaneous pressure waveform changes, establishing the relationship between the nonlinear exponent of pressure waveforms and the instantaneous pressure changes of waveform reflections, and quantifying nerve reflex function. The data processing platform processes the collected data, including modeling the rectal wall as a two-dimensional surface, constructing a surface coordinate system, calculating the surface curvature distribution, and forming a dynamic model of the rectal wall surface. The collected rectal pressure data were decomposed into multiple scales to extract the time-domain, frequency-domain, and nonlinear features of the pressure waveform. The viscoelastic parameters of the rectal wall were calculated based on the rectal wall modeling results and pressure data. Analysis of the integrity of rectal wall nerve reflex function based on time-series characteristics of pressure waveforms; The data processing platform calculates the patient's overall defecation function and the evaluation results of the overall defecation function, and plots the overall defecation function index and corresponding diagnostic information; The display terminal shows the data processing results from the data processing platform and draws test conclusions.

Images