Method and system for dynamic pelvic floor function detection and evaluation

Abstract

The embodiment of the invention relates to the technical field of detection and evaluation, and discloses a method for detecting and evaluating a dynamic pelvic floor function, which comprises the following steps: filtering surface electromyogram signals of a first detection area according to a first area filtering matrix, and extracting the filtered surface electromyogram signals to obtain a first area feature group; filtering the surface electromyogram signals of the second detection area according to the second area filtering matrix, and extracting the filtered surface electromyogram signals to obtain a second area feature group; determining a third region feature group according to the fast muscle signal and the slow muscle signal; and performing evaluation calculation on the first region feature group, the second region feature group and the third region feature group by adopting a pelvic floor evaluation algorithm. According to the method for detecting and evaluating the dynamic pelvic floor function in the embodiment of the invention, each part of the pelvic floor is accurately detected by adopting the surface myoelectricity detection assembly, and the myoelectricity signals detected by each part are optimized through a specific filtering algorithm.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of detection and evaluation, in particular to a method and system for dynamic pelvic floor function detection and evaluation. BACKGROUND

[0002] At present, the incidence of pelvic floor dysfunction diseases (such as vaginal wall prolapse, pelvic organ prolapse, urinary incontinence, etc.) is high, which seriously affects the quality of life of patients. The traditional pelvic floor function detection method mainly relies on the subjective judgment of doctors, lacks quantitative and objective evaluation means, and the diagnosis process has certain subjectivity and uncertainty. Therefore, it has become a technical problem to be solved by those skilled in the art to design a scheme that can comprehensively and accurately detect the pelvic floor function. SUMMARY

[0003] In view of the defects, the embodiment of the present application discloses a method for dynamic pelvic floor function detection and evaluation, which can realize accurate evaluation of pelvic floor function.

[0004] The first aspect of the embodiment of the present application discloses a method for dynamic pelvic floor function detection and evaluation, comprising:

[0005] The surface electromyography detection component is used to acquire the surface electromyography signals of the corresponding detection areas in each detection stage, and the surface electromyography detection component includes a first detection component, a second detection component and a third detection component; the detection areas include a first detection area, a second detection area and a third detection area;

[0006] The acquired surface electromyography signals are subjected to frequency domain conversion to obtain corresponding frequency domain signals, and the frequency bands of interest of the frequency domain signals of each detection area are determined according to the attribute characteristics of each detection area, wherein the frequency bands of interest include a first frequency band of interest and a second frequency band of interest corresponding to the first detection area and the second detection area respectively;

[0007] A first covariance matrix of the first detection area and the second detection area, the third detection area in the first frequency band of interest is calculated, and a first area filtering matrix is constructed according to the first covariance matrix; the surface electromyography signals of the first detection area are subjected to filtering processing according to the first area filtering matrix, and the filtered surface electromyography signals are extracted to obtain a first area feature group;

[0008] A second covariance matrix of the second detection area and the first detection area in the second frequency band of interest is calculated, and a second area filtering matrix is constructed according to the second covariance matrix; the surface electromyography signals of the second detection area are subjected to filtering processing according to the second area filtering matrix, and the filtered surface electromyography signals are extracted to obtain a second area feature group;

[0009] The surface electromyography signals of the third detection area are analyzed by using the improved independent component analysis algorithm to obtain fast muscle signals and slow muscle signals in the third detection area, and a third area feature group is determined according to the fast muscle signals and the slow muscle signals;

[0010] The first area feature group, the second area feature group and the third area feature group are evaluated and calculated by using the pelvic floor evaluation algorithm to determine the score results of the detection areas and the comprehensive score result.

[0011] As an optional implementation, in the first aspect of the embodiment of the present application, the evaluation and calculation of the first area feature group, the second area feature group and the third area feature group by using the pelvic floor evaluation algorithm to determine the score results of the detection areas and the comprehensive score result, comprises:

[0012] The first area feature group in the first detection area in the corresponding detection stage is determined, the first area feature group comprises peak electromyography amplitude, contraction rising time and high-frequency signal proportion, and the area score result of the first detection area is determined according to the peak electromyography amplitude, the contraction rising time and the high-frequency signal proportion;

[0013] The second area feature group in the second detection area in the corresponding detection stage is determined, the second area feature group comprises electromyography signal standard deviation, endurance contraction end and initial amplitude ratio and low-frequency signal proportion, and the area score result of the second detection area is determined according to the electromyography signal standard deviation, the endurance contraction end and initial amplitude ratio and the low-frequency signal proportion;

[0014] The third area feature group of the third detection area in the corresponding detection stage is determined, the third area feature group comprises peak value of fast contraction and endurance parameter of sustained contraction, the synergy result of the fast muscle and the slow muscle in the third detection area is determined according to the peak value of fast contraction and the endurance parameter of sustained contraction, and the area score result of the third detection area is determined according to the peak value of fast contraction, the endurance parameter of sustained contraction and the synergy result;

[0015] The corresponding comprehensive score result is determined according to the area score result of the first detection area, the area score result of the second detection area and the area score result of the third detection area.

[0016] As an optional implementation, in the first aspect of the embodiment of the present application, the method for detection evaluation further comprises:

[0017] The surface electromyography signals of the first detection area after filtering are processed to determine the corresponding contraction rising rate and peak decay rate; wherein the contraction rising rate is the slope of the baseline rising to the peak value, and the peak decay rate is the ratio of the decline amplitude to the time of adjacent two peak values;

[0018] The surface electromyogram signal of the second detection area after filtering is processed to determine a corresponding contraction stability rate and a fatigue attenuation rate, wherein the contraction stability rate is a fluctuation rate of a low-frequency component amplitude; and the fatigue attenuation rate includes an amplitude difference between a durable contraction end and an initial stage and a time ratio;

[0019] The surface electromyogram signal of the second detection area after filtering is processed to determine a synergistic contraction synchronization rate, wherein the synergistic contraction synchronization rate is a change rate of an activation time difference between a fast muscle high-frequency component and a slow muscle low-frequency component;

[0020] The corresponding rate parameters in the first detection area, the second detection area and the third detection area are matched with a rate detection model to determine a current warning level.

[0021] As an optional implementation, in the first aspect of the embodiment of the present application, the matching of the corresponding rate parameters in the first detection area, the second detection area and the third detection area with the rate detection model to determine the current warning level comprises:

[0022] The corresponding rate parameters in the first detection area, the second detection area and the third detection area are respectively matched with an individual baseline and a group baseline in the rate detection model to determine the current warning level and a corresponding functional defect result.

[0023] As an optional implementation, in the first aspect of the embodiment of the present application, after the score results of the respective detection areas and the comprehensive score result are determined, the method further comprises:

[0024] The score conditions of the respective areas are output, and the score conditions of the respective areas are matched with a pre-set knowledge database to obtain dynamic evaluation results of the respective detection areas;

[0025] The dynamic evaluation results of the respective detection areas of the to-be-detected person and a rehabilitation training model are determined, wherein the rehabilitation training model includes: a plurality of standard rehabilitation training actions obtained after the rehabilitation training model is trained by using a historical rehabilitation training plan of a historical rehabilitation person, the historical rehabilitation training plan including a plurality of rehabilitation training actions.

[0026] The dynamic evaluation results of the respective detection areas are input into the rehabilitation training model, so that the rehabilitation training model selects a standard rehabilitation training action matching the dynamic evaluation results from the plurality of standard rehabilitation training actions, to generate and output a final rehabilitation training plan according to the to-be-detected person information and the selected standard rehabilitation training action;

[0027] The final rehabilitation training plan is sent to a corresponding display terminal.

[0028] As an optional implementation, in the first aspect of the embodiment of the present application, the rehabilitation training model is constructed by the following steps:

[0029] obtain a historical rehabilitation training plan and a dynamic evaluation result of each detection area of a historical rehabilitation personnel, wherein the historical rehabilitation training plan comprises a plurality of rehabilitation training actions;

[0030] perform feature recognition on the dynamic evaluation result to obtain rehabilitation demand information corresponding to the dynamic evaluation result;

[0031] generate a rehabilitation training action corresponding to each kind of rehabilitation demand information according to the rehabilitation demand information and the dynamic evaluation result;

[0032] train the initial rehabilitation training model based on the dynamic evaluation result of each detection area, the rehabilitation training action corresponding to each detection area, and the historical rehabilitation personnel information, to obtain a final rehabilitation training model.

[0033] As an optional implementation, in the first aspect of the embodiment of the present application, the training of the initial rehabilitation training model based on the dynamic evaluation result of each detection area, the rehabilitation training action corresponding to each detection area, and the historical rehabilitation personnel information, to obtain a final rehabilitation training model, comprises:

[0034] input the dynamic evaluation result of each detection area and the historical rehabilitation personnel information into the initial rehabilitation training model to generate an initial rehabilitation action corresponding to each kind of dynamic evaluation result;

[0035] determine an action difference and an intensity difference between the initial rehabilitation action corresponding to each detection area and the standard rehabilitation action corresponding to each detection area;

[0036] train the initial rehabilitation training model based on the action difference and the intensity difference between the initial rehabilitation action and the standard rehabilitation action to obtain a final rehabilitation training model.

[0037] The second aspect of the embodiment of the present application discloses a system for dynamic pelvic floor function detection and evaluation, comprising:

[0038] an acquisition module configured to acquire a surface electromyography signal of a corresponding detection area in each detection stage through a surface electromyography detection assembly, wherein the surface electromyography detection assembly comprises a first detection assembly, a second detection assembly, and a third detection assembly, and the detection area comprises a first detection area, a second detection area, and a third detection area;

[0039] The conversion module is used for converting the obtained surface electromyogram signals into frequency domain signals, and determining a frequency band of interest of the frequency domain signals of each detection area according to the attribute characteristics of each detection area, wherein the frequency band of interest includes a first frequency band of interest corresponding to the first detection area and a second frequency band of interest corresponding to the second detection area.

[0040] The first calculation module is used for calculating a first covariance matrix of the first detection area, the second detection area and the third detection area in the first frequency band of interest, constructing a first area filter matrix according to the first covariance matrix, filtering the surface electromyogram signals of the first detection area according to the first area filter matrix, and extracting the filtered surface electromyogram signals to obtain a first area feature group.

[0041] The second calculation module is used for calculating a second covariance matrix of the second detection area and the first detection area in the second frequency band of interest, determining a second area filter matrix according to the second covariance matrix, filtering the surface electromyogram signals of the second detection area according to the second area filter matrix, and extracting the filtered surface electromyogram signals to obtain a second area feature group.

[0042] The third calculation module is used for analyzing the surface electromyogram signals of the third detection area by using an improved independent component analysis algorithm to obtain fast muscle signals and slow muscle signals in the third detection area, and determining a third area feature group according to the fast muscle signals and the slow muscle signals.

[0043] The analysis module is used for evaluating and calculating the first area feature group, the second area feature group and the third area feature group by using a pelvic floor evaluation algorithm to determine score results of each detection area and a comprehensive score result.

[0044] The third aspect of the embodiment of the application discloses an electronic device, including: a memory in which executable program codes are stored; a processor coupled with the memory; the processor invokes the executable program codes stored in the memory, and is used for executing the method for performing dynamic pelvic floor function detection and evaluation disclosed in the first aspect of the embodiment of the application.

[0045] The fourth aspect of the embodiment of the application discloses a computer readable storage medium which stores a computer program, wherein the computer program causes a computer to execute the method for performing dynamic pelvic floor function detection and evaluation disclosed in the first aspect of the embodiment of the application.

[0046] Compared with the prior art, the embodiment of the application has the following beneficial effects:

[0047] The method for performing dynamic pelvic floor function detection and evaluation in the embodiment of the present application realizes accurate detection of each part of the pelvic floor by using a surface electromyography detection assembly, and optimizes the electromyography signals detected by each part through a specific filtering algorithm, thereby improving the accuracy of the final recognition result and providing a targeted basis for subsequent rehabilitation scheme formulation. BRIEF DESCRIPTION OF DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0049] Figure 1 is a flowchart of the method for performing dynamic pelvic floor function detection and evaluation disclosed in the embodiments of the present application;

[0050] Figure 2 is a flowchart of the regional score calculation disclosed in the embodiments of the present application;

[0051] Figure 3 is a flowchart of the early warning level determination disclosed in the embodiments of the present application;

[0052] Figure 4 is a flowchart of the rehabilitation training display disclosed in the embodiments of the present application;

[0053] Figure 5 is a structural diagram of a system for performing dynamic pelvic floor function detection and evaluation provided in the embodiments of the present application;

[0054] Figure 6 is a structural diagram of an electronic device provided in the embodiments of the present application. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort fall within the scope of protection of the present application.

[0056] It should be noted that the terms "first", "second", "third", "fourth" and the like in the description and claims of the application are used for distinguishing between similar objects, and are not necessarily used to describe a particular sequential order. The terms "include", "has", "having" and any variations thereof in the embodiments of the application are intended to cover a non-exclusive inclusion, such that a process, method, system, product or apparatus that comprises a list of steps or units is not necessarily limited to those steps or units that are clearly listed, but can include other steps or units that are not clearly listed or inherent to such processes, methods, products or apparatus.

[0057] Embodiment one

[0058] Please refer to Figure 1 , Figure 1 is a flowchart of the method for performing dynamic pelvic floor function detection and evaluation disclosed in the embodiments of the application. The execution subject of the method described in the embodiments of the application is an execution subject composed of software or / and hardware, which can receive relevant information through wired or / and wireless means and can send certain instructions. Of course, it can also have certain processing functions and storage functions. The execution subject can control multiple devices, such as remote physical servers or cloud servers and related software, or it can be a local host or server and related software that performs related operations on a device placed somewhere. In some scenarios, it can also control multiple storage devices, which can be placed in the same place or different places. As Figure 1 shown, the method based on performing dynamic pelvic floor function detection and evaluation includes the following steps:

[0059] S101: acquiring surface electromyography signals of corresponding detection areas in each detection stage through a surface electromyography detection assembly, the surface electromyography detection assembly including a first detection assembly, a second detection assembly and a third detection assembly; the detection areas including a first detection area, a second detection area and a third detection area;

[0060] S102: performing frequency domain conversion on the acquired surface electromyography signals to obtain corresponding frequency domain signals, determining a frequency band of interest of the frequency domain signals of each detection area according to the attribute characteristics of each detection area, the frequency band of interest including a first frequency band of interest and a second frequency band of interest corresponding to the first detection area and the second detection area, respectively;

[0061] S103: calculating a first covariance matrix of the first detection area and the second detection area, the third detection area in the first frequency band of interest, constructing a first area filtering matrix according to the first covariance matrix; filtering the surface electromyography signals of the first detection area according to the first area filtering matrix, and extracting the filtered surface electromyography signals to obtain a first area feature group;

[0062] S104: Calculate the second covariance matrix of the second detection area and the first detection area in the second frequency band of interest, determine the second region filtering matrix according to the second covariance matrix; filter the surface electromyogram signal of the second detection area according to the second region filtering matrix, and extract the filtered surface electromyogram signal to obtain a second region feature group;

[0063] S105: Analyze the surface electromyogram signal of the third detection area by using the improved independent component analysis algorithm to obtain fast muscle signals and slow muscle signals in the third detection area, and determine a third region feature group according to the fast muscle signals and the slow muscle signals;

[0064] S106: Perform evaluation calculation on the first region feature group, the second region feature group and the third region feature group by using the pelvic floor evaluation algorithm to determine the score results of each detection area and the comprehensive score result.

[0065] In the embodiment of the application, the specific region division can be divided into multiple regions. Specifically, the pelvic floor muscle is usually divided into three main regions with the urogenital hiatus and the anal hiatus as boundaries, corresponding to different functional divisions of the pelvic cavity:

[0066] The anterior region (urogenital region) is located between the pubic symphysis and the ischial tuberosity, and contains the urethral sphincter, bulbospongiosus muscle, superficial perineal transverse muscle and the like, mainly surrounds the urethra and vagina, and is responsible for controlling urination and supporting the anterior wall of the vagina.

[0067] The middle region (central region) takes the levator ani muscle as the core and is the main region of the pelvic floor muscle, supports the core organs such as the uterus, bladder and rectum, and is the key structure for preventing pelvic organ prolapse.

[0068] The posterior region (anal region) surrounds the anus and contains the external anal sphincter and coccygeal muscle, and is mainly responsible for controlling defecation and maintaining the closing and opening functions of the anus.

[0069] Each stage in the embodiment of the application includes a resting stage, a fast contraction stage, a sustained contraction stage, a durable contraction stage and the like; and the main purpose is to perform complete evaluation of the pelvic floor muscle. The surface electromyogram detection assembly acquires the surface electromyogram signals of different detection regions, can accurately capture the electrophysiological activities of the pelvic floor muscle in each detection stage, can diagnose the dysfunction of the pelvic floor muscle such as muscle weakness and poor muscle coordination, and provides basic data for subsequent analysis.

[0070] The embodiment of the present application converts the surface electromyogram signal into frequency domain, determines the frequency band of interest, decomposes the muscle electro-physiological signal into components of different frequencies, evaluates the activity mode and frequency characteristics of the muscle, helps to identify the abnormal activity mode of the muscle, sets different frequency bands of interest for different detection areas, can more targetedly analyze the muscle characteristics of each area, and improves the accuracy of analysis.

[0071] The filter matrix is constructed by calculating the covariance matrix, and the surface electromyogram signals of each detection area are filtered, which can remove the noise and interference in the signals, improve the clarity and stability of the signals, and the filtered signals can more accurately reflect the real state of the pelvic floor muscle, which is helpful for subsequent feature extraction and analysis, and further improves the reliability of detection and evaluation. The improved independent component analysis algorithm is used to analyze the surface electromyogram signals of the third detection area, and the fast muscle signal and the slow muscle signal can be separated. The fast muscle and the slow muscle play different roles in the function of the pelvic floor muscle, and the signals of the two can help to understand the contraction characteristics and endurance of the pelvic floor muscle, provide more detailed information for evaluating the function of the pelvic floor muscle, and facilitate to judge whether the muscle function is normal and to develop more accurate rehabilitation training plan.

[0072] The pelvic floor evaluation algorithm is used to evaluate and calculate each area feature group, and the score results of each detection area and the comprehensive score results are obtained, which can quantify the pelvic floor function condition.

[0073] Specifically, the filter matrix is constructed in the following way:

[0074] Preprocessing: The 50Hz notch filter is used for the multi-channel original signal to remove the power frequency interference, and then the band-pass filter is used to retain the effective frequency band of the electromyogram. This is a preliminary processing of the collected original signal to remove the interference signal, so that the subsequent analysis is more accurate.

[0075] Covariance matrix construction: The covariance matrix of the multi-channel signal is calculated to reflect the signal correlation between channels, and three principal components (cumulative variance contribution rate > 90%) are extracted by eigenvalue decomposition to construct a spatial filter matrix. The construction of the covariance matrix and the spatial filter matrix is for the subsequent separation of regional signals, which separates different regional signals by analyzing the signal correlation.

[0076] Regional signal separation: For the front zone channel, the weight of high frequency signal (> 100Hz) is strengthened, and the low frequency interference of the middle zone / back zone is suppressed; for the middle zone channel, the gain of low frequency signal (< 50Hz) is improved, and the high frequency interference of the front zone fast muscle contraction is weakened. This step realizes the preliminary separation of different regional signals and reduces the signal interference between regions.

[0077] Independent Component Analysis Optimization: For the mixed muscle area (posterior region), an improved ICA algorithm is used to separate fast and slow muscle signals. The number of iterations is set to 150, and the tanh non-linear function is selected. The separation effect is verified by muscle recruitment timing. Fast muscle signals should be activated in the rapid contraction phase, and slow muscle signals should be dominant in the sustained contraction phase. The separation accuracy is high. This step further separates the fast and slow muscle signals in the mixed muscle area, solving the problem of mixed signals in this area.

[0078] Differential measurement parameters of fast / slow muscle:

[0079] Anterior region (fast muscle dominant region) parameter design: Peak electromyographic amplitude (taking the maximum amplitude value of each contraction), contraction rise time (time interval from 10% to 90% of peak amplitude), and high-frequency component ratio (calculated by short-time Fourier transform, 100-200 Hz frequency band energy ratio) are measured in the rapid contraction phase. These parameters are used to evaluate the explosive power, activation speed and functional status of fast muscle.

[0080] Mid-region (slow muscle dominant region) parameter design: Contraction stability (calculated as the ratio of standard deviation to mean value of electromyographic signal during each contraction) is measured in the sustained contraction phase; fatigue index (amplitude ratio at the end of contraction to the initial stage) and low-frequency component ratio (calculated as the energy ratio in the 20-50 Hz frequency band) are measured in the endurance contraction phase. These parameters are used to evaluate the sustained contraction stability, endurance and recruitment of slow muscle.

[0081] Posterior region (mixed muscle area) synergy analysis: Peak time difference (contraction start interval between anterior and posterior regions) and amplitude correlation (signal correlation coefficient between mid and posterior regions) of sustained contraction are recorded synchronously to evaluate the synergistic function of fast and slow muscles. This step comprehensively analyzes the synergistic effect of fast and slow muscles in the mixed muscle area, providing a basis for comprehensive evaluation of pelvic floor muscle function.

[0082] Improved ICA algorithm is selected, with 150 iterations and tanh non-linear function. The improved ICA algorithm may be optimized based on traditional FastICA and other algorithms for the characteristics of electromyographic signals, such as adjusting the objective function and optimizing the iteration steps, to improve the accuracy and stability of signal separation.

[0083] Algorithm initialization: Set the initial parameters of the algorithm, such as the initial estimate of the mixing matrix, learning rate, etc. The selection of initial parameters will affect the convergence speed and separation effect of the algorithm, which can be reasonably set through multiple trials or based on experience.

[0084] Iterative calculation: Calculate according to the iterative formula of the improved ICA algorithm. In each iteration, the input signal is transformed according to the current estimated mixing matrix and separation matrix, and the transformed signal is processed using a nonlinear function (such as the tanh function) to update the separation matrix, so that it gradually approaches the true demixing matrix. After 150 iterations, the final separation matrix is obtained.

[0085] Signal separation: Use the obtained separation matrix to separate the preprocessed mixed electromyographic signal to obtain independent component signals. These independent component signals correspond to the fast muscle and slow muscle activity signals in the posterior region.

[0086] When performing subsequent separation verification, the separation effect can be verified based on the muscle recruitment timing

[0087] Determine the contraction phase: According to the experimental design and muscle physiological characteristics, determine the fast contraction phase and sustained contraction phase. During the experiment, the start time and end time of muscle contraction can be recorded, or specific stimulation signals can be used to determine these phases.

[0088] Feature extraction: Extract the features of the fast contraction phase and the sustained contraction phase from the independent component signals obtained by separation. Time domain features (such as root mean square value, zero crossing rate, etc.) and frequency domain features (such as power spectral density, intermediate frequency, etc.) can be used to describe the characteristics of the signals in different phases.

[0089] Time domain features: Calculate the root mean square value of each phase signal, which reflects the energy of the signal. Fast muscle produces relatively large energy in the fast contraction phase, so the root mean square value of its corresponding signal in this phase should be relatively large; slow muscle dominates muscle activity in the sustained contraction phase, so the root mean square value of its signal in this phase should be large.

[0090] Frequency domain features: Convert the time domain signal to the frequency domain signal by Fourier transform, and calculate the power spectral density. The frequency components of fast muscle and slow muscle electromyographic signals are different, with fast muscle signals having relatively more high frequency components and slow muscle signals having relatively more low frequency components. The frequency characteristics of signals in different phases can be analyzed according to the distribution of power spectral density.

[0091] Timing matching analysis: Match the extracted features with the muscle recruitment timing. Determine whether the separated signals conform to the rule that fast muscle signals are activated in the fast contraction phase (0-0.5 seconds) and slow muscle signals dominate in the sustained contraction phase (1-5 seconds). If an independent component signal has obvious features in the fast contraction phase and weak features in the sustained contraction phase, and is consistent with the known fast muscle activity characteristics, it can be considered as a fast muscle signal; conversely, if the signal has prominent features in the sustained contraction phase and is consistent with the slow muscle activity characteristics, it can be determined as a slow muscle signal.

[0092] More preferably, as shown in Figure 2 The evaluation and calculation of the first region feature group, the second region feature group and the third region feature group by the pelvic floor assessment algorithm to determine the score results of each detection region and the comprehensive score results include:

[0093] S1061: Determine the first region feature group in the first detection region in the corresponding detection stage, the first region feature group including peak electromyographic amplitude, contraction rise time and high-frequency signal proportion, to determine the region score result of the first detection region according to the peak electromyographic amplitude, the contraction rise time and the high-frequency signal proportion;

[0094] S1062: Determine the second region feature group in the second detection region in the corresponding detection stage, the second region feature group including electromyographic signal standard deviation, endurance contraction end and initial amplitude ratio and low-frequency signal proportion, to determine the region score result of the second detection region according to the electromyographic signal standard deviation, the endurance contraction end and initial amplitude ratio and the low-frequency signal proportion;

[0095] S1063: Determine the third region feature group of the third detection region in the corresponding detection stage, the third region feature group including peak value of fast contraction and endurance parameter of sustained contraction, determine the synergy result of fast muscle and slow muscle of the third detection region according to the peak value of fast contraction and the endurance parameter of sustained contraction, and determine the region score result of the third detection region according to the peak value of fast contraction, the endurance parameter of sustained contraction and the synergy result;

[0096] S1064: Determine the corresponding comprehensive score result according to the region score result of the first detection region, the region score result of the second detection region and the region score result of the third detection region.

[0097] In the embodiment of the application, the first detection region focuses on peak electromyographic amplitude, contraction rise time, high-frequency signal proportion and the like. The peak electromyographic amplitude directly reflects muscle contraction strength, the contraction rise time embodies muscle starting speed, and the high-frequency signal proportion is related to muscle fast contraction ability (such as fast muscle fiber activity). Through quantitative analysis of these three dimensions, the explosive power, reaction speed and fast muscle dominant functional state of the muscles in this region can be accurately evaluated, and problems such as muscle weakness and contraction delay can be identified.

[0098] The second detection area: taking the standard deviation of the electromyogram signal, the amplitude ratio of the end and initial stage of the endurance contraction, and the low frequency signal proportion as the core. The standard deviation of the electromyogram signal reflects the stability and coordination of muscle contraction, the endurance contraction amplitude ratio evaluates the muscle endurance (the ability to avoid fatigue), and the low frequency signal proportion is related to the sustained activity of slow muscle fibers. These features can accurately capture the endurance level, contraction stability and slow muscle dominant functional state of the muscles in this area, and effectively identify abnormalities such as muscle fatigue and uncoordinated contraction.

[0099] The third detection area: combining the rapid contraction peak value, the sustained contraction endurance parameter, and the fast muscle and slow muscle coordination. The rapid contraction peak value represents the fast muscle explosive force, the sustained contraction endurance parameter reflects the slow muscle endurance, and the coordination result evaluates the cooperation efficiency of the two (such as the cooperation ability of the pelvic floor muscles in complex movements). Through the combination of the three, the fast and slow muscle functions and coordination of the muscles in this area are comprehensively covered, avoiding the limitations of a single indicator, and more accurately reflecting the overall functional state of the muscles.

[0100] The scheme of the embodiment of the application reflects the pelvic floor function from the local to the whole (such as focusing on the fast muscle explosive force in the first area, focusing on the slow muscle endurance in the second area, and focusing on the coordination in the third area, which together constitute the core dimension of the pelvic floor function); identifies the overall coordination problem that is difficult to find by identifying a single area (such as the pelvic floor disorder caused by the normal function of a certain area but poor overall coordination); and provides a more comprehensive basis for clinical diagnosis and treatment plan formulation (such as when the comprehensive score is low, the specific weak links can be located by combining the scores of each area).

[0101] The multi-region parameter fusion and comprehensive scoring of the embodiment of the application constructs a three-dimensional scoring system of region-muscle fiber-function:

[0102] The front area: fast muscle function score (weight 30%) = rapid contraction peak value x 0.6 + rise time x 0.4; the middle area: slow muscle function score (weight 40%) = sustained contraction stability x 0.5 + endurance contraction fatigue index x 0.5; the overall score result is improved by the high frequency signal proportion and the low frequency signal proportion; the rear area: mixed function score (weight 30%) = fast muscle peak value x 0.3 + slow muscle endurance x 0.3 + coordinated contraction synchronization x 0.4

[0103] The final comprehensive score directly reflects the functional state of each area (such as the specific ability deficiency of the fast muscle weakness in the front area, the slow muscle endurance deficiency in the middle area, and the delayed fast muscle response of the anal sphincter in the rear area).

[0104] In addition to the above parameters, a high-frequency band power spectrum peak value can be used for the front area, which is more resistant to high-frequency noise; the stability of fast muscle activation is evaluated by the coefficient of variation of the contraction rise time, reducing the fluctuation misjudgment caused by artifacts. For the middle area, the low-frequency band power spectrum density integral is used instead of the original amplitude; the slow muscle fatigue is evaluated by the entropy value of the endurance contraction (the lower the entropy value, the more stable the signal), and a sudden increase in the entropy value indicates real fatigue (not noise interference).

[0105] More preferably, as shown in Figure 3 The method for detecting and evaluating further comprises:

[0106] S1051: performing data processing on the surface electromyogram signal of the first detection area after filtering to determine the corresponding contraction rise rate and peak decay rate; wherein the contraction rise rate is the slope from the baseline to the peak, and the peak decay rate is the ratio of the decline amplitude to the time of adjacent two peaks;

[0107] S1052: performing data processing on the surface electromyogram signal of the second detection area after filtering to determine the corresponding contraction stability rate and fatigue decay rate, wherein the contraction stability rate is the fluctuation rate of the amplitude of the low-frequency component; and the fatigue decay rate includes the amplitude difference between the end and the beginning of the endurance contraction and the ratio of the time;

[0108] S1053: performing data processing on the surface electromyogram signal of the second detection area after filtering to determine the synergistic contraction synchronization rate, wherein the synergistic contraction synchronization rate is the change rate of the activation time difference between the fast muscle high-frequency component and the slow muscle low-frequency component;

[0109] S1054: matching the corresponding rate parameters in the first detection area, the second detection area and the third detection area with the rate detection model to determine the current warning level.

[0110] The scheme of the present application refines the analysis of the function of the pelvic floor muscles from the perspective of dynamic changes by extracting rate parameters, making up for the limitations of static characteristics (such as peak value and amplitude ratio):

[0111] The contraction rise rate and the peak decay rate of the first detection area:

[0112] The contraction rise rate (slope from baseline to peak) directly reflects the starting efficiency of the muscle from rest to exertion, and can identify muscle contraction retardation; the peak decay rate (ratio of adjacent peak decline amplitude to time) reflects the strength maintenance ability of the muscle during continuous contraction, and is used to determine whether the muscle has rapid fatigue or abnormal contraction continuity.

[0113] The contraction stability rate and the fatigue decay rate of the second detection area: the contraction stability rate (low-frequency component amplitude fluctuation rate) focuses on the stability of the muscle during sustained contraction, and excessive fluctuations indicate poor contraction coordination (such as imbalance between the pelvic floor muscles and the surrounding muscle groups); the fatigue decay rate (amplitude difference at the end of the endurance contraction and at the beginning of the contraction / time) quantifies the speed of muscle endurance decline and more accurately reflects the degree of fatigue of slow muscle fibers (such as the progress speed of postpartum pelvic floor muscle endurance deficiency).

[0114] The synchronous rate of the third detection area: the change rate of the activation time difference between the high-frequency component of fast muscle and the low-frequency component of slow muscle, which directly evaluates the coordination efficiency of fast and slow muscles in complex movements (such as coughing and lifting heavy objects), and abnormal synchronous rate indicates muscle coordination disorder (such as the asynchronous activation of fast and slow muscles commonly seen in stress urinary incontinence). These parameters reveal the progress trend of muscle dysfunction from the dynamic change speed level, rather than just evaluating the static state.

[0115] The scheme of the embodiment of the present application converts the rate parameters into warning levels (such as low, medium, and high risk) based on preset models (such as normal range thresholds and abnormal progress speed thresholds), so that users can quickly identify the urgency of functional abnormalities, such as high warning levels indicating rapid deterioration of muscle function, which requires immediate intervention.

[0116] The timing of targeted intervention: different warning levels correspond to different intervention strategies, such as low risk can be improved through home training, and high risk requires clinical treatment, avoiding the one-size-fits-all approach to all abnormal situations and improving the timeliness and accuracy of intervention. Changes in rate parameters can be used to track the effectiveness of treatment or training, such as an increase in contraction rate and a decrease in fatigue decay rate after intervention, indicating improved function, making the warning level a dynamic indicator of treatment efficacy.

[0117] Specifically, the model can also automatically recommend training adjustment strategies based on the warning level:

[0118] First-level warning: maintain basic training and increase the frequency of rate monitoring, such as 2 times of evaluation per week;

[0119] Second-level warning: strengthen training for abnormal rate parameters, such as increasing 0.5-second fast contraction training, 3 sets per day x 20 times, for low fast muscle contraction rate;

[0120] Third-level warning: start individualized rehabilitation programs, such as using 20-second contraction and 10-second rest intermittent training to reduce the fatigue progress speed for slow muscle fatigue decay, and quantifying the rehabilitation effect through the above-mentioned ways: comparing the rate parameters before and after training to intuitively reflect the improvement degree of fast muscle explosive power.

[0121] The scheme of the embodiment of the present application can also perform regional comparison monitoring: synchronously comparing the rate parameters of the front region, the middle region and the rear region, and identifying regional rate abnormalities: if only the front region has abnormal fast muscle ascending rate, and the middle region and the rear region are normal, it is suggested that the problem is limited to the front region fast muscle; if the middle region slow muscle fatigue decay rate and the rear region coordinated synchronous rate are abnormal at the same time, it is suggested that there is multi-region linkage dysfunction.

[0122] More preferably, the matching of the corresponding rate parameters in the first detection region, the second detection region and the third detection region with the rate detection model to determine the current warning level comprises:

[0123] The corresponding rate parameters in the first detection region, the second detection region and the third detection region are matched with the individual baseline and the group baseline in the rate detection model respectively to determine the current warning level and the corresponding functional defect result.

[0124] The individual baseline of the embodiment of the present application is a benchmark established based on historical detection data (such as rate parameters in a healthy state) of the same subject, and by comparing the current rate parameters with the individual baseline, the dynamic change trend of the individual function can be accurately captured. This method avoids the problem of individual differences being hidden due to the reliance on group standards only, and is especially suitable for monitoring the change of pelvic floor function over time (such as postpartum rehabilitation tracking and middle-aged functional decline warning).

[0125] The group baseline is a normal range obtained by statistics based on a large number of people of the same type (such as people of the same age, the same sex and without pelvic floor diseases), and by comparing with the group baseline, the deviation of the current rate parameters of the subject from the overall population can be quickly judged. This horizontal reference can clearly determine the general abnormality of the functional defect, and provides a basis for distinguishing between individual specific fluctuations and pathological abnormalities.

[0126] By combining the individual baseline with the group baseline, a double standard of longitudinal tracking + horizontal reference is constructed, making the matching of the rate parameters and the determination of the warning level more scientific and operable. This system not only retains individual differences, but also ensures the reference of the evaluation results.

[0127] More preferably, as shown in Figure 4 After the score results of each detection region and the comprehensive score result are determined, the method further comprises:

[0128] S107: outputting the score of each region, and matching the score of each region with a pre-set knowledge database to obtain a dynamic evaluation result of each detection region;

[0129] S108: determine the dynamic evaluation result of each detection area of the to-be-detected person and a rehabilitation training model, wherein the rehabilitation training model comprises a plurality of standard rehabilitation training actions obtained by training the rehabilitation training model respectively using a historical rehabilitation training plan of a historical rehabilitation person, and the historical rehabilitation training plan comprises a plurality of rehabilitation training actions;

[0130] S109: input the dynamic evaluation result of each detection area into the rehabilitation training model, so that the rehabilitation training model selects a standard rehabilitation training action matching the dynamic evaluation result from the plurality of standard rehabilitation training actions, to generate and output a final rehabilitation training plan according to the to-be-detected person information and the selected standard rehabilitation training action;

[0131] S1010: send the final rehabilitation training plan to a corresponding display terminal.

[0132] The scheme of the embodiment of the application generates a dynamic evaluation result by matching the scores of each area with a preset knowledge database, and the dynamic evaluation result is no longer limited to a static score value, but is converted into a functional description that can be directly understood in combination with information such as clinical cases, pathological characteristics, and functional correspondence stored in the database.

[0133] The model is trained based on the training plan of the historical rehabilitation person, covers a plurality of standard rehabilitation training actions, and is essentially machine learning and refining of the functional defect-effective training scheme correspondence. Such a model based on real rehabilitation data can avoid the subjectivity of traditional training schemes (such as relying on experience to develop) and ensure the effectiveness and safety of the recommended actions.

[0134] By inputting the dynamic evaluation result into the model, the model selects a training action highly matching the current functional defect from the standard actions (such as a rapid contraction training for improving explosive force when the first area has insufficient fast muscle function, and a sustained contraction training when the second area has insufficient endurance). In combination with individual information (such as age, physical condition, and medical history) of the to-be-detected person, the intensity and frequency of the training action are further optimized, and the final generated plan truly realizes personalized rehabilitation training.

[0135] The output final rehabilitation training plan contains specific and standardized actions, avoiding the ambiguity of traditional verbal guidance or written description. The patient can directly refer to the plan for execution, reducing the risk of ineffective training or injury caused by non-standard actions. The plan is received by a display terminal (such as a mobile phone or a tablet), and the patient can check and track the training progress at any time. At the same time, the terminal pushes the plan, which is convenient for subsequent real-time adjustment of the plan in combination with monitoring data (such as electromyographic feedback during home training).

[0136] The method of the embodiment of the application directly associates the pelvic floor function detection and the rehabilitation intervention through the flow of evaluation results, training plan generation, and terminal pushing, avoids the disconnection problem of separately formulating a training plan after detection, and ensures the continuity and pertinence of the rehabilitation process and improves the rehabilitation effect, with the dynamic evaluation results in the subsequent detection being updated, the rehabilitation training model being re-matched with the training actions, and the plan being adjusted in real time with the function improvement (such as gradually transitioning from basic endurance training to high-intensity coordination training).

[0137] More preferably, the rehabilitation training model is constructed through the following steps:

[0138] obtaining a historical rehabilitation training plan and dynamic evaluation results of each detection area of a historical rehabilitation personnel, wherein the historical rehabilitation training plan comprises a plurality of rehabilitation training actions;

[0139] performing feature recognition on the dynamic evaluation results to obtain rehabilitation demand information corresponding to the dynamic evaluation results;

[0140] generating a rehabilitation training action corresponding to each kind of rehabilitation demand information according to the rehabilitation demand information and the dynamic evaluation results;

[0141] training the initial rehabilitation training model based on the dynamic evaluation results of the each detection area and the rehabilitation training actions corresponding thereto and the historical rehabilitation personnel information to obtain a final rehabilitation training model. The rehabilitation training model herein can be a neural network model or a gradient boosting model.

[0142] The embodiment of the application generates a rehabilitation training action corresponding to each kind of demand according to the rehabilitation demand information, and essentially constructs a training action library covering a plurality of pelvic floor function defects. For example, for the endurance deficiency demand, the library contains sustained contraction, stepwise endurance training, and the like; for the poor coordination demand, the library contains fast-slow muscle alternate contraction, synchronous activation training, and the like. Such a structured correspondence relationship enables the model to cope with diversified function defects of different individuals (such as postpartum female pelvic floor muscle relaxation and male urinary incontinence related dysfunction).

[0143] The historical rehabilitation personnel information (such as age, physical fitness, medical history, rehabilitation cycle, and the like) is incorporated when training the model, so that the model not only learns the association between function defects and actions, but also learns the influence of individual differences on training effect. For example, the model will identify the rule that young patients have better tolerance to high-intensity training, patients with lumbar vertebrae disease need to avoid some actions with high abdominal pressure, and the like, so as to take into account the individual particularity when subsequently recommending, and further improve the safety and adaptability of the training plan.

[0144] The training process of the model is based on a clear link between dynamic evaluation results, rehabilitation needs, training actions, and individual information, and each recommended action can be traced back to the corresponding rehabilitation needs and historical case basis. This explainability facilitates user understanding and verification of the rationality of the model recommendations, and also facilitates patient trust and cooperation with the training plan. Specific training actions include: rehabilitation actions, action time groups, action group numbers, action intensity, presence or absence of resistance, resistance strength, and the like.

[0145] As new historical rehabilitation data is continuously added, the model can update the demand-action correspondence library by retraining, continuously optimizing the recommendation logic. For example, when a new collaborative training action is proven to be more effective for a specific coordination defect, the model can quickly learn and incorporate it into the recommendation range, ensuring that the model remains adaptable to the latest rehabilitation techniques at all times. It can achieve precise action recommendation.

[0146] More preferably, the initial rehabilitation training model is trained based on the dynamic evaluation results of the respective detection areas and their respective corresponding rehabilitation training actions, as well as the historical rehabilitation personnel information, to obtain a final rehabilitation training model, comprising:

[0147] The dynamic evaluation results of each detection area and the historical rehabilitation personnel information are input into the initial rehabilitation training model to generate an initial rehabilitation action corresponding to each dynamic evaluation result;

[0148] For the dynamic evaluation results of each detection area, the action difference and intensity difference between the corresponding initial rehabilitation action and the corresponding standard rehabilitation action are determined;

[0149] Based on the action difference and intensity difference between the corresponding standard rehabilitation action, the initial rehabilitation training model is trained to obtain a final rehabilitation training model.

[0150] The embodiment of the present application compares the initial rehabilitation action generated by the initial rehabilitation training model with the standard rehabilitation action (effective action verified by historical data) to accurately identify the differences between the two in action details (such as contraction angle, force position) and intensity parameters (such as contraction duration, interval time, and repetition number). This difference provides a clear correction target for model optimization, avoiding the blindness of model training.

[0151] The standard rehabilitation action is used as a benchmark to train the model, ensuring that the actions output by the final model always conform to the standard paradigm verified by clinical practice. For example, for rehabilitation needs of insufficient fast muscle explosive force, the standard action may be rapid contraction for 3 seconds - relaxation for 2 seconds, repeated 10 times. If the action generated by the initial model is contraction for 5 seconds - relaxation for 1 second, the parameters can be corrected through difference comparison to strictly match the recommended action to the standard, ensuring the stability of the training effect.

[0152] Based on the difference training process, the essence is to let the model continuously learn how to generate more standard action recommendations from dynamic evaluation results. For example, when the model repeatedly identifies bias due to insufficient slow muscle endurance requirements, the initial action deviates greatly from the standard action, and the training process focuses on optimizing the feature recognition weight of this type of requirement, so that the model gradually improves the sensitivity to specific rehabilitation requirements, and finally realizes the accurate mapping of evaluation results to action recommendations.

[0153] When training combined with historical rehabilitation information, difference comparison can further refine the adjustment rules of different individuals to the same standard action. For example, for slow muscle endurance training of elderly patients, the standard action intensity needs to be reduced, and the model will gradually master the correction coefficient of the age factor on the strength parameter by learning the difference between the initial action and the standard action adapted to this type of individual, so that the recommended action takes into account individual specialities on the basis of standardization.

[0154] By locating the deviation points (such as action type error, unreasonable intensity parameter) between the initial action and the standard action through difference comparison, the model training does not need to make indiscriminate adjustments to all parameters, but focuses on key dimensions with significant differences, greatly improving training efficiency.

[0155] Each model optimization can correspond to a specific difference point, such as correcting the time parameter due to contraction duration difference and adjusting the action description feature due to force application position deviation, so that the decision logic of the model is interpretable and traceable. This transparent training process facilitates clinicians to verify the reliability of the model, and also provides a clear direction for further improvement of the model, such as tracing a certain type of difference that repeatedly occurs to the feature recognition link.

[0156] In the electrode circuit of the embodiment of the application, a 500Hz weak test current (<10μA) is integrated, and the signal quality is judged by measuring the electrode-skin contact impedance: when the impedance is <5kΩ, the indicator light shows green (signal qualified); when the impedance is 5-10kΩ, a slight vibration is issued (the electrode position needs to be adjusted); when the impedance is >10kΩ, an alarm is triggered (the skin needs to be cleaned or the electrode sheet needs to be replaced), ensuring that the signal path is stable before collection.

[0157] The method for performing dynamic pelvic floor function detection and evaluation in the embodiment of the application realizes accurate detection of each part of the pelvic floor by using a surface electromyography detection component, optimizes the electromyography signals detected by each part through a specific filtering algorithm, improves the accuracy of the final recognition result, and provides targeted basis for subsequent rehabilitation scheme formulation.

[0158] Embodiment two

[0159] Please refer to Figure 5 , Figure 5is a structural schematic diagram of a system for dynamic pelvic floor function detection and evaluation disclosed by an embodiment of the present application. As shown in Figure 5 The system for dynamic pelvic floor function detection and evaluation can include:

[0160] The acquisition module 21 is configured to acquire surface electromyography signals of respective detection areas in each detection stage by a surface electromyography detection assembly, the surface electromyography detection assembly including a first detection assembly, a second detection assembly and a third detection assembly, and the detection areas including a first detection area, a second detection area and a third detection area.

[0161] The conversion module 22 is configured to perform frequency domain conversion on the acquired surface electromyography signals to obtain respective frequency domain signals, and determine a frequency band of interest of the frequency domain signals of each detection area according to attribute characteristics of each detection area, the frequency band of interest including a first frequency band of interest corresponding to the first detection area and a second frequency band of interest corresponding to the second detection area.

[0162] The first calculation module 23 is configured to calculate a first covariance matrix of the first detection area and the second detection area and the third detection area in the first frequency band of interest, construct a first region filter matrix according to the first covariance matrix, perform filter processing on the surface electromyography signals of the first detection area according to the first region filter matrix, and extract the surface electromyography signals after the filter processing to obtain a first region feature group.

[0163] The second calculation module 24 is configured to calculate a second covariance matrix of the second detection area and the first detection area in the second frequency band of interest, determine a second region filter matrix according to the second covariance matrix, perform filter processing on the surface electromyography signals of the second detection area according to the second region filter matrix, and extract the surface electromyography signals after the filter processing to obtain a second region feature group.

[0164] The third calculation module 25 is configured to analyze the surface electromyography signals of the third detection area by using an improved independent component analysis algorithm to obtain fast muscle signals and slow muscle signals in the third detection area, and determine a third region feature group according to the fast muscle signals and the slow muscle signals.

[0165] The analysis module 26 is configured to perform evaluation calculation on the first region feature group, the second region feature group and the third region feature group by using a pelvic floor evaluation algorithm to determine score results of each detection area and a comprehensive score result.

[0166] The method for performing dynamic pelvic floor function detection and evaluation in the embodiment of the present application realizes accurate detection of each part of the pelvic floor by adopting a surface electromyography detection assembly, optimizes the electromyography signals detected by each part through a specific filtering algorithm, improves the accuracy of the final recognition result, and provides a targeted basis for subsequent rehabilitation scheme formulation.

[0167] Embodiment three

[0168] Please refer to Figure 6 , Figure 6 is a structural schematic diagram of an electronic device disclosed by the embodiment of the present application. The electronic device can be a computer, a server, and the like. Of course, in certain cases, it can also be a smart device such as a mobile phone, a tablet computer, and a monitoring terminal, and an image acquisition device with a processing function. As shown in Figure 6 , the electronic device can include:

[0169] a memory 510 storing executable program codes;

[0170] a processor 520 coupled with the memory 510;

[0171] The processor 520 calls the executable program codes stored in the memory 510 to execute part or all of the steps in the method for performing dynamic pelvic floor function detection and evaluation in the embodiment one.

[0172] The embodiment of the present application discloses a computer readable storage medium storing a computer program, wherein the computer program causes a computer to execute part or all of the steps in the method for performing dynamic pelvic floor function detection and evaluation in the embodiment one.

[0173] The embodiment of the present application further discloses a computer program product, wherein when the computer program product runs on a computer, it causes the computer to execute part or all of the steps in the method for performing dynamic pelvic floor function detection and evaluation in the embodiment one.

[0174] The embodiment of the present application further discloses an application publishing platform, wherein the application publishing platform is used to publish a computer program product, wherein when the computer program product runs on a computer, it causes the computer to execute part or all of the steps in the method for performing dynamic pelvic floor function detection and evaluation in the embodiment one.

[0175] In various embodiments of the present application, it should be understood that the size of the serial number of the processes does not mean the inevitable sequence of execution, and the execution sequence of the processes should be determined according to their functions and inherent logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0176] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, i.e. may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0177] In addition, each functional unit in each embodiment of the application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0178] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer accessible memory. Based on such understanding, the technical scheme of the application, essentially or the part that contributes to the prior art, or all or part of the technical scheme, can be embodied in the form of a software product. The computer software product is stored in a memory, and includes a part or all steps of the method for enabling a computer device (which can be a personal computer, a server or a network device, etc., and specifically can be a processor in the computer device) to execute the embodiments of the application.

[0179] In the embodiments provided by the application, it should be understood that "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that the determination of B according to A does not mean that B is determined only according to A, but also can be determined according to A and / or other information.

[0180] Those skilled in the art can understand that part or all of the steps in the various methods of the embodiments can be completed by instructing the relevant hardware by a program, and the program can be stored in a computer readable storage medium, including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk memories, magnetic disk memories, magnetic tape memories, or any other medium capable of carrying or storing data which can be read by a computer.

[0181] The method, system, electronic device and storage medium for dynamic pelvic floor function detection and evaluation disclosed in the embodiments of the present application are described in detail, the principles and implementation manners of the present application are described by applying specific examples in the present application, and the above embodiment description is only used to help understand the method of the present application and the core idea thereof; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manners and application ranges will be changed, and the above description should not be understood as a limitation on the present application.

Claims

1. A method for dynamic pelvic floor function testing and evaluation, characterized in that, include: The surface electromyography (SEMG) detection component is used to acquire surface electromyography (SEMG) signals in corresponding detection areas during each detection stage. The SEMG detection component includes a first detection component, a second detection component, and a third detection component. The detection areas include a first detection area, a second detection area, and a third detection area. The acquired surface electromyography signals are frequency domain converted to obtain corresponding frequency domain signals. The frequency bands of interest for each detection region are determined according to the attribute characteristics of each detection region. The frequency bands of interest include the first frequency band and the second frequency band corresponding to the first detection region and the second detection region, respectively. Calculate the first covariance matrix of the first detection region, the second detection region, and the third detection region in the first frequency band of interest; construct a first region filtering matrix based on the first covariance matrix; filter the surface electromyography (EMG) signal of the first detection region based on the first region filtering matrix; and extract the filtered EMG signal to obtain the first region feature group. Calculate the second covariance matrix between the second detection region and the first detection region in the second frequency band of interest, and determine the second region filtering matrix based on the second covariance matrix; The surface electromyography (EMG) signal of the second detection region is filtered according to the second region filtering matrix, and the filtered EMG signal is extracted to obtain the second region feature group. An improved independent component analysis algorithm is used to analyze the surface electromyography signal in the third detection region to obtain the fast-twitch and slow-twitch signals in the third detection region, and the feature group of the third region is determined based on the fast-twitch and slow-twitch signals. The pelvic floor assessment algorithm is used to evaluate and calculate the feature groups of the first, second, and third regions to determine the scores of each detection region and the overall score.

2. The method for dynamic pelvic floor function testing and evaluation as described in claim 1, characterized in that, The pelvic floor assessment algorithm is used to evaluate and calculate the feature groups of the first, second, and third regions to determine the score results for each detection region and the overall score result, including: A first region feature group is determined in the first detection region of the corresponding detection stage. The first region feature group includes peak electromyographic amplitude, contraction rise time and high frequency signal ratio, so as to determine the region score result of the first detection region based on peak electromyographic amplitude, contraction rise time and high frequency signal ratio. The second region feature group in the second detection area of ​​the corresponding detection stage is determined. The second region feature group includes the standard deviation of electromyography signal, the amplitude ratio at the end of endurance contraction to the initial stage, and the proportion of low-frequency signal. The region score result of the second detection area is determined based on the standard deviation of electromyography signal, the amplitude ratio at the end of endurance contraction to the initial stage, and the proportion of low-frequency signal. The third region feature group of the third detection area in the corresponding detection stage is determined. The peak value of rapid contraction and the endurance parameter of sustained contraction of the third region feature group are used to determine the synergy result of fast and slow muscle groups in the third detection area based on the peak value of rapid contraction and the endurance parameter of sustained contraction. The region score result of the third detection area is determined based on the peak value of rapid contraction, the endurance parameter of sustained contraction and the synergy result. The corresponding comprehensive score is determined based on the regional score results of the first detection area, the regional score results of the second detection area, and the regional score results of the third detection area.

3. The method for dynamic pelvic floor function testing and evaluation as described in claim 1, characterized in that, The detection and evaluation method further includes: The surface electromyography (EMG) signal of the first detection area after filtering is processed to determine the corresponding contraction rise rate and peak decay rate; wherein, the contraction rise rate is the slope of the rise from the baseline to the peak, and the peak decay rate is the ratio of the decrease in amplitude to the time between two adjacent peaks; The surface electromyography (EMG) signals of the second detection area after filtering are processed to determine the corresponding contraction stabilization rate and fatigue decay rate. The contraction stabilization rate is the fluctuation rate of the low-frequency component amplitude. The fatigue decay rate includes the ratio of the amplitude difference between the end and beginning of the endurance contraction to the time. Data processing is performed on the surface electromyography signal of the second detection area after filtering to determine the synergistic contraction synchronization rate, wherein the synergistic contraction synchronization rate is the rate of change of the activation time difference between the high-frequency component of fast muscle and the low-frequency component of slow muscle. The corresponding rate parameters in the first, second, and third detection areas are matched with the rate detection model to determine the current warning level.

4. The method for dynamic pelvic floor function testing and evaluation as described in claim 3, characterized in that, The step of matching the corresponding rate parameters in the first, second, and third detection regions with the rate detection model to determine the current warning level includes: The corresponding rate parameters in the first, second, and third detection areas are matched with the individual and group baselines in the rate detection model to determine the current warning level and the corresponding functional defect results.

5. The method for dynamic pelvic floor function testing and evaluation as described in claim 3, characterized in that, After determining the scores for each detection area and the overall score, the following steps are also included: Output the scores for each region, and match the scores of each region with a pre-set knowledge database to obtain dynamic evaluation results for each detection region; Determine the dynamic assessment results of each testing area of ​​the person to be tested and the rehabilitation training model, wherein the rehabilitation training model includes: a variety of standard rehabilitation training movements obtained by training the rehabilitation training model with the historical rehabilitation training plan of the historical rehabilitation person, and the historical rehabilitation training plan includes multiple rehabilitation training movements. The dynamic evaluation results of each detection area are input into the rehabilitation training model, so that the rehabilitation training model selects standard rehabilitation training actions that match the dynamic evaluation results from the multiple standard rehabilitation training actions, and generates and outputs the final rehabilitation training plan based on the information of the person to be tested and the selected standard rehabilitation training actions. The final rehabilitation training plan is sent to the corresponding display terminal.

6. The method for dynamic pelvic floor function testing and evaluation as described in claim 5, characterized in that, The rehabilitation training model is constructed through the following steps: Obtain historical rehabilitation training plans and dynamic assessment results of various testing areas of historical rehabilitation personnel, wherein the historical rehabilitation training plans include multiple rehabilitation training actions; Feature recognition is performed on the dynamic assessment results to obtain the rehabilitation needs information corresponding to the dynamic assessment results; Based on the rehabilitation needs information and dynamic assessment results, generate rehabilitation training actions corresponding to each type of rehabilitation needs information; Based on the dynamic evaluation results of each detection area and their corresponding rehabilitation training actions, as well as the historical rehabilitation personnel information, the initial rehabilitation training model is trained to obtain the final rehabilitation training model.

7. The method for dynamic pelvic floor function testing and evaluation as described in claim 6, characterized in that, The process of training the initial rehabilitation training model based on the dynamic evaluation results of each detection area and its corresponding rehabilitation training actions, as well as the historical rehabilitation personnel information, to obtain the final rehabilitation training model includes: The dynamic evaluation results of each detection area and the historical rehabilitation personnel information are input into the initial rehabilitation training model to generate the initial rehabilitation action corresponding to each dynamic evaluation result. Based on the dynamic evaluation results of each detection area, determine the differences in movement and intensity between the corresponding initial rehabilitation movement and the corresponding standard rehabilitation movement; Based on the differences in movement and intensity between the corresponding standard rehabilitation movements, the initial rehabilitation training model is trained to obtain the final rehabilitation training model.

8. A system for dynamic pelvic floor function testing and evaluation, characterized in that, include: Acquisition module: used to acquire surface electromyography (SEMG) signals of corresponding detection areas in each detection stage through surface electromyography (SEMG) detection components, the surface electromyography (SEMG) detection components including a first detection component, a second detection component, and a third detection component; the detection areas include a first detection area, a second detection area, and a third detection area; The conversion module is used to perform frequency domain conversion on the acquired surface electromyography signals to obtain corresponding frequency domain signals. It determines the frequency band of interest for each detection region based on the attribute characteristics of each detection region. The frequency band of interest includes the first frequency band of interest and the second frequency band of interest corresponding to the first detection region and the second detection region, respectively. The first calculation module is used to calculate the first covariance matrix of the first detection region, the second detection region, and the third detection region in the first frequency band of interest; construct a first region filtering matrix based on the first covariance matrix; filter the surface electromyography (EMG) signal of the first detection region based on the first region filtering matrix; and extract the filtered EMG signal to obtain the first region feature group. The second calculation module is used to calculate the second covariance matrix between the second detection region and the first detection region in the second frequency band of interest, and to determine the second region filtering matrix based on the second covariance matrix. The surface electromyography (EMG) signal of the second detection region is filtered according to the second region filtering matrix, and the filtered EMG signal is extracted to obtain the second region feature group. The third calculation module is used to analyze the surface electromyography signal of the third detection area using an improved independent component analysis algorithm to obtain the fast muscle signal and slow muscle signal in the third detection area, and to determine the feature group of the third area based on the fast muscle signal and slow muscle signal. Analysis module: Used to evaluate and calculate the first region feature group, the second region feature group, and the third region feature group using the pelvic floor assessment algorithm to determine the score results of each detection region and the comprehensive score result.

9. An electronic device, characterized in that, include: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the method for dynamic pelvic floor function detection and evaluation as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program causes a computer to perform the method for dynamic pelvic floor function detection and evaluation as described in any one of claims 1 to 7.

Images