A mobile terminal based pelvic floor muscle training system

Abstract

The present application relates to the technical field of intelligent health management, in particular to a pelvic floor muscle training system based on a mobile terminal, which comprises a signal coordination acquisition module, a multi-parameter linkage identification module, a three-stage action decomposition module, a time sequence voice guidance module and an execution deviation evaluation module. Through synchronous time sequence analysis of the pelvic floor muscle electrical signal and the heart rate signal during the training process, the present application excavates the coordinated changes of signal mutation nodes, combines multi-source physiological parameter linkage identification and induction of action stage morphological characteristics, decomposes the training action into a multi-stage process, dynamically maps the signal nodes and the time sequence of voice guidance, realizes precise docking of step-by-step voice instructions and action states, quantizes the performance of each stage of the training process into data characteristics, intuitively evaluates the training performance differences through the time sequence deviation degree, and enhances the scientificity and continuity of rehabilitation guidance through the step-by-step feedback mechanism, thereby providing data-based support for individual rehabilitation path adjustment and progress tracking.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent health management, and particularly relates to a pelvic floor muscle training system based on a mobile terminal. BACKGROUND

[0002] Intelligent health management involves using wearable devices, mobile terminals, sensor networks, and health informatics to continuously monitor, collect data, and evaluate and analyze individual physiological parameters, behavior habits, and lifestyles, so as to realize multi-level health management services such as disease prevention, health promotion, and rehabilitation guidance, and is widely used in many fields such as chronic disease management, sports rehabilitation, elderly care, and psychological adjustment; among them, the traditional pelvic floor muscle training system refers to offline rehabilitation training services carried out by the hospital physical therapy department, mainly guided by rehabilitation therapists to train patients to contract muscles, and often uses visual prompts, language guidance, and electrical stimulation instrument assistance to repeatedly train the pelvic floor muscle group of the patient, and usually relies on fixed places and full-time staff to carry out services.

[0003] The prior art relies on manual observation and traditional instruments for pelvic floor training, cannot automatically collect and link multiple physiological signals, cannot finely disassemble the process of the training action stage, cannot accurately identify the key nodes, mainly relies on perception or static data review for the action execution status, lacks data features for the training performance, cannot quantize the differences between stages, has a fault in the process feedback mechanism, lacks continuous step-by-step assistance and scientific guidance in the rehabilitation training process, and the data tracing and progress evaluation are not coherent, so that the user cannot obtain timely, detailed, and individualized rehabilitation adjustment support. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art and to provide a pelvic floor muscle training system based on a mobile terminal.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: a pelvic floor muscle training system based on a mobile terminal, the system comprising:

[0006] The signal cooperative collection module analyzes the pelvic floor muscle electrical signal and the heart rate signal based on the training action process, judges the synchronous fluctuation of the two on the same time axis, compares the order of sudden changes, determines the synchronous sudden change nodes and counts the number, and obtains the synchronous sudden change node characteristics;

[0007] The multi-parameter linkage identification module filters the signal synchronization nodes based on the synchronous sudden change node characteristics, compares the collection time of the muscle electrical signal, the photoelectric heart rate, and the three-axis displacement signal, analyzes the corresponding relationship, judges the node linkage performance, and obtains the three-parameter linkage feature group;

[0008] The three-segment action decomposition module judges the rising, maintaining and falling stages of the displacement signal based on the three-parameter linkage feature group, compares the signal changes in the difference stages through continuity analysis, induces the difference performance in each stage, and obtains the action segmentation morphology feature;

[0009] The timing voice guide module compares the signal nodes and the voice prompt trigger time points based on the action segmentation morphology feature, analyzes the key node and the voice instruction response timing, screens the optimal node of response difference, establishes the corresponding relationship, and obtains the voice guide timing feature.

[0010] The synchronous mutation node feature includes the node occurrence time, the synchronous distribution type and the associated signal number, the three-parameter linkage feature group includes the multi-signal coupling type, the event discrimination category and the linkage amplitude identifier, the action segmentation morphology feature includes the action structure parameter, the segmentation morphology category and the action cycle marker, and the voice guide timing feature includes the trigger signal type, the segmented voice node and the guide timing label.

[0011] The signal cooperative acquisition module includes:

[0012] The timing fluctuation analysis submodule analyzes the collected pelvic floor muscle electrical signals and heart rate signals based on the training action process, identifies the continuous change segments of each signal in the training process, calculates the change rate of each segment signal, judges the trend change of the signal on the time axis, optimizes the positioning of the signal inflection point, and obtains the timing change trend quantity;

[0013] The mutation relationship judgment submodule screens the key nodes with obvious signal jump based on the timing change trend quantity, judges the order of occurrence of mutations of various signals, analyzes the primary and secondary relationship of the mutations, and calculates the time difference between the nodes, and obtains the mutation priority information;

[0014] The synchronous node identification submodule screens the nodes with close time difference based on the mutation priority information, calculates the number and distribution of synchronous mutation nodes, compares the synchronous change of each signal type, analyzes the coupling characteristics of the signals in each time period, and obtains the synchronous mutation node feature.

[0015] The multi-parameter linkage recognition module includes:

[0016] The signal synchronization extraction submodule analyzes the collected muscle electrical signals, heart rate signals and displacement signals based on the synchronous mutation node feature, judges the synchronous change trend of each signal on the time axis, screens the signal combination that changes synchronously in the same time period, and obtains the synchronous node set;

[0017] The node matching discrimination sub-module compares the change direction and fluctuation of each signal in the recording period based on the synchronization node set, judges the consistency of the difference signal on the same node, analyzes the time sequence and trend corresponding relationship, and obtains a node coupling distribution group;

[0018] The linkage feature generation sub-module judges the change combination of each signal based on the node coupling distribution group, analyzes the synchronization and linkage feature of the signal under the key node, filters the change type with correlation, and obtains a three-parameter linkage feature group.

[0019] The three-stage action decomposition module comprises:

[0020] The stage recognition sub-module analyzes the cooperative performance of each signal based on the three-parameter linkage feature group, judges the rising, stable and falling trend of the displacement signal curve in the difference time period, optimizes the start and end nodes of the stage division, filters the effective interval, and obtains a stage boundary time sequence.

[0021] The signal comparison sub-module compares the displacement signal and the electromyographic signal trajectory of each stage based on the stage boundary time sequence, analyzes the fluctuation amplitude and change consistency of the two types of signal trajectories in each stage, calculates the fluctuation difference between the trajectories in each stage, and obtains a stage signal difference amplitude range through fluctuation difference collection.

[0022] The form induction sub-module judges the stage number with the most performance based on the stage signal difference amplitude range, analyzes the change process of the displacement signal and the electromyographic signal response in the stage, induces the difference characteristics of the three stages, and obtains an action segmentation form feature.

[0023] The time sequence voice guidance module comprises:

[0024] The voice node alignment sub-module compares the sequence of each stage signal node and voice prompt in time sequence based on the action segmentation form feature, combines the corresponding time point of the voice prompt, filters the signal nodes and voice prompts with similar time for pairing, and obtains a node alignment time interval group.

[0025] The response difference calculation sub-module compares the time interval between each pair of signal nodes and voice prompts based on the node alignment time interval group, calculates the deviation of the signal and voice synchronous change, and obtains an average response difference amplitude.

[0026] The guidance matching construction sub-module optimizes the pairing result of each stage signal node and voice prompt based on the average response difference amplitude, filters the node combination with the optimal pairing effect, establishes the one-to-one corresponding relationship between the signal node and the voice prompt, and obtains a voice guidance time sequence feature.

[0027] The system further comprises:

[0028] The execution deviation evaluation module calculates the difference between the start and end time periods of each stage of the user's actual action based on the voice guidance timing feature, analyzes the execution time, weights the stage difference, judges the consistency with the target, and obtains the execution deviation index;

[0029] The execution deviation index includes a timing deviation level, an action stage evaluation, and an execution consistency classification.

[0030] The execution deviation evaluation module includes:

[0031] The training period extraction submodule matches the instruction time points marked by the voice prompt with the start and end nodes of the user's action one by one based on the voice guidance timing feature, extracts the action start and end time corresponding to each stage by judging the node coincidence degree and the timing closeness degree, and obtains the stage execution time period sequence.

[0032] The stage difference calculation submodule compares the stage execution time period sequence with the preset action timing arrangement, focuses on the distribution and difference of the start and end nodes of each stage action in the time sequence, determines the timing error type of each stage by analyzing the node offset performance before and after, and obtains the time matching offset set.

[0033] The deviation degree summary submodule judges the relationship between the rhythm characteristics in the action structure parameters based on the time matching offset set, analyzes the phenomena of advance, lag, and rhythm asynchronization exhibited by the actual action in the difference stage, and integrates the offset performance of all stages to obtain the execution deviation index.

[0034] Compared with the prior art, the advantages and positive effects of the present application are:

[0035] In the present application, through the synchronous timing analysis of the pelvic floor muscle electrical signal and the heart rate signal during the training process, the cooperative change of the signal mutation node is excavated, the multi-source physiological parameter linkage recognition is combined with the induction of the action stage form feature, the training action is disassembled into a multi-stage process, the signal node and the time sequence of the voice guidance are dynamically mapped, the precise docking of the step-by-step voice instruction and the action state is realized, the training performance of each stage is quantized into data features, the training performance difference is intuitively evaluated through the timing deviation degree, and the step-by-step feedback mechanism enhances the scientificity and continuity of the rehabilitation guidance, thereby providing data basis support for individual rehabilitation path adjustment and progress tracking. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The system flowchart of the present application is shown in the figure;

[0037] Figure 2 The flowchart of the signal cooperative acquisition module in the present application is shown in the figure;

[0038] Figure 3 Flow chart of the multi-parameter linkage identification module in the application;

[0039] Figure 4 Flow chart of the three-segment action decomposition module in the application;

[0040] Figure 5 Flow chart of the time sequence voice guide module in the application;

[0041] Figure 6 Flow chart of the execution deviation evaluation module in the application. DETAILED DESCRIPTION

[0042] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0043] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0044] Embodiment, please refer to Figure 1 The present application provides a technical solution: a pelvic floor muscle training system based on a mobile terminal, comprising:

[0045] The signal cooperative acquisition module analyzes the collected pelvic floor muscle electrical signal and heart rate signal based on the training action process, judges the time sequence fluctuation thereof on the same time axis, compares the order of sudden changes of the collected signals, determines the nodes of simultaneous sudden changes, and counts the number of synchronous sudden change nodes to obtain the synchronous sudden change node features;

[0046] The multi-parameter linkage identification module compares the acquisition time of the Bluetooth synchronous muscle electrical signal collector, the photoelectric heart rate belt and the three-axis displacement sensor based on the synchronous sudden change node features after screening the signal synchronization nodes, analyzes the mutual correspondence of each type of signal under each node, judges the signal linkage performance between the nodes, and obtains a three-parameter linkage feature group;

[0047] The three-segment action decomposition module judges the rising stage, maintaining stage and falling stage of the displacement signal based on the three-parameter linkage feature group, analyzes the continuity of the signal in each stage, compares the signal change characteristics in each stage, and induces the difference stage performance to obtain the action segmentation morphology feature;

[0048] The timing voice guide module compares the signal nodes in each stage with the trigger time points of the voice prompts based on the action segmentation morphology feature, analyzes the key nodes in each stage and the response timing of the voice instructions, selects the nodes with the optimal difference between the signal and the instruction time point response, establishes a corresponding relationship, and obtains the voice guide timing feature;

[0049] The execution deviation evaluation module calculates the difference between the voice guide timing feature and the actual training action of the user in each stage, analyzes the actual execution time in each stage, judges the consistency of the action execution and the target through the weighted processing of the difference in each stage, and obtains the execution deviation index.

[0050] The synchronization mutation node feature includes the node occurrence time, synchronization distribution type and associated signal number, the three-parameter linkage feature group includes the multi-signal coupling type, event discrimination category and linkage amplitude identifier, the action segmentation morphology feature includes the action structure parameter, segmentation morphology category and action cycle marker, the voice guide timing feature includes the trigger signal type, segmented voice node and guide timing label, and the execution deviation index includes the timing deviation level, action stage evaluation and execution consistency classification.

[0051] In the signal cooperative acquisition module, the pelvic floor muscle electrical signal refers to the pelvic floor muscle electrical activity signal acquired by the surface electromyographic sensor in real time, reflecting the contraction and relaxation physiological process of the pelvic floor muscle group; the heart rate signal refers to the heart rate signal obtained by using an electrocardiogram or photoelectric sensor, reflecting the activity state of the autonomic nervous system; the timing fluctuation refers to the dynamic process of the change of the signal intensity or value on the time axis, i.e., the up and down fluctuation of the signal with time; the order of mutation refers to which item appears first in the signal acquisition, and the priority and causal relationship between different physiological events can be judged by comparison; the nodes with simultaneous mutation refer to the nodes where multiple signals are mutated (such as synchronous changes in muscle contraction and heart rate) at the same time point or within a very short time, which are used to mark the time of the key physiological event.

[0052] In the multi-parameter linkage identification module, the Bluetooth synchronous electromyography collector refers to a special device that collects and synchronously transmits electromyography signals to a mobile terminal through Bluetooth wireless mode; the photoelectric heart rate belt refers to a wearable device for measuring heart rate by using photoelectric sensing technology (such as PPG), which is usually wound around the chest or wrist and used for real-time heart rate monitoring; the three-axis displacement sensor refers to a sensor for detecting displacement changes in three-dimensional space, which can reflect physical motion information such as user actions and body positions; the mutual correspondence refers to the correspondence and matching between various signals (electromyography, heart rate, displacement) in time at the same event node; the signal linkage performance refers to whether multiple physiological signals change coordinately at the event node and whether they show synchronization or certain linkage rules, which are used to judge the coupling characteristics of actions and physiological states.

[0053] In the three-stage action decomposition module, the displacement signal refers to a physical signal collected by the three-axis displacement sensor to reflect the change of user body position or action; the signal change characteristics of each stage refer to the continuity and change rate of displacement or related physiological signals in different stages (such as rising, maintaining, and falling) of action execution; the difference stage performance refers to the typical performance of each stage by comparing and analyzing the change mode of signals in each stage, for example, the difference between the action force stage and the relaxation stage.

[0054] In the timing voice guidance module, the signal node of each stage refers to the physiological or displacement signal feature node at the start and end time points and the turning points of each stage in the three-stage cycle of action; the trigger time point refers to the specific time node at which the voice prompt module issues a prompt instruction, which needs to be matched with the signal node; the voice instruction response timing refers to the actual timing relationship of the user action signal in response after the voice prompt is issued, which is used to evaluate the timeliness of guidance; the node with optimal response difference refers to the key node with the smallest time difference and the closest cooperation between the signal node and the voice prompt trigger time; the establishment of the correspondence refers to the one-to-one matching of the signal stage characteristics and the voice prompt time, which realizes the linkage between the step-by-step voice guidance and the action state; the voice guidance timing feature is the time response data matched one by one between the "voice prompt" and the "actual action node" of the user, which is used to realize the accuracy, personalized adjustment, and quantitative evaluation of training feedback of voice guidance, and provides a basis for training data analysis and individualized rehabilitation path optimization.

[0055] In the execution deviation evaluation module, the difference of the start and end time period of each stage refers to the difference between the standard stage start and end time determined by the system and the time of the actual action stage completed by the user; the actual execution time refers to the time used by the user to actually complete each stage of action, which can be calculated from the start and end nodes detected by the signal; the weighted processing refers to the comprehensive processing of the time difference of each stage according to a certain proportion, which reflects the deviation degree of the overall action execution; the target consistency refers to the evaluation of the consistency of the actual action of the user and the preset target action in terms of time, rhythm, etc.

[0056] Please refer toFigure 2 The signal cooperative acquisition module comprises:

[0057] The timing fluctuation analysis submodule performs a process based on the training action, analyzes the collected pelvic floor muscle electrical signal and heart rate signal, identifies continuous change segments of each signal in the training process, calculates the change rate of each segment of signal, judges the trend change of the signal on the time axis, optimizes the positioning of the signal inflection point, and obtains the timing change trend quantity;

[0058] Identify the training period the user is currently in, group the pelvic floor muscle electrical signal and heart rate signal according to the sampling frequency respectively, and generate a data segment corresponding to the time axis in combination with the action start time. Each segment of signal is composed of multiple continuous time sampling points. After each group of signal sequences is input, the signal amplitude difference between adjacent sampling points is extracted, and then the change rate value is calculated according to a fixed time interval. Subsequently, a rate sequence graph is established according to the time sequence, and the change trend is analyzed continuously. If the change directions of three or more sampling points are the same, i.e., the values are all increasing or decreasing, the segment is marked as a continuous change segment. For example, in a one-time anal contraction training, the muscle electrical signal amplitude of a certain user continuously rises from the 1.0th second to the 1.6th second during the training, and the change amplitude of each sampling point is relatively stable. It is judged that this segment is an “upward segment”. Then, the change amplitude range of each segment is counted. If the difference range exceeds a certain threshold, it is marked as a “rapid change segment”, otherwise it is a “slow change segment”. For example, if the total amplitude difference of the muscle electrical signal in the upward segment reaches 1.2 and the time duration is less than 0.6 seconds, the segment is identified as a rapid upward segment. Then, the trend signs of adjacent change segments are compared. If the previous segment rises and the next segment falls, and the difference between the end and the start of the two segments reaches the inflection point judgment requirement, the intersection sampling point between the segments is marked as a signal inflection point. In the implementation process, if the muscle electrical signal reaches a peak value at 1.6 seconds and then immediately decreases, and the decrease amplitude is obvious, the inflection point position is accurately positioned at 1.6 seconds. All identified change segments and inflection point times are recorded into a database as the timing change trend quantity output of the training process.

[0059] The mutation relationship judgment submodule is based on the timing change trend quantity, screens key nodes with obvious signal jump, judges the order of occurrence of various signal mutations, analyzes the primary and secondary relationship of the mutations, and calculates the time difference between the nodes to obtain mutation priority information;

[0060] Extract all trend segments and signal inflection points from the recorded electromyography and heart rate signals, and determine whether a mutation event is formed by calculating the amplitude difference before and after the change segment. If the change amplitude exceeds the set threshold, it is marked as a mutation node. For example, the electromyography signal changes by 1.5 at 1.8 seconds, and the heart rate signal changes by 0.9 at 1.85 seconds, both of which exceed the corresponding mutation recognition threshold, and are identified as electromyography mutation nodes and heart rate mutation nodes. Then, the mutation nodes are timestamped, and each type of mutation event is sorted by time and compared for time difference. For example, the time difference between the two nodes is 50 milliseconds, which is identified as a time-related event. Then, the primary and secondary nodes are determined based on the change amplitude. The node with a larger change amplitude is considered the primary signal, and the node with a smaller change amplitude is considered the secondary signal. In this example, the electromyography signal is the primary signal, and the heart rate signal is the secondary signal, which is marked as an "electromyography dominant mutation event". The time difference, primary and secondary order, and signal source number of all event pairs are included in the mutation priority information. If multiple similar nodes are identified, for example, electromyography mutations occur about 30 to 60 milliseconds before heart rate mutations during a muscle training, and the amplitude difference is significant, then this pattern is classified as a "high priority electromyography dominant event pattern", and a mutation priority information table is generated for the entire training period.

[0061] The synchronization node identification submodule filters nodes with close time differences based on the mutation priority information, calculates the number and distribution of synchronization mutation nodes, compares the synchronization changes of each signal type, analyzes the coupling characteristics of the signals in each time period, and uses the formula:

[0062] ;

[0063] Get synchronization mutation node characteristics wherein, represents the time interval between two signal mutations in the first group of nodes, represents the number of signal types in the first group, represents the time standard deviation of each group of mutation points, represents the absolute value sum of the time difference of each signal mutation, indicates the total number of mutation node groups.

[0064] Synchronization mutation node characteristics refer to the characteristic statistical indicators formed when different physiological signals (such as pelvic floor electromyography and heart rate signals) change significantly at almost the same time during the same training period. This characteristic is used to quantitatively describe the synchronization of mutation of different signals during the training action period, reflect the synchronization and coupling tightness of different signals at key nodes, and be used for further analysis and evaluation of training action quality and physiological response synchronization in the subsequent stage;

[0065] The mutation node groups with time intervals in the range of ±0.15 seconds are screened out, the number of signal types involved in each group of mutation nodes is counted, and the time interval between two mutation signals in each group is recorded , the time standard deviation between mutation points , and the absolute value sum of the time difference between each mutation signal To ensure the consistency of the dimensions of each participating item, the data with non-time dimensions are normalized and then involved in subsequent calculations. The number of signal types after normalization is: in group 1, in group 2, in group 3; The normalized standard deviation is: , , ; The normalized sum of mutation time difference is: , , When substituting the formula for multiple group operations, first calculate the numerator part:

[0066] ;

[0067] Substitute each group calculation item of the denominator in turn:

[0068] Group 1: ;

[0069] Group 2: ;

[0070] Group 3: ;

[0071] Take the average value of the denominator: ;

[0072] Substitute the main formula to get: ;

[0073] The results show that the calculated synchronous mutation node characteristics , after comparison with the synchronous evaluation benchmark interval , fall near the upper limit of the interval, indicating that the time distribution between the mutation nodes has a strong degree of concentration and signal coupling. This value reflects the clear synchronous mutation performance of multiple signals at key nodes during the training process, the linkage of mutation events between signals tends to be consistent, and the response delay of signal variation is short. Therefore, this result can be used as a quantitative basis for the characteristics of synchronous mutation nodes, and used as a standard reference for subsequent modules of voice guidance optimization and action stage decomposition strategy based on node synchronization.

[0074] Please refer to Figure 3 , the multi-parameter linkage recognition module includes:

[0075] The signal synchronization extraction submodule analyzes the collected electromyographic signals, heart rate signals and displacement signals based on the synchronization mutation node characteristics, judges the synchronization change trend of each signal on the time axis, screens the signal combination that changes synchronously in the same time period, and obtains a synchronization node set;

[0076] All the previously identified synchronization mutation node list is called, which contains the time label and amplitude change information of the electromyographic signal, heart rate signal and displacement signal mutation occurring in the training process at different time points. The time axis of each type of signal data is unified, all signals are aligned in the same training period, the corresponding time accuracy is uniformly set to 0.01 seconds, and the nodes are arranged in order of the time label of the mutation node. It is judged whether the other two types of signals also mutate in the window range of ±0.05 seconds at this time point. If the amplitude of the three types of signals changes more than the set mutation recognition threshold in this window, the node is identified as a preliminary synchronization node. For example, the electromyographic signal rises by more than 1.0, the heart rate signal drops by 0.7, and the displacement signal moves upward by 0.9 at 7.32 seconds, and the time interval is within 0.04 seconds. The time point is identified as a synchronization mutation node. Then all the candidate synchronization nodes are further compared to judge whether the change direction of the three types of signals is consistent, that is, whether the signals all rise, fall or have the same trend at this time. If the electromyographic signal and the displacement signal rise at the same time, and the heart rate slightly decreases but does not exceed the direction consistency threshold, it is identified as a "partially consistent" node. Only when the three types of signals meet the consistency determination conditions in both the change direction and the change rate level, can it be included in the final synchronization node set. The direction consistency determination is based on the direction symbol judgment, and the rate level difference does not exceed the set upper limit of one level as a reference. For example, the rate level is set to three levels: slow, medium and fast. If two of the three types of signals are medium and one is fast, it meets the condition, otherwise the node is excluded. Finally, the nodes that meet all the conditions are output as the synchronization node set.

[0077] The node matching discrimination submodule compares the change direction and fluctuation of each signal in the recording period based on the synchronization node set, judges the consistency of the difference signals at the same node, analyzes the time sequence and trend correspondence, and obtains a node coupling distribution group.

[0078] The synchronous nodes are taken out one by one, the change direction and fluctuation amplitude of the electromyographic signal, the heart rate signal and the displacement signal under the node in the recording period are read, the continuous change data segment of each type of signal within 0.2 seconds before and after the node is extracted, and the amplitude difference between the start point and the end point is compared to judge the change trend. If the change direction of the three signals in the same node segment is increasing or decreasing, it is marked as consistent direction event. If the direction of a signal is different from the other two, it is marked as a difference signal. The amplitude difference of the difference signal and the other two types of signals is calculated. If the difference is within the set matching error threshold, for example, the fluctuation difference tolerance between electromyography and heart rate is ±0.4, and if the measured difference is 0.35, it is considered as fluctuation matching. Through the above process, the event nodes of each signal that show consistency under the same node are screened out, and the direction consistency level, amplitude difference size, fluctuation correspondence degree and other data are combined into a comprehensive node feature group. Further analyze the start time difference of the change trend of each signal at the node. If the starting time is within 0.1 seconds, it is marked as high time sequence correspondence. If it exceeds 0.2 seconds, it is determined as low correspondence. According to this rule, all synchronous nodes are traversed to form three types of indexes of direction consistency, amplitude coincidence and time sequence response difference of each node. By setting a comprehensive index threshold, for example, at least two of the three indexes of a node meet the set reference standard, the node can be determined as an effective coupling node. All nodes that pass the determination are classified into a node coupling distribution group.

[0079] The linkage feature generation submodule judges the change combination of each signal based on the node coupling distribution group, analyzes the synchronization and linkage characteristics of the signals at the key nodes, screens the change types with relevance, and obtains a three-parameter linkage feature group;

[0080] The change type combination of the electromyographic, heart rate and displacement signals in each node is analyzed. For each node, the change path type of the signal combination in the corresponding time period is extracted, such as the combination of electromyographic rising, heart rate falling and displacement rising, which is classified as the "contraction excitation-autonomic regulation-movement propulsion" type. Each type of combination is recorded as an independent linkage mode number, and the occurrence frequency and time period distribution of each combination in the complete training period are counted to determine whether there is relevance. If a certain signal combination appears in the initial, middle and final segments of the training, and the interval time of each appearance does not exceed the set upper time limit, such as 10 seconds, the combination is marked as a periodic linkage mode. Then, the consistency ratio of the change direction of the three types of signals, the amplitude synchronization rate and the start response difference are calculated to determine whether the combination meets the linkage condition. If at least two of the three meet the set threshold, for example, the consistency ratio is higher than 70% and the amplitude synchronization rate is higher than 65%, it is determined that it has relevance linkage characteristics. All node combinations that meet the above standards are identified as three-parameter linkage events. Finally, a three-parameter linkage feature group is constructed and outputted to support the action segmentation structure recognition and speech guidance strategy distribution in the subsequent training process.

[0081] Please see Figure 4 The three-stage motion decomposition module includes:

[0082] The stage identification submodule is based on the three-parameter linkage feature group to analyze the coordinated performance of each signal, determine the rising, stable and falling trends of the displacement signal curve in different time periods, optimize the start and end nodes of stage division, screen the effective intervals, and obtain the stage boundary time series.

[0083] Iterate through all linked feature nodes, extracting the trend sequences of electromyography (EMG), heart rate, and displacement signals within a 2-second range before and after each node. Establish fluctuation curve groups for the three types of signals according to the sampling time axis. Prioritize judging the numerical fluctuation of the displacement signal during this period, marking all upward trend segments as candidates for the rising phase, stable fluctuation segments as candidates for the maintenance phase, and continuously decreasing segments as candidates for the falling phase. The judgment criterion is whether the direction of the difference between consecutive sampling points remains consistent. Further screening of the trend nature is combined with the magnitude of the rate difference between each segment. For example, in a user's contraction action, if the displacement signal continuously rises between 2.1 and 2.9 seconds with an amplitude change exceeding 1.0, it is marked as a candidate segment for the rising phase. Immediately following, from 2.9 to 3.3 seconds, the signal fluctuation amplitude is less than 0.1, and the segment is judged as a candidate segment for the falling phase. The segment is defined as a stable segment. Then, the displacement signal continuously decreases from 3.3 seconds to 4.0 seconds, which is determined to be a declining segment. After identifying the trend of each segment, the consistency of the three types of signals in this segment is further compared. If the electromyographic signal rises synchronously and the accompanying change in the heart rate signal is within the threshold, the segment is confirmed as a valid segment. The start and end times of all valid segment nodes are extracted for segment boundary calculation. Then, all action cycles are numbered according to the start and end times. Each action cycle is divided into three segments. Then, segments whose amplitude fluctuation does not reach the set recognition benchmark value are screened out from all segments. For example, segments with fluctuation amplitude less than 0.3 or segment duration less than 0.4 seconds will not be included in the segment sequence. Finally, all segment numbers that meet the amplitude threshold and duration requirements are confirmed, and the start and end times are combined to form a segment boundary time series for output.

[0084] The signal comparison submodule compares the displacement signal and electromyographic signal trajectory at each stage based on the stage boundary time series, analyzes the fluctuation amplitude and consistency of the two signal trajectories at each stage, and uses the following formula:

[0085] ;

[0086] Calculate the fluctuation differences between trajectories at each stage By aggregating fluctuation differences, the amplitude of stage signal differences is obtained, where... Indicates the first Phase 1 displacement signal trajectory, representing the equalization level of all displacement signal trajectories within the phase, representing the muscle electrical signal amplitude within the phase, representing the muscle electrical signal amplitude within the phase, representing the number of samples of the displacement signal trajectory or the muscle electrical signal amplitude within the phase;

[0087] The inter-trajectory fluctuation difference refers to the degree of dispersion between the change amplitude of the user's displacement signal in the process of performing each phase action and the overall average performance, and the response strength of the muscle electrical signal within the phase, comprehensively reflecting the synchronous fluctuation and cooperative performance of different signals in the process of action;

[0088] The original displacement velocity and muscle electrical amplitude signals obtained by sampling are grouped according to the phase number, and each group of normalized signal sequences is processed in sections within each phase. Then, the degree of dispersion between the normalized displacement signal sample and its equalization level within the phase is analyzed, the change amplitude is measured by using the square difference of the trajectory, the absolute response level of the normalized muscle electrical signal amplitude within the phase is superimposed, and on this basis, the formula is calculated. Taking phase 2 as an example, the normalized sampling data is as follows:

[0089] Displacement signal sequence: ;

[0090] Muscle electrical signal sequence: ;

[0091] First, calculate :

[0092] ;

[0093] Then calculate the sum of the square differences of the numerator part:

[0094] ;

[0095] ;

[0096] Calculate the absolute value of the denominator part of the muscle electrical response strength and add 1:

[0097] ;

[0098] Substitute the formula to calculate:

[0099] ;

[0100] The result shows that the inter-trajectory fluctuation difference in phase 2 is The value is compared with the fluctuation difference grading interval, if the set normalized fluctuation difference grading interval is:

[0101] low fluctuation section ( );

[0102] medium fluctuation section ( );

[0103] high fluctuation section ( ), then the value is in the medium fluctuation section, indicating that the displacement trajectory in this stage has a certain degree of change compared with the equilibrium level, and does not show too violent or extreme fluctuation behavior. Combined with the continuous response of the electromyographic signal amplitude, it is deduced that there is a certain fluctuation in the action rhythm in this stage, the signal cooperativity is relatively obvious, but still maintains in the controllable range of structure, so the numerical result can be used as one of the basic characteristic indexes for determining the stage type to support the construction and classification basis of the stage signal difference amplitude.

[0104] The morphology induction submodule judges the key stage number based on the stage signal difference amplitude, analyzes the change process of the displacement signal and the electromyographic signal response in the stage, induces the difference characteristics of the three stages, and obtains the action segmentation morphology characteristics;

[0105] The start and end time range of each stage, and the corresponding displacement signal and electromyographic signal change process are extracted from the identified stage sequence. First, the total change value, change direction and duration of the displacement signal in each stage are calculated, then the change synchronization of the electromyographic signal in the stage is counted, for example, in the rising stage, if the total displacement signal changes more than 1.5, the average rate is greater than 0.3 per second, and the duration is 1.2 seconds, it is recorded as a high-intensity change stage. Then extract the fluctuation amplitude and initial response time of the electromyographic signal in the same time period, judge the response delay between them, such as delay less than 0.2 seconds and electromyographic fluctuation amplitude more than 0.8, determine that the stage is a muscle-electricity-displacement coupling enhancement section. Then compare the change of heart rate signal in the stage, if the heart rate fluctuation direction is consistent with the electromyographic trend, and the amplitude change is more than 0.5, mark the stage as a three-part linkage highlighted section. The characteristic induction operation of each stage is completed in turn, and all stages are archived according to three states: rising period, maintenance period and falling period. The corresponding displacement response amplitude interval, electromyographic average change value and heart rate fluctuation amplitude are recorded in each state to form an action stage morphology database. The morphology characteristics of the above three stages are arranged and combined to determine which combination morphology appears most frequently, has the highest similarity in change path, and has the highest coincidence degree of signal fluctuation characteristics in the training period. Confirm the repeated structure as the core action segmentation morphology characteristics, label the number and generate complete action segmentation morphology characteristics output.

[0106] Please refer to Figure 5The timing voice guide module comprises:

[0107] The voice node alignment sub-module compares the order of each stage signal node and voice prompt in time sequence based on the action segment form feature and the corresponding time point of the voice prompt, filters the signal nodes with similar time and the voice prompts for pairing, and obtains a node alignment time interval group;

[0108] The starting and ending time of each action stage and the list of stage corresponding signal nodes output by the three-section action decomposition module are called, and then the preset prompt voice playing time point sequence in the voice prompt module is read. The trigger time point of each voice prompt and the time point of the signal node in the action stage are compared item by item, and the time interval of each voice time point and all signal nodes within a range of not more than 0.5 seconds before and after it is recorded. If the absolute time difference between the signal node time in a stage and the voice prompt is less than the set alignment tolerance threshold, for example, the tolerance is set to 0.3 seconds, the signal node is the candidate pairing object. The time interval of all candidate pairing objects is sorted, and the signal node with the shortest time interval is paired with the voice prompt, for example, in the maintenance stage, the voice prompt is played at 4.2 seconds, and there are signal nodes at 4.05 seconds, 4.25 seconds and 4.45 seconds in this stage. The calculation of each time difference is 0.15 seconds, 0.05 seconds and 0.25 seconds. Because 0.05 seconds is the smallest, the signal node at 4.25 seconds is paired with the voice prompt. Then this operation is completed for each stage in turn to generate a one-to-one pairing result between the signal node and the voice prompt in each stage, and the time interval information of all paired nodes is recorded as a node alignment time interval group. Each record in the group contains stage number, signal node time, voice prompt time and time difference value, for example, the output content is marked: stage two, signal node 4.25 seconds, voice prompt 4.2 seconds, interval 0.05 seconds. This data is used for subsequent analysis of the matching degree of pairing.

[0109] The response difference calculation sub-module compares the time interval between each pair of signal nodes and voice prompts based on the node alignment time interval group, calculates the deviation of the synchronous change of the signal and the voice, and uses the formula:

[0110] ;

[0111] The average response difference amplitude is obtained , wherein, represents the total number of paired node groups of the signal node and the voice prompt, represents the time sequence of the signal node in the i-th group, represents the time sequence of the voice prompt in the i-th group.

[0112] ​​The average response difference amplitude refers to the overall dispersion degree of the response difference in time sequence of all signal nodes and corresponding voice prompts, and reflects whether the response of each group of signal nodes and voice prompts in time is synchronized and whether the deviation is large. The smaller the value is, the closer the response of the signal and the voice prompt is, and the higher the synchronization is. The larger the value is, the more obvious the time sequence difference between the two is, and the lower the synchronization is. The index is used to measure the close degree of the action execution and the voice guidance, and is the core quantitative basis for evaluating the effectiveness of voice guidance;

[0113] The time sequence response difference between each pair of signal nodes and voice prompts is analyzed, the original collection time of each group of signal nodes and the trigger time of the corresponding voice prompts are obtained respectively, and it is set that five groups of paired node data are formed in the current training period. The original data are as follows:

[0114] The time of the first group of signal nodes is 3.21 seconds, and the corresponding voice prompt is 3.20 seconds;

[0115] The second group is 5.45 seconds and 5.41 seconds;

[0116] The third group is 7.62 seconds and 7.59 seconds;

[0117] The fourth group is 10.88 seconds and 10.86 seconds;

[0118] The fifth group is 13.74 seconds and 13.69 seconds;

[0119] After normalization processing is performed on all time data, the normalized signal time sequence is obtained as follows:

[0120] [0.276, 0.451, 0.63, 0.902, 1];

[0121] The normalized voice time sequence is:

[0122] [0.275, 0.445, 0.627, 0.9, 0.997];

[0123] The square difference of the response time difference of each group of pairs is calculated by calling the formula, and the following is obtained:

[0124] The first group is ;

[0125] The second group is ;

[0126] The third group is ;

[0127] The fourth group is ;

[0128] The fifth group is ;

[0129] The above results are summed up as , and the formula is:

[0130] ;

[0131] The results show that the current five groups of signal nodes and voice prompts have a high degree of synchronization response on a unified time scale, and the average response difference amplitude calculated is much lower than the upper limit of the timing synchronization reference interval of 0.01, belonging to the complete synchronization response interval. By quantifying the response deviation between the signal and the voice prompt, the result provides a criterion for subsequent pairing relationship screening and optimization. Therefore, the result not only reflects the timing coupling degree between the nodes, but also can be used as a pairing judgment condition in the guided matching construction module. Based on this value, further comparison is made between the difference and the mean deviation of each group, so as to deduce and form the signal-voice one-to-one matching relationship supported by the average response difference amplitude.

[0132] The guided matching construction module optimizes the pairing results of the signal nodes and the voice prompts at each stage based on the average response difference amplitude, screens the node combination with the best pairing effect, establishes a one-to-one correspondence between the signal nodes and the voice prompts, and obtains the voice guide timing characteristics.

[0133] The time interval values of each pairing group are extracted from the node alignment time interval group, and the average response difference values of all pairing pairs in each stage are calculated according to Action Three, for example, there are three pairings in the rising stage, and the time differences are 0.12 seconds, 0.08 seconds, and 0.15 seconds, respectively. The average value is 0.116 seconds, which is taken as the response difference amplitude reference of the rising stage. Then, it is compared with the set reference response difference threshold. If the average response difference is lower than the threshold value, for example, 0.2 seconds, the current pairing in this stage is valid, otherwise, the node needs to be reselected to establish a new pairing combination. All the selectable pairing nodes in each stage are reordered, and one or more signal nodes are replaced to find a new pairing combination, and the average response difference of the replaced combination is calculated again. If a new combination with a smaller response difference is found, for example, the time differences of the three groups after replacement are 0.09 seconds, 0.07 seconds, and 0.10 seconds, respectively, and the average value decreases to 0.086 seconds. This combination is taken as the optimal pairing. Then, the combination with the smallest response difference in each stage is selected, and the final one-to-one correspondence between the signal nodes and the voice prompts is formed in each combination. This relationship group is composed of stage number, signal node time, and voice prompt time, and is constructed into a complete voice guide timing characteristic data structure. The final output result identifies the voice guide trigger time point and the key response point of signal change in each stage to form a pairing, for example, in stage three, the pairing result is signal node 6.35 seconds and voice prompt 6.4 seconds, and the response time difference is 0.05 seconds. This structure supports the release of prompt instructions in the subsequent voice guide module according to the rhythm.

[0134] Referring to Figure 6 , the execution deviation evaluation module comprises:

[0135] The training period extraction submodule matches the instruction time points marked by the voice prompts with the start and end nodes of the user actions one by one based on the voice-guided timing features, extracts the corresponding action start and end times in each stage by judging the degree of node coincidence and the degree of timing closeness, and obtains a sequence of stage execution time periods;

[0136] Read the instruction time points corresponding to each stage voice prompt, and call the action segmentation signal nodes identified in the previous step. Compare the start voice prompt time of each stage with the time of the signal node at the start of the stage, calculate the time difference, and if the time difference is within the tolerance threshold, for example, not more than 0.25 seconds, mark it as "closely aligned", if the time difference is between 0.25 and 0.5 seconds, mark it as "delayed alignment", and if the time difference exceeds 0.5 seconds, mark it as "misaligned". Then, sequentially judge the time difference between each voice prompt time and the subsequent signal end node, and record the time interval value group of all voice prompt time points and signal node times. For example, in a one-repetition maintenance training, the maintenance stage start voice prompt is identified at 3.5 seconds, and the corresponding electromyographic signal node is at 3.7 seconds, with a time difference of 0.2 seconds. This pairing is marked as closely aligned. Subsequently, all voice prompts and signal nodes in the action stage are paired, and the start and end times of each stage are formed in turn, for example, the rising stage is from 2.2 seconds to 3.1 seconds, the maintenance stage is from 3.5 seconds to 4.4 seconds, and the descending stage is from 4.6 seconds to 5.3 seconds. The start and end periods of each stage aligned by voice signals and action nodes are uniformly arranged to form a complete sequence of stage execution time periods.

[0137] The stage difference calculation submodule compares the sequence of stage execution time periods with the preset action timing arrangement, focuses on the distribution and difference of the start and end nodes of each stage action in the time sequence, determines the timing error type of each stage by analyzing the node offset before and after, and obtains a set of time matching offsets.

[0138] The built-in training standard action template is called, which records the standard rhythm arrangement and preset start and end time points of each training action. For example, the template sets that the rising stage should be completed between 2.0 seconds and 3.0 seconds, the maintenance stage is 3.0 seconds to 4.0 seconds, and the falling stage is 4.0 seconds to 5.0 seconds. Then, the actual start and end time of each stage of the user is compared with the template time, and the start and end offset values are calculated in sequence. The offset direction is classified as advance or lag. For example, if the user's rising stage starts at 2.2 seconds and ends at 3.1 seconds, the start offset is +0.2 seconds, and the end offset is +0.1 seconds, mark this stage as "start and end lag type". For example, the maintenance stage starts at 3.5 seconds and ends at 4.4 seconds, with a lag of 0.5 seconds and 0.4 seconds respectively. According to the offset interval division standard, 0.0 to ±0.2 seconds is recognized as mild offset, ±0.2 to ±0.5 seconds is recognized as moderate offset, and more than ±0.5 seconds is recognized as severe offset. In the above example, the maintenance stage is determined to be moderate lag type. The start and end offset values, offset type and corresponding rhythm number of each stage are arranged as structured data, named time matching offset set.

[0139] The deviation degree summary submodule judges the relationship between the rhythm characteristics in the action structure parameters based on the time matching offset set, analyzes the advance, lag and rhythm out-of-sync phenomenon exhibited by the actual action in the difference stage, and integrates the offset performance of all stages to obtain the execution deviation degree index.

[0140] The start and end offset values of each stage and the rhythm characteristic number are extracted, the difference between the start and end rhythm numbers is compared with the standard rhythm interval, and it is judged whether there is rhythm continuity abnormality in the stage. If the start and end rhythm difference exceeds the standard rhythm length tolerance such as 0.3 seconds, it is marked as rhythm discontinuity segment. Then, the offset direction of each stage is counted. If there are two stages in succession whose start or end are lag type, mark this combination as "overall lag segment". Otherwise, it is "advance segment" or "rhythm inconsistency segment". For example, the user's actual performance is lagging by 0.2 seconds in the first stage, 0.1 seconds in advance in the second stage, and 0.6 seconds in lag in the third stage. It is judged that there is a problem of inter-stage rhythm fracture and end delay in this training cycle. Then, the offset type is valued and scored according to mild, moderate and severe. Mild is 1 point, moderate is 2 points, and severe is 3 points. Then, the weighted calculation is performed on the offset score of each stage. If the total score after weighting exceeds the preset upper limit value such as 6 points, it is marked as "significant deviation". Otherwise, it is "slight deviation" or "good rhythm". In the output execution deviation degree index, the offset score of each stage, rhythm misplacement distribution and training cycle comprehensive score are recorded for subsequent quality analysis and feedback control.

[0141] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. A mobile terminal based pelvic floor muscle training system, characterized in that, The system comprises: The signal cooperative acquisition module analyzes the pelvic floor muscle electrical signal and the heart rate signal based on the training action, judges the synchronous fluctuation of the two on the same time axis, compares the order of sudden change occurrence, determines the synchronous sudden change node and counts the number, and obtains the synchronous sudden change node feature; The multi-parameter linkage identification module filters the signal synchronization node based on the synchronous sudden change node feature, compares the acquisition time of the muscle electrical signal, the photoelectric heart rate and the three-axis displacement signal, analyzes the corresponding relationship, judges the node linkage performance, and obtains the three-parameter linkage feature group; The three-segment action decomposition module judges the rising, maintaining and falling stages of the displacement signal based on the three-parameter linkage feature group, compares the signal changes in the difference stage through continuity analysis, induces the difference performance in each stage, and obtains the action segmentation morphology feature; The timing voice guidance module compares the signal node and the voice prompt trigger time point based on the action segmentation morphology feature, analyzes the key node and the voice instruction response timing, filters the optimal node of response difference, establishes the corresponding relationship, and obtains the voice guidance timing feature.

2. The mobile terminal based pelvic floor muscle training system according to claim 1, characterized in that, The synchronous sudden change node feature includes the node occurrence time, the synchronous distribution type, and the associated signal number, the three-parameter linkage feature group includes the multi-signal coupling type, the event discrimination category, and the linkage amplitude identifier, the action segmentation morphology feature includes the action structure parameter, the segmentation morphology category, and the action cycle marker, and the voice guidance timing feature includes the trigger signal type, the segmentation voice node, and the guidance timing label.

3. The mobile terminal based pelvic floor muscle training system of claim 1, wherein, The signal cooperative acquisition module comprises: The timing fluctuation analysis submodule analyzes the collected pelvic floor muscle electrical signal and heart rate signal based on the training action, identifies the continuous change segment of each signal in the training process, calculates the change rate of each signal segment, judges the trend change of the signal on the time axis, optimizes the positioning of the signal inflection point, and obtains the timing change trend quantity; The sudden change relationship judgment submodule filters the key nodes with obvious signal jump based on the timing change trend quantity, judges the order of occurrence of the sudden change of various signals, analyzes the primary and secondary relationship of the sudden change occurrence, and calculates the time difference between the nodes, and obtains the sudden change priority information; The synchronous node identification submodule filters the nodes with time difference based on the sudden change priority information, calculates the number and distribution of the synchronous sudden change nodes, compares the synchronous change of each signal type, analyzes the coupling features of the signals in each time segment, and obtains the synchronous sudden change node feature.

4. The mobile terminal based pelvic floor muscle training system of claim 1, wherein, The multi-parameter linkage identification module comprises: The signal synchronization extraction submodule analyzes the collected muscle electrical signal, heart rate signal and displacement signal based on the synchronous sudden change node feature, judges the synchronous change trend of each signal on the time axis, filters the signal combination with synchronous change in the same time segment, and obtains the synchronous node set; The node matching discrimination submodule compares the change direction and fluctuation of each signal in the recording period based on the synchronous node set, judges the consistency of the difference signals on the same node, analyzes the timing and trend corresponding relationship, and obtains the node coupling distribution group; The linkage feature generation submodule judges the combination of changes of each signal based on the node coupling distribution group, analyzes the synchronization and linkage features of the signals under the key nodes, filters the change types with relevance, and obtains a three-parameter linkage feature group.

5. The mobile terminal based pelvic floor muscle training system of claim 1, wherein, The three-stage action decomposition module includes: The phase identification submodule analyzes the cooperative performance of each signal based on the three-parameter linkage feature group, judges the rising, stable, and falling trends of the displacement signal curve in the difference time period, optimizes the start and end nodes of the phase division, filters the effective interval, and obtains a phase boundary time sequence; The signal comparison submodule compares the displacement signal and the electromyographic signal trajectory of each phase based on the phase boundary time sequence, analyzes the fluctuation amplitude and change consistency of the two types of signal trajectories in each phase, calculates the fluctuation difference between the trajectories in each phase, and obtains a phase signal difference amplitude range through fluctuation difference collection; The form induction submodule judges the phase number with the most performance based on the phase signal difference amplitude range, analyzes the change process of the displacement signal and the electromyographic signal response in the phase, induces the difference features of the three phases, and obtains an action segmented form feature.

6. The mobile terminal based pelvic floor muscle training system of claim 1, wherein, The time sequence voice guidance module includes: The voice node alignment submodule compares each phase signal node and voice prompt in time sequence based on the action segmented form feature in combination with the corresponding time points of the voice prompt, filters the signal nodes and voice prompts with similar time to pair them, and obtains a node alignment time interval group; The response difference calculation submodule compares the time interval between each pair of signal nodes and voice prompts based on the node alignment time interval group, calculates the deviation of the signal and voice synchronous change, and obtains an average response difference amplitude; The guidance matching construction submodule optimizes the pairing results of the signal nodes and voice prompts in each phase based on the average response difference amplitude, filters the node combination with the optimal pairing effect, establishes the one-to-one correspondence between the signal nodes and voice prompts, and obtains a voice guidance time sequence feature.

7. The mobile terminal based pelvic floor muscle training system of claim 1, wherein, The system further includes: The execution deviation evaluation module calculates the difference between the voice guidance time sequence feature and the actual action start and end time period of the user, analyzes the execution time, weights the phase difference, judges the consistency with the target, and obtains an execution deviation index; The execution deviation index includes a time sequence deviation level, an action phase evaluation, and an execution consistency classification.

8. The mobile terminal based pelvic floor muscle training system according to claim 7, characterized in that, The execution deviation evaluation module includes: The training time period extraction submodule one-to-one matches the instruction time points marked by the voice prompt and the user action start and end nodes based on the voice guidance time sequence feature, extracts the corresponding action start and end time in each phase by judging the node coincidence degree and time sequence closeness degree, and obtains a phase execution time period sequence; The phase difference calculation submodule compares the phase execution time period sequence with the preset action time sequence arrangement, focuses on the distribution and difference of the start and end nodes of each phase action in time sequence, determines the time sequence error type of each phase by combing the node offset performance before and after, and obtains a time matching offset set; The deviation degree aggregation submodule determines the relationship between the rhythm characteristics in the action structure parameters based on the time matching offset set, analyzes the advance, lag and rhythm asynchronization phenomenon of the actual action in the difference stage, and integrates the offset performance of all stages to obtain the execution deviation degree index.

Images