A method and system for evaluating pelvic floor muscle rehabilitation training of stress urinary incontinence

Abstract

The present application relates to the technical field of rehabilitation evaluation, in particular to a stress urinary incontinence pelvic floor muscle rehabilitation training evaluation method and system, comprising the following steps: based on rehabilitation training data, uniform same session time identifier and upload sequence, in turn screening resting reference segment, dividing contraction attribution sequence, positioning fatigue boundary sequence, and combining abdominal pressure peak segment, pelvic floor muscle contraction coverage interval and urine leakage trigger relationship, obtaining rehabilitation state grade. The present application unifies training session time identifier and corrects multi-source record sequence, forms same session time sequence relationship of muscle electromyography start and stop, abdominal pressure trigger and urine leakage record, resting reference screening and contraction segment attribution determination can exclude resting drift and abdominal pressure compensation interference, fatigue boundary positioning and urine control time sequence checking can distinguish effective load, fatigue influence and urine leakage trigger relationship, thereby converting original training times into rehabilitation evaluation basis with quality attribution, solving original times mixed calculation and evaluation basis dispersion problems.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation assessment technology, and in particular to a method and system for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence. Background Technology

[0002] The field of rehabilitation assessment technology involves the collection, organization, analysis, and interpretation of information related to patients' functional status, training process, symptom changes, compliance records, and rehabilitation stages during medical rehabilitation. It primarily provides data support for rehabilitation program development, training process management, and follow-up assessment in rehabilitation departments, urology departments, gynecology departments, pelvic floor rehabilitation centers, and medical health management institutions. Traditional assessment methods for stress urinary incontinence pelvic floor muscle rehabilitation training assess training records, pelvic floor muscle contraction ability, urinary control-related performance, symptom scale information, and rehabilitation follow-up data during pelvic floor muscle rehabilitation training. These methods typically employ pelvic floor muscle function testing, urinary incontinence symptom questionnaires, training logs, medical data statistical analysis, and rehabilitation assessment models to comprehensively evaluate patients' pelvic floor muscle training-related information.

[0003] Current assessments of pelvic floor muscle rehabilitation training for stress urinary incontinence primarily rely on pelvic floor muscle function tests, symptom questionnaires, training logs, and follow-up records. Training records are summarized by session or date, and pelvic floor muscle electromyographic signals, changes in abdominal pressure, and leakage trigger information are often interpreted as independent items. The original number of training sessions is easily included in the rehabilitation dosage. It is difficult to correlate abdominal pressure compensation, changes in contraction quality after fatigue, and the timing of leakage occurrence with a single contraction event, resulting in insufficient differentiation between the actual training dosage, compensatory movements, and urinary incontinence mismatch status, and a fragmented basis for determining the rehabilitation level. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a method and system for assessing and training pelvic floor muscles for stress urinary incontinence. The technical solution is as follows: On the one hand, a method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence is provided, including the following steps: S1: Based on rehabilitation training data, analyze the time stamps and upload order of the same training session, compare the electromyographic start and stop points of the pelvic floor muscles with the abdominal pressure triggering time, determine the training stage to which the urinary leakage record belongs, and obtain the session time sequence chain. S2: Based on the time sequence of the aforementioned sessions, obtain the electromyographic signals of the pelvic floor muscles during the resting phase, screen out stable segments before abdominal pressure fluctuations, compare the connection state between the active contraction and the resting segment, and obtain the resting reference segment. S3: Based on the resting reference segment, analyze the relationship between the starting point and the ending point of electromyography, screen continuous contraction segments, compare the changes and maintenance states of pelvic floor muscle pressure, determine the attribution of abdominal pressure compensation, and obtain the contraction attribution sequence. S4: Based on the contraction attribution sequence, analyze the time sequence of effective contraction events, select the first-stage target events as tolerance references, compare the deviation positions of the subsequent rising rhythm and the maintenance state, and obtain the fatigue boundary sequence. S5: Based on the fatigue boundary sequence, check the peak time of abdominal pressure, compare the proximity relationship between the pelvic floor muscle contraction coverage area and the leakage trigger record, determine the order of load, fatigue and compensation, and obtain the recovery status level.

[0005] On the other hand, the session time sequence includes sampling channel number, event anchor point and frame number; the resting reference segment includes electromyographic baseline segment, signal drift amplitude and fallback stabilization segment; the contraction attribution sequence includes contraction event index, quality classification label and compensation classification label; the fatigue boundary sequence includes tolerance benchmark segment, turning point event location and load cutoff interval; and the rehabilitation status level includes training load level, urinary control time sequence level and compensation interference level.

[0006] On the other hand, the specific steps for obtaining the time-series chain of the sessions are as follows:

[0007] S101: Based on rehabilitation training data, the sampling channel number is attached according to the source of the acquisition device, the time identifier and the order of device upload are checked, and the mis-uploaded records are uploaded according to the frame sequence number and the status mark is continued. The mis-uploaded records are then assigned to the acquisition sequence to obtain the record sequence index. S102: Based on the recorded sequence index, mark the start and end points of the pelvic floor muscle electromyography (EMG) signals, align the start and end points with the abdominal pressure triggering time, and determine the triggering phase according to the sequential relationship between EMG activation, abdominal pressure change, and EMG fallback, thereby obtaining the rise and fall triggering relationship. S103: Based on the aforementioned rise and fall triggering relationship, analyze the training phase state interval into which the urine leakage recording time falls, compare the proximity relationship between the urine leakage recording time and the electromyography interval and the abdominal pressure triggering time, determine the arrangement position of the uploaded misaligned recording within the session, adjust the connection relationship between the event anchor point and the adjacent frame sequence number, and obtain the session time sequence chain.

[0008] On the other hand, the specific steps for obtaining the resting reference segment are as follows:

[0009] S201: Based on the time sequence chain of the training sessions, analyze the connection relationship between the state intervals and frame numbers during the training phase, compare the fluctuation trajectory of electromyography recording with the starting point of changes in abdominal pressure recording, determine the continuous frame segments in which no active contraction occurs and the abdominal pressure recording remains stable, classify the electromyography signals of the pelvic floor muscles into the resting phase, and obtain the resting electromyography segments. S202: Based on the resting electromyography segmentation, analyze the electromyography fluctuation trajectory within the recording interval before the abdominal pressure fluctuation, compare the electromyography amplitude change and the starting point of the abdominal pressure change in adjacent frames, identify the interval where the electromyography fluctuation is continuous and the abdominal pressure recording has not changed, and obtain stable candidate segments. S203: Based on the stable candidate segments, analyze the connection state between the end point of the active contraction segment and the beginning point of the resting segment, compare the electromyographic trajectory after the fall with the stable state of the candidate interval, determine whether the baseline segment belongs to the resting segment or the fall stable segment, adjust the position of the baseline segment, and obtain the resting reference segment.

[0010] On the other hand, the specific steps for obtaining the contracted attribution sequence are as follows:

[0011] S301: Based on the resting reference segment, analyze the starting point of the electromyographic fluctuation trajectory after it leaves the resting segment, compare the connection state of the electromyographic trajectory with the resting segment after it falls back to the end point, determine that the starting point and the end point correspond to the same event anchor point, and obtain the relationship between the rising and falling endpoints. S302: Based on the relationship between the rising and falling endpoints, analyze the sequence of the endpoints recorded by the pelvic floor muscle training acquisition device, compare the connection status of the rising segment, holding segment and falling segment in the time sequence of the session, screen out the recording intervals that are not separated by abdominal pressure fluctuations, and obtain the complete contraction interval. S303: Based on the complete contraction interval, analyze the pelvic floor muscle pressure change record and the state of the holding segment, compare the overlap between the pressure change direction and the electromyographic holding interval, determine the active contraction or compensatory segment corresponding to the increase in abdominal pressure, and obtain the contraction assignment sequence.

[0012] On the other hand, the specific steps for obtaining the fatigue boundary sequence are as follows:

[0013] S401: Based on the contraction attribution sequence, analyze the order of effective contraction events, compare the corresponding status of adjacent contraction event indices and quality classification labels, determine the intervention position of compensation classification labels and the position of ineffective contraction discontinuity, adjust the correlation between event anchor points and the temporal position of the event, and obtain event ranking mapping data. S402: Based on the event ranking mapping data, obtain the pre-training contraction events, compare the initial rhythm, the maintenance state and the fallback connection state, filter the quality classification labels and classify them into effective contraction events and the compensation classification labels do not mark the interference events, adjust the continuous assignment interval of the events, and obtain the tolerance reference interval. S403: Based on the tolerance reference segment, obtain the contraction performance in the post-training phase, compare the changes in the rising rhythm, holding state, and event interval relative to the previous reference segment, determine the adjacent event boundaries corresponding to the delayed rising start and the interruption of the holding segment, adjust the load cutoff position, and obtain the fatigue boundary sequence.

[0014] On the other hand, the steps for obtaining the rehabilitation status level are as follows: S501: Based on the fatigue boundary sequence, analyze the position of the tolerance benchmark segment and the load cutoff segment in the urinary continence training scenario, compare the connection between the abdominal pressure rise segment and the peak segment stay segment, determine the event anchor point corresponding to the abdominal pressure peak segment moment, adjust the correlation between the peak segment moment and the fatigue boundary, and obtain the peak segment boundary index. S502: Based on the peak segment boundary index, compare the relative positions of the pelvic floor muscle contraction initiation, maintenance coverage and fall-off end times to the leakage record, determine the stage of the leakage record's adjacent contraction coverage interval, adjust the association relationship between the contraction event index and the leakage record, and obtain the leakage proximity relationship. S503: Based on the aforementioned proximity relationship of urinary leakage, analyze the order of occurrence of effective load, fatigue transition and compensation attribution in the training sequence, determine the attribution of training load level, urinary control sequence level and compensation interference level, adjust the rehabilitation assessment judgment level, and obtain the rehabilitation status level.

[0015] On the other hand, the abdominal pressure triggering moment refers to the time point at which the abdominal pressure acquisition device records the change in abdominal pressure, and the training phase refers to the intervals of resting preparation, active contraction, relaxation and recovery, abdominal pressure induction, and training end state in the same training session, divided in chronological order.

[0016] On the other hand, the resting phase refers to the recording phase in which the patient does not actively contract the pelvic floor muscles and there is no movement-induced change in abdominal pressure, while the abdominal pressure fluctuation refers to the changes in abdominal pressure caused by coughing, standing up, bending over, holding one's breath, or exertion.

[0017] On the other hand, a pelvic floor muscle rehabilitation training assessment system for stress urinary incontinence is provided. This system is applied to the assessment method of pelvic floor muscle rehabilitation training for stress urinary incontinence, including:

[0018] The time sequence construction module is based on rehabilitation training data. It analyzes the time markers and upload order of the same training session, compares the electromyographic start and stop points of the pelvic floor muscles with the triggering time of abdominal pressure, determines the training stage to which the urinary leakage record belongs, and obtains the session time sequence chain.

[0019] The resting reference module acquires the electromyographic signals of the pelvic floor muscles during the resting phase based on the time sequence chain of the field, filters the stable segments before the abdominal pressure fluctuation, compares the connection state between the active contraction and the resting segment, and obtains the resting reference segment.

[0020] Based on the resting reference segment, the contraction attribution module analyzes the relationship between the starting point and the ending point of electromyography, filters continuous contraction segments, compares the changes and maintenance states of pelvic floor muscle pressure, determines the attribution of abdominal pressure compensation, and obtains the contraction attribution sequence.

[0021] Based on the contraction attribution sequence, the fatigue localization module analyzes the timing of effective contraction events, selects the first-stage target events as tolerance references, compares the deviation positions of the subsequent rising rhythm and the maintenance state, and obtains the fatigue boundary sequence.

[0022] The grading module, based on the fatigue boundary sequence, checks the peak time of abdominal pressure, compares the proximity of the pelvic floor muscle contraction coverage area with the leakage trigger record, determines the order of load, fatigue, and compensation, and obtains the recovery status level.

[0023] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0024] By standardizing the time signature of training sessions and correcting the sequence of multi-source recordings, a temporal relationship between the start and end of electromyography, abdominal pressure triggering, and urinary leakage records within the same session is established. Resting reference screening and contraction segment attribution determination are used to eliminate resting drift and abdominal pressure compensation interference. Fatigue boundary localization and urinary control sequence verification are used to distinguish the effective load, fatigue effects, and urinary leakage triggering relationship. In this way, the original training counts are transformed into rehabilitation assessment criteria with quality attribution, achieving the effect of distinguishing the actual training dose, compensatory movements, and urinary control mismatch status, and solving the problems of mixed counting of original counts and scattered assessment criteria in traditional assessments. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart of the main steps of the present invention; Figure 2 This is a flowchart of steps S1 of the present invention; Figure 3 This is a flowchart of steps S2 of the present invention; Figure 4 This is a flowchart of steps S3 of the present invention; Figure 5 This is a flowchart of step S4 of the present invention; Figure 6 This is a flowchart of steps S5 of the present invention; Figure 7 This is a system block diagram of the present invention. Detailed Implementation

[0027] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0028] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0029] This invention provides a method for assessing pelvic floor muscle rehabilitation training in cases of stress urinary incontinence, such as... Figure 1 As shown, it includes the following steps: S1: Based on rehabilitation training data, analyze the time stamps of mobile acquisition terminals and the order of device uploads within the same training session, compare the relationship between the start and end points of pelvic floor muscle electromyography signals and the timing of abdominal pressure triggering, determine the training stage corresponding to the time of urine leakage recording, adjust the time sequence of each record, and obtain the session time sequence chain. S2: Based on the time sequence of the sessions, obtain the electromyographic signals of the pelvic floor muscles during the resting phase, screen the continuous and stable segments before the abdominal pressure fluctuation, compare the connection state between the end point of the active contraction segment and the resting segment, adjust the position of the baseline segment, and obtain the resting reference segment. S3: Based on the resting reference segment, analyze the correspondence between the rising start point and the falling end point in the electromyographic fluctuation trajectory, screen the continuous contraction segments recorded by the pelvic floor muscle training acquisition device, compare the pelvic floor muscle pressure change records with the hold segment status, determine the category of the abdominal pressure compensation segment, and obtain the contraction classification sequence. S4: Based on the contraction attribution sequence, analyze the arrangement of effective contraction events in the training time sequence, select the first phase of qualified events as tolerance references, compare the deviation positions of the rising rhythm and the maintenance state in the second phase, determine the continuous event boundary of fatigue transition, and obtain the fatigue boundary sequence. S5: Based on the fatigue boundary sequence, the peak time of abdominal pressure is checked according to the urinary control training scenario. The proximity relationship between the pelvic floor muscle contraction coverage area and the leakage trigger record is compared. The corresponding order of effective load, fatigue transition and compensation is determined. The rehabilitation assessment level is adjusted to obtain the rehabilitation status level.

[0030] The session time sequence includes sampling channel number, event anchor point and frame number; the resting reference segment includes electromyographic baseline segment, signal drift amplitude and fall-off stabilization segment; the contraction attribution sequence includes contraction event index, quality classification label and compensation classification label; the fatigue boundary sequence includes tolerance benchmark segment, turning point event location and load cutoff interval; and the rehabilitation status level includes training load level, urinary continence time sequence level and compensation interference level.

[0031] In S1, the mobile acquisition terminal refers to the terminal used to receive pelvic floor muscle electromyography (EMG) records, abdominal pressure records, and leakage trigger records when the patient is undergoing pelvic floor muscle rehabilitation training for stress urinary incontinence, such as a handheld device; the device upload order refers to the order in which different acquisition devices transmit training records to the mobile acquisition terminal; the abdominal pressure trigger moment refers to the time point at which the abdominal pressure acquisition device records the beginning of changes in abdominal pressure; the leakage record moment refers to the time point of urinary control failure recorded in the leakage trigger recording device or the patient's operation record; and the training phase refers to the time intervals in the same training session divided chronologically into resting preparation, active contraction, relaxation recovery, abdominal pressure induction, and training end states.

[0032] In S2, the resting phase refers to the recording phase in which the patient does not actively contract the pelvic floor muscles and there are no movement-induced changes in abdominal pressure; abdominal pressure fluctuations refer to changes in abdominal pressure caused by coughing, standing up, bending over, holding one's breath, or exertion; continuous stable segments refer to recording segments in which the pelvic floor muscle electromyographic signals and abdominal pressure remain stable over a continuous period of time; active contraction segments refer to recording segments in which the electromyographic rise, hold, and fall occur when the patient actively contracts the pelvic floor muscles according to training requirements; the resting phase refers to the segment in the training recording in which the pelvic floor muscles are in a relaxed state; and the baseline segment refers to the reference segment selected from the resting phase or stable recording.

[0033] In S3, the electromyographic fluctuation trajectory refers to the rising, holding, and falling changes in the electromyographic signal of the pelvic floor muscles over training time; the training acquisition device refers to the electromyographic acquisition device or pelvic floor muscle pressure acquisition device used to record pelvic floor muscle rehabilitation training data, such as pelvic floor muscle electromyographic acquisition probes, pelvic floor muscle surface electromyographic electrodes, pelvic floor muscle pressure acquisition probes, or pelvic floor muscle rehabilitation training instruments; the continuous contraction segment refers to the complete segment formed by the electromyographic signal from the beginning of contraction to the end of relaxation; the pressure change record refers to the pressure change content recorded by the pressure acquisition device during the contraction of the pelvic floor muscles; the hold segment refers to the state maintained after the pelvic floor muscle contraction reaches its peak; and the abdominal pressure compensation segment refers to the recorded segment that accompanies the rise in abdominal pressure during the contraction of the pelvic floor muscles.

[0034] In S4, an effective contraction event refers to a pelvic floor muscle contraction event that is identified as eligible for rehabilitation assessment after contraction classification; an initial target event refers to a contraction event that occurs in the early stages of training and is classified as an effective contraction event; a subsequent rising rhythm refers to the rhythmic change in electromyographic signals as they transition from a resting state to a contraction state in the later stages of training; a deviation location refers to the location of an event where the performance of the subsequent contraction is inconsistent with the tolerance reference in the early stages; a fatigue transition point refers to the dividing point where the quality of pelvic floor muscle contraction begins to decline during training; and a continuous event boundary refers to the dividing point between adjacent contraction events used to locate the fatigue transition point.

[0035] In S5, the peak abdominal pressure moment refers to the time position when the change in abdominal pressure reaches its peak; the contraction coverage range refers to the time range covered by the contraction of the pelvic floor muscles from initiation, maintenance to fall; and the effective load refers to the amount of training included in the rehabilitation assessment after contraction classification and fatigue boundary treatment.

[0036] like Figure 2 As shown, the specific steps for obtaining the time series chain of a session are as follows:

[0037] S101: Based on rehabilitation training data, the sampling channel number is attached according to the source of the acquisition device, the time identifier and the order of device upload are checked, and the mis-uploaded records are uploaded according to the frame sequence number and the status mark is continued. The mis-uploaded records are then assigned to the acquisition sequence to obtain the record sequence index. Rehabilitation training data is first broken down into electromyography (EMG) acquisition records, abdominal pressure acquisition records, urinary leakage trigger records, and handheld terminal reception logs. Within the same training session, the device source, channel name, local generation time, terminal reception time, upload batch, and frame number are read for each record. The EMG probe is assigned as A1, the abdominal pressure probe as B1, and the urinary leakage button as C1. The frame number continuity of the same device is checked one by one. If the difference between adjacent frame numbers is 1, it is recorded as a continuous item; if the adjacent frame numbers are reversed but the local time is still increasing, it is recorded as an upload misalignment item; if the difference between adjacent frame numbers is greater than 1, it is recorded as a missing frame item. For example, if frame 126 of A1 was formed at 10:03:15.240, frame 127 at 260, and frame 128 at 280, but the terminal received frame 128 before frame 127, then frame 128 is registered as an early upload item, and frame 127 is registered as a delayed upload item. During sorting, the order of terminal reception is not considered; instead, the order is determined by device origin, local formation time, and the continuity of frame number, placing the frame after frame 126. If frame 78 of B1 was received by the terminal 820 milliseconds later than frame 214 of A1, but its local formation time was 160 milliseconds earlier, then frame 78 of B1 is placed before the corresponding adjacent EMG frame, forming a record sequence index.

[0038] S102: Based on the recording sequence index, the start and end points of the pelvic floor muscle electromyography (EMG) signals are marked, the start and end points are aligned with the abdominal pressure triggering time, and the triggering phase is determined according to the order of EMG initiation, abdominal pressure change and EMG fallback, thus obtaining the initiation and fall triggering relationship. Retrieve the amplitude sequence, frame number sequence, and formation time sequence of the electromyography (EMG) channel; then retrieve the pressure sequence and formation time sequence of the abdominal pressure channel. The resting EMG reference can be the median level of a continuous, stable segment. For example, if the EMG value is mostly between 6 and 8 microvolts within a 10-second resting segment, 7 microvolts is used as the reference level for this session. When three consecutive frames are all 5 microvolts or more above the resting reference level, and the amplitude increases frame by frame, the first frame that meets the condition is registered as the EMG initiation point. For example, if frame 181 is 8 microvolts, frame 182 is 14 microvolts, frame 183 is 23 microvolts, and frame 184 is 31 microvolts, then frame 182 is registered as the initiation point. The fallback endpoint is registered when two consecutive frames return to within 2 microvolts above or below the resting reference level. For example, if frame 246 is 18 microvolts, frame 247 is 11 microvolts, frame 248 is 8 microvolts, and frame 249 is 7 microvolts, then frame 248 is registered as the endpoint. The abdominal pressure trigger point is recorded when the abdominal pressure rises by 2 cmH2O relative to the resting level and is maintained for two consecutive frames. For example, if the resting abdominal pressure is 12 cmH2O, the 61st frame is 13 cmH2O, and the 62nd frame is 18 cmH2O, then the 62nd frame is recorded as the abdominal pressure trigger point. Subsequently, the order of the electromyographic (EMG) initiation point, the abdominal pressure trigger point, and the EMG endpoint is compared. If the EMG starts first, the abdominal pressure is in the middle, and the EMG falls later, it is recorded as the contraction-preceding phase, thus obtaining the rise-fall trigger relationship.

[0039] S103: Based on the rise and fall trigger relationship, analyze the training phase state interval into which the urine leakage recording time falls, compare the proximity relationship between the urine leakage recording time and the electromyography interval and the abdominal pressure trigger time, determine the arrangement position of the uploaded misaligned record in the session, adjust the connection relationship between the event anchor point and the adjacent frame sequence number, and obtain the session time sequence chain. The system acquires the start point, end point, abdominal pressure trigger point, abdominal pressure peak segment, and phase label for each contraction; then it reads the formation time, terminal reception time, and event frame number of the urine leakage record. Training sessions are divided into resting preparation, active contraction, relaxation recovery, abdominal pressure induction, and post-contraction intervals based on actual records. The urine leakage formation time is compared with the interval boundaries one by one, and the interval it falls into is registered as the corresponding stage. For example, resting preparation is from 10:00:00 to 10:01:30, active contraction is from 10:01:30 to 10:04:10, relaxation recovery is from 10:04:10 to 10:04:50, and abdominal pressure induction is from 10:04:50 to 10:05:20; if urine leakage occurs at 10:05:03, it is registered as the abdominal pressure induction stage. Next, the time of urinary leakage is compared with the nearest EMG interval. If the EMG start point is 10:04:56 and the end point is 10:05:06, then the urinary leakage falls within the EMG coverage area. If the abdominal pressure peak is 10:05:02.600 milliseconds and the urinary leakage is 10:05:03.000 milliseconds, the difference is 400 milliseconds. The proximity judgment value is registered according to the sampling interval and the manual button delay. When the EMG interval is 20 milliseconds, the abdominal pressure interval is 50 milliseconds, and the manual button delay is 500 milliseconds, the value within 500 milliseconds is registered as the peak proximity. Then, the uploaded misaligned records are placed back between adjacent events according to the local formation time and frame number, and the event anchor points are adjusted to connect with the adjacent frame numbers to obtain the scene time sequence chain.

[0040] like Figure 3 As shown, the specific steps for obtaining the resting reference fragment are as follows:

[0041] S201: Based on the time sequence chain of the training sessions, analyze the connection between the state intervals and frame numbers during the training phase, compare the fluctuation trajectory of electromyography (EMG) recordings with the starting point of changes in abdominal pressure recordings, determine the continuous frame segments where no active contraction occurs and the abdominal pressure recording remains stable, classify the pelvic floor muscle EMG signals into the resting phase, and obtain the resting EMG segments. Within the same training session, consecutive frame numbers from the electromyography (EMG) channel A1 and the abdominal pressure channel B1 are read. Each frame is then linked to the corresponding time intervals of resting preparation, active contraction, relaxation recovery, abdominal pressure induction, and training end. Frame number connections are checked against adjacent differences; a difference of 1 between adjacent frame numbers is recorded as continuous, a difference greater than 1 is recorded as a breakpoint, and if adjacent frame numbers are reversed, the local formation time is retrieved and reordered. Subsequently, the EMG amplitude and abdominal pressure values ​​are read frame by frame. The resting EMG reference can be determined from the recording of the first 10 seconds of the session. If there are 500 frames in the first 10 seconds at 50Hz sampling, with EMG concentrated between 6 and 8 microvolts and abdominal pressure concentrated between 11 and 13 cmH2O, then the stable range of EMG is recorded as 5 to 10 microvolts, and the stable range of abdominal pressure is recorded as 10 to 14 cmH2O. If, within a consecutive frame segment, the electromyography (EMG) does not exceed 12 microvolts for three consecutive frames, and the abdominal pressure does not exceed 15 cmH2O for two consecutive frames, and the frame segment does not fall between the active contraction anchor point and the abdominal pressure-induced anchor point, then it is registered as a frame segment with no active contraction and stable abdominal pressure. For example, frames 300 to 520 correspond to 10:02:00 to 10:02:04.400 milliseconds. The EMG values ​​are between 6.5 microvolts and 8.2 microvolts, and the abdominal pressure is between 11.8 cmH2O and 12.6 cmH2O. Since no contraction event anchor point is connected, the pelvic floor muscle EMG signals within this interval are classified as resting phase, resulting in a resting EMG segment.

[0042] S202: Based on resting electromyography segmentation, analyze the electromyography fluctuation trajectory within the recording interval before abdominal pressure fluctuation, compare the electromyography amplitude changes in adjacent frames with the starting point of abdominal pressure changes, identify the interval where the electromyography fluctuations are continuous and the abdominal pressure recording has not changed, and obtain stable candidate segments. Each frame of the recording before each point of abdominal pressure change was checked. First, the range where the abdominal pressure was still within a stable range was identified, and then the amplitude jumps of the electromyography (EMG) in adjacent frames were checked. The stability judgment value was set with reference to the resting EMG fluctuations in the same session. If the EMG in the previous resting segment was a minimum of 6 microvolts and a maximum of 8 microvolts, and the difference between adjacent frames was usually between 0.1 microvolts and 0.6 microvolts, then the allowable change in adjacent frames was recorded as 1.5 microvolts. If the resting abdominal pressure range was 11 cmH2O to 13 cmH2O, then the starting point of the movement change was recorded as two consecutive frames reaching 15 cmH2O or above. During the examination, if the electromyography (EMG) values ​​from frames 610 to 690 are successively between 6.8 μV and 8.1 μV, with adjacent frame variations not exceeding 1.5 μV, and the abdominal pressure is between 11.7 cmH2O and 12.8 cmH2O, then this interval is registered as a candidate interval. If the abdominal pressure is 13.4 cmH2O in frame 691, 15.6 cmH2O in frame 692, and 18.2 cmH2O in frame 693, then frame 692 is registered as the starting point of the abdominal pressure change, and frames 610 to 691 are recorded as the interval before the abdominal pressure fluctuation. If an EMG segment shows 7.2 μV in frame 640, 12.5 μV in frame 641, and 18.4 μV in frame 642, even if the abdominal pressure remains unchanged, frames after 641 are discarded, and frames 610 to 640 are retained. After checking the amplitude of adjacent frames, the starting point of the abdominal pressure change, and the length of consecutive frames, stable candidate segments are obtained.

[0043] S203: Based on stable candidate segments, analyze the connection between the end point of the active contraction segment and the beginning point of the resting segment, compare the electromyographic trajectory after the fall with the stable state of the candidate interval, determine whether the baseline segment belongs to the resting segment or the stable fall segment, adjust the position of the baseline segment, and obtain the resting reference segment. Acquire candidate initiation points, candidate endpoints, candidate segment electromyographic (EMG) ranges, adjacent active contraction fallback endpoints, and subsequent resting segment initiation points. Verify the fallback process frame-by-frame after active contraction, first checking if the EMG returns to the stable range of the candidate segment, then checking for any jumps or drifts between the fallback endpoint and the resting segment initiation point. Fallback stability is determined based on the fluctuation settings of the candidate segment itself. If the candidate segment EMG is between 6.4 μV and 8.0 μV, and the maximum change between adjacent frames is 0.8 μV, then if the EMG falls between 5.5 μV and 9.0 μV for one consecutive second after fallback, and the change between adjacent frames does not exceed 1.5 μV, it is registered as fallback stable. For example, if the peak value of an active contraction is 32 microvolts, it drops to 14 microvolts in frame 820, 10 microvolts in frame 821, 8.4 microvolts in frame 822, 7.6 microvolts in frame 823, and 7.2 microvolts in frame 824, and remains between 6.8 and 8.1 microvolts in the next 50 frames, then the segment after frame 822 can be registered as the stable decline segment. If the candidate segment from frame 760 to frame 810 is located before contraction, with electromyography values ​​of 6.7 to 7.9 microvolts and abdominal pressure around 12 cmH2O, then this segment is registered as the resting baseline. If the difference between frames 822 to 872 after contraction and the candidate segment before contraction does not exceed 1.2 microvolts, and the intra-abdominal pressure does not exceed 14 cmH2O, then this segment after contraction can be registered together with the resting segment; if it remains between 10.5 and 12 microvolts after contraction, it is registered as an unstable contraction and not included in the resting baseline. Subsequently, the baseline segment is reattached according to the location of the candidate segment to obtain the resting reference segment.

[0044] like Figure 4 As shown, the specific steps for obtaining the contracted sequence are as follows:

[0045] S301: Based on the resting reference segment, analyze the rising starting point of the electromyographic fluctuation trajectory after it leaves the resting segment, compare the connection state of the electromyographic trajectory with the resting segment after the falling end point, determine that the starting point and the ending point correspond to the same event anchor point, and obtain the relationship between the rising and falling endpoints. The electromyography (EMG) amplitude, formation time, difference between adjacent frames, and adjacent event anchor points are acquired for each frame. First, the resting range of this session is recorded using continuous 5-second intervals within the resting segment. For example, if the EMG amplitude of 250 frames in the resting segment falls between 6.2 μV and 8.4 μV, the median level is recorded as 7.1 μV. The upper edge of the resting range is recorded as 10.4 μV, increasing by 2 μV from 8.4 μV. When three consecutive frames reach 10.4 μV or higher and increase frame by frame, this is recorded as the starting point of the rise from the resting range. If frame 436 is 8.1 μV, frame 437 is 11.2 μV, frame 438 is 16.8 μV, and frame 439 is 25.6 μV, then frame 437 is recorded as the starting point. Then, the fall trajectory is read backwards from the peak, with the fall endpoint recorded as a continuous 5-frame interval between 5.1 μV and 9.1 μV. The voltage registration is based on a range of 2 microvolts above and below the resting median level. If frame 502 is 14.3 microvolts, frame 503 is 10.1 microvolts, frame 504 is 8.7 microvolts, frame 505 is 7.8 microvolts, frame 506 is 7.3 microvolts, frame 507 is 7.5 microvolts, and frame 508 is 7.1 microvolts, then frame 504 is registered as the endpoint. Next, check whether there are other contraction anchor points or leakage anchor points inserted between the start and end points. If the frame numbers between the start and end points are continuous, the interval between adjacent frames is maintained at 20 milliseconds, the electromyography peak appears between the two, and after falling back, it connects with the resting segment. Then, the start and end points are attached to the same event anchor point. If there is a frame missing of more than 300 milliseconds or another start point in the middle, it is split into different events to obtain the relationship between the start and end points.

[0046] S302: Based on the relationship between the rising and falling endpoints, analyze the sequence of the endpoints in the recordings of the pelvic floor muscle training acquisition device, compare the connection status of the rising segment, holding segment and falling segment in the time sequence of the session, screen out the recording intervals that are not separated by abdominal pressure fluctuations, and obtain the complete contraction interval. Acquire the start frame, peak frame, end frame, EMG acquisition channel, pressure acquisition channel, and abdominal pressure acquisition channel. Verify the position of each endpoint in the time sequence of the sessions. The start frame must precede the peak frame, and the peak frame must precede the end frame. If frame 437 is the start frame, frame 463 is the peak frame, and frame 504 is the end frame, then the endpoint order is registered as a valid order. If the end frame is earlier than the peak frame, the event is registered as an endpoint mismatch. Then, register the period from the start frame to the peak frame as the rising segment, the continuous stable portion near the peak frame as the holding segment, and the period from the end of the holding segment to the end frame as the falling segment. The amplitude range of the holding segment is registered as 70% to 10% of the peak value. For example, if the peak value is 36 microvolts, 25 to 40 microvolts are registered as the holding range. If frames 462 to 488 all fall within the holding range... If the voltage is between 26 and 38 microvolts and the duration is 540 milliseconds, it is recorded as a short hold. If the duration is between 1000 and 4000 milliseconds, it is recorded as a regular hold. Then, the changes in the abdominal pressure channel within this event are read. If the resting abdominal pressure is 12 cmH2O, and two consecutive frames reach 15 cmH2O or more, it is recorded as the start of the abdominal pressure fluctuation. If the start of the abdominal pressure fluctuation falls within 500 milliseconds before the rising segment and the fluctuation continues to pass through the hold segment, then this interval is recorded as being separated by abdominal pressure. If the abdominal pressure is always between 11 and 14 cmH2O, and the rising, hold, and falling segments of the electromyography are connected consecutively according to the frame number, then this recording interval is retained. If there is a leakage anchor point between the start and end points but the abdominal pressure does not fluctuate, it is still recorded as the same contraction interval, thus obtaining a complete contraction interval.

[0047] S303: Based on the complete contraction interval, analyze the pelvic floor muscle pressure change records and the state of the maintenance segment, compare the overlap between the pressure change direction and the electromyographic maintenance interval, determine whether the increase in abdominal pressure corresponds to active contraction or compensatory segment, and obtain the contraction assignment sequence. Acquire the electromyographic amplitude sequence, pelvic floor muscle pressure sequence, abdominal pressure sequence, start and end frames of the hold interval, and event index. Read the overlap between the pressure change direction and the electromyographic hold interval for each sequence. The resting pressure reference is recorded 2 seconds before contraction. For example, if the pelvic floor muscle pressure falls between 17 and 20 cmH2O within 100 frames before contraction, the median level is recorded as 18.5 cmH2O. The active pressure change determination value is recorded as 23.5 cmH2O, which is 5 cmH2O higher than the median resting level. If, within 300 milliseconds after the electromyographic initiation, the pelvic floor muscle pressure rises from 18.5 cmH2O to 27 cmH2O, the direction of pressure increase is consistent with the direction of electromyographic increase, and the interval with pressure above 23.5 cmH2O overlaps with the electromyographic hold interval for more than 1 second, then this event is recorded as an active contraction. If, after the electromyography (EMG) enters the hold phase, the pelvic floor muscle pressure does not exceed 23.5 cmH2O, but the abdominal pressure rises from 12 cmH2O to 21 cmH2O, and the increase in abdominal pressure occurs 200 milliseconds earlier than the EMG initiation point, then the event is registered as abdominal pressure compensation. If the pelvic floor muscle pressure rises to 25 cmH2O but the abdominal pressure simultaneously rises to 18 cmH2O, and the high value of the abdominal pressure overlaps with the EMG hold phase for more than half the hold duration, then it is registered as active contraction with compensatory interference. The hold phase state is categorized by duration: less than 800 milliseconds is registered as insufficient hold, 800 to 3000 milliseconds is registered as adequate hold, and more than 3000 milliseconds is registered as prolonged hold. Finally, the index number, start and end frames, active contraction category, hold state, and compensation category of each event are written in sequence to obtain the contraction attribution sequence.

[0048] like Figure 5 As shown, the specific steps for obtaining the fatigue boundary sequence are as follows:

[0049] S401: Based on the shrinkage attribution sequence, analyze the order of effective shrinkage events, compare the corresponding status of adjacent shrinkage event indices and quality classification labels, determine the intervention position of compensation classification labels and the position of ineffective shrinkage discontinuity, adjust the correlation between event anchor points and the temporal position of the event, and obtain event ranking mapping data; Each event is read from index number in ascending order. First, the continuity of adjacent indices is checked, then the interval between adjacent start times is checked. The interval is obtained by subtracting the start time of the previous event from the start time of the subsequent event. For example, if the first start time is 10:02:10 and the second start time is 10:02:16, the interval is recorded as 6 seconds. If the training instruction beat is contraction every 6 seconds, and the allowable deviation is recorded as 1.5 seconds (25% of the beat), then 4.5 seconds to 7.5 seconds are classified as rhythmic continuity, less than 4.5 seconds as premature triggering, and more than 7.5 seconds as delayed triggering. Then, the quality classification label is read. Active contraction and maintaining the standard are recorded as valid, while insufficient maintenance, endpoint mismatch, and no pressure response are recorded as invalid. Compensation classification labels are checked separately. If an event is registered as active contraction with compensatory interference, the event is not included in the tolerance continuum; only the interference marker is retained at its original position. If the first to fourth events are all effective, the fifth event is abdominal pressure compensation, and the sixth event is effective, then the fifth event is registered as the location of the compensatory intervention, and the period between the fourth and sixth events is registered as a discontinuous connection. If the eighth and ninth events are both invalid, and there are no effective events between them, then the period from the start of the eighth event to the end of the ninth event is registered as an invalid discontinuous interval. Subsequently, the event anchor points are reattached to the frame number position in the session time sequence chain. If the time difference between the anchor point and the adjacent electromyographic frame is less than 40 milliseconds, it is attached to that frame; if it exceeds 40 milliseconds, it is attached between the two preceding and following frames, thus obtaining the event ranking mapping data.

[0050] S402: Based on the event ranking mapping data, obtain the contraction events in the early stage of training, compare the initial rhythm, the maintenance state and the fallback connection state, filter the quality classification labels and classify them into effective contraction and the compensation classification labels are not marked as interference events, adjust the continuous assignment interval of events, and obtain the tolerance reference interval. Acquire the target number of contractions, the ranking of each event, the initial rhythm, the duration of hold, the resting transition after the fall, and the compensation marker for each session. For the initial training phase, the top 30% of the target number of contractions are used; if the target is 20 contractions, the top 30% are equivalent to 6 contractions; if the target is less than 10 contractions, the top 3 are used. Compare the initial rhythm of each event in the initial phase. If the intervals between the starting points of the 1st to 6th contractions are 6.0 seconds, 6.2 seconds, 5.8 seconds, 6.1 seconds, and 6.0 seconds respectively, all falling within the range of 4.5 seconds to 7.5 seconds, then the rhythm is considered acceptable. Next, compare the hold status. The acceptable hold range follows the 800 milliseconds to 3000 milliseconds range from the contraction classification. If the hold durations of the 1st to 6th contractions are 2100 milliseconds, 2250 milliseconds, 2180 milliseconds, 2300 milliseconds, 2050 milliseconds, and 2200 milliseconds respectively, then all are considered acceptable hold durations. The fallback transition is checked by checking if the electromyography (EMG) returns to the resting reference level within 2 microvolts within 1 second after the endpoint. If the resting median is 7.1 microvolts, the fallback transition range is 5.1 to 9.1 microvolts. If the EMG values ​​are consistently between 6.8 and 8.4 microvolts within 50 frames after the third endpoint, it is considered a stable transition. Events labeled as invalid (quality-based) and events labeled as interference (compensation-based) are then removed. If the first four transitions are valid and interference-free, the fifth transition is a compensatory interference, and the sixth transition is valid, then the first four transitions are registered as a continuous tolerance segment, and the sixth transition is considered a temporary segment. If there are fewer than three consecutive valid events in a current segment, the next valid event without interference is added until three consecutive events are obtained, thus determining the tolerance reference segment.

[0051] S403: Based on the tolerance reference segment, obtain the contraction performance in the later stage of training, compare the changes in the rising rhythm, holding state and event interval relative to the previous reference segment, determine the adjacent event boundaries corresponding to the delayed rising start and the interruption of the holding segment, adjust the load cutoff position, and obtain the fatigue boundary sequence. The system retrieves the rise time, hold time, event interval, and fallback range from the preceding reference. Taking the first four rises in the preceding period as an example, the rise times are 500ms, 540ms, 520ms, and 560ms respectively. The preceding reference is registered as around 520ms. The hold time is between 2050ms and 2300ms, and the event interval is between 5.8 seconds and 6.2 seconds. The training process starts reading from the first event after the tolerance reference segment, comparing the rise time, hold time, and event interval item by item. The rise delay is determined based on the preceding rise time setting. If the preceding reference is 520ms, anything exceeding half the increment of the preceding period (780ms) is registered as delayed. If the 12th rise time is 820ms, it is marked as a delayed rise. Interruptions are recorded when the duration is less than 70% of the preceding segment's lower limit or less than 800 milliseconds. When the preceding segment's lower limit is 2050 milliseconds, 70% is 1435 milliseconds. If the 13th hold lasts only 1200 milliseconds, it is marked as a shortened hold. If the electromyographic value drops from 30 microvolts to 18 microvolts and then rises again during the hold, it is recorded as an intra-segment interruption. Event intervals are recorded as 1.5 times the preceding segment's upper limit of 6.2 seconds; intervals exceeding 9.3 seconds are marked as prolonged intervals. If the 12th event shows a delayed rise, and the 13th event shows both a shortened hold and a prolonged interval, the interval between the 11th and 12th events is recorded as an adjacent boundary of the fatigue transition. If only a single event is abnormal, the next event is checked. If two consecutive subsequent events are abnormal, the load position is truncated to obtain the fatigue boundary sequence.

[0052] like Figure 6 As shown, the specific steps for obtaining the recovery status level are as follows:

[0053] S501: Based on the fatigue boundary sequence, analyze the position of the tolerance benchmark segment and the load cutoff interval in the urinary continence training scenario, compare the connection between the abdominal pressure rise segment and the peak segment stay segment, determine the event anchor point corresponding to the abdominal pressure peak segment moment, adjust the correlation between the peak segment moment and the fatigue boundary, and obtain the peak segment boundary index. Acquire the markers for the urinary control training scenario: cough trigger is recorded as P1, standing trigger as P2, and breath-holding trigger as P3. Then, read the starting frame, peak start frame, peak end frame, and adjacent electromyographic event indexes for the rise in abdominal pressure. Abdominal pressure stability is referenced by registering the data 2 seconds before triggering, for example, from 11 cmH2O to 13 cmH2O. Two consecutive frames reaching 15 cmH2O are registered as the start of the rise, and three consecutive frames maintaining above 22 cmH2O are registered as the peak end. If the starting point of the 12th event is 10:06:20.120 ms, the holding period is from 10:06:20.760 ms to 10:06:22.50 ms, and the peak abdominal pressure is from 10:06:21.100 ms to 10:06:21.90 ms, then the peak segment is included in the 12th holding period. If the fatigue boundary is registered between the 11th and 12th events, the 12th peak segment is classified as a post-boundary event. If the peak segment's start point is less than 300 milliseconds from the 11th end point and more than 900 milliseconds from the 12th start point, it is classified as a post-11th peak segment. Peak segments and event anchor points are registered as adjacent if the distance is less than 500 milliseconds, as delayed adjacent if the distance is between 500 and 1500 milliseconds, and as non-adjacent if the distance is greater than 1500 milliseconds. The index is then reconnected according to the event to which the peak segment belongs, the position before and after the boundary, and the load cutoff position to obtain the peak segment boundary index.

[0054] S502: Based on the peak segment boundary index, compare the relative positions of the pelvic floor muscle contraction initiation, maintenance coverage and fall-off end times to the leakage record, determine the stage of the leakage record's adjacent contraction coverage interval, adjust the association between the contraction event index and the leakage record, and obtain the leakage proximity relationship. The device formation time, manual trigger time, and terminal reception time are used to record urine leakage. The device formation time is preferred, and the manual recording delay is recorded in 500 milliseconds. When comparing, first check if the urine leakage time is earlier than the start of electromyography (EMG). If it is more than 500 milliseconds earlier than the start, it is recorded as pre-contraction urine leakage; if it falls between the start and the beginning of the hold phase, it is recorded as rising phase urine leakage; if it falls within the start and end time of the hold phase, it is recorded as hold phase urine leakage; if it falls between the end of the hold phase and the end of the fall phase, it is recorded as fall phase urine leakage; if it is later than the end of the fall phase but not more than 1500 milliseconds later, it is recorded as adjacent urine leakage after fall. For example, if the 12th EMG initiation occurs at 10:06:20:120, the hold period is from 10:06:20:760 to 10:06:22:500, and the decline ends at 10:06:23:100, and the urine leakage is recorded at 10:06:21:600, then it is classified as urine leakage during the hold period. If the urine leakage is recorded at 10:06:23:800, then it differs from the end of the decline by 700 milliseconds, and is classified as urine leakage adjacent to the end of the decline. If two contraction events are adjacent to the same urine leakage time, the event with the smaller time difference is selected; if the differences are the same, the event belonging to the peak of abdominal pressure is selected, thus obtaining the urine leakage proximity relationship.

[0055] S503: Based on the proximity relationship of urinary leakage, analyze the order of occurrence of effective load, fatigue transition and compensation in the training sequence, determine the training load level, urinary control sequence level and compensation interference level, adjust the rehabilitation assessment judgment level and obtain the rehabilitation status level. Effective load is registered based on the effective contraction events before load cutoff. With a target training duration of 20 repetitions, 16 to 20 effective events before cutoff are registered as Level 1 load, 10 to 15 as Level 2 load, and less than 10 as Level 3 load. Urinary control timing is registered based on the relationship between leakage and contraction coverage. No leakage and the peak within the contraction maintenance phase are registered as Level 1 timing; leakage only occurs within 1500 milliseconds after the fallback and is registered as Level 2 timing; leakage occurs before the rise or within the maintenance phase and is registered as Level 3 timing. Compensatory interference is registered based on the proportion of compensatory events to effective events. No more than 2 compensations in the 20-repetition target are registered as Level 1 interference, 3 to 6 as Level 2 interference, and more than 6 as Level 3 interference. If compensatory events are concentrated after the fatigue boundary, they are marked as late-stage compensation; if compensatory events occur before the fatigue boundary and are adjacent to the leakage event, they are marked as early-stage compensation. For example, if there are 14 valid events before the cutoff in a certain session, it is classified as Level 2 of the load. After the 12th event, a fatigue transition occurs, and urinary leakage occurs during the maintenance period of the 13th event, which is classified as Level 3 of the timing. There are 5 compensatory events, 3 of which are after the transition, which is classified as Level 2 of the interference. The judgment level is registered in sequence according to the timing level, interference level, and load level to obtain the recovery status level.

[0056] like Figure 7 As shown, a pelvic floor muscle rehabilitation training and assessment system for stress urinary incontinence includes:

[0057] The time sequence construction module is based on rehabilitation training data. It analyzes the time markers and upload order of the same training session, compares the electromyographic start and stop points of the pelvic floor muscles with the triggering time of abdominal pressure, determines the training stage to which the urinary leakage record belongs, and obtains the session time sequence chain.

[0058] The resting reference module is based on the time sequence of the sessions to obtain the electromyographic signals of the pelvic floor muscles during the resting phase, screen the stable segments before the abdominal pressure fluctuations, and compare the connection state between the active contraction return endpoint and the resting segment to obtain the resting reference segment.

[0059] The contraction attribution module is based on the resting reference segment. It analyzes the relationship between the starting point and the ending point of electromyography, selects continuous contraction segments, compares the changes and maintenance states of pelvic floor muscle pressure, determines the attribution of abdominal pressure compensation, and obtains the contraction attribution sequence.

[0060] The fatigue localization module analyzes the timing of effective contraction events based on the contraction attribution sequence, selects the first-stage target events as tolerance references, compares the deviation of the subsequent rising rhythm from the maintenance state, and obtains the fatigue boundary sequence.

[0061] The grading module is based on the fatigue boundary sequence, checks the peak time of abdominal pressure, compares the proximity of the pelvic floor muscle contraction coverage area with the leakage trigger record, determines the order of load, fatigue and compensation, and obtains the rehabilitation status level.

[0062] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for assessing pelvic floor muscle rehabilitation training in patients with stress urinary incontinence, characterized in that, The method includes: S1: Based on rehabilitation training data, analyze the time stamps and upload order of the same training session, compare the electromyographic start and stop points of the pelvic floor muscles with the abdominal pressure triggering time, determine the training stage to which the urinary leakage record belongs, and obtain the session time sequence chain. S2: Based on the time sequence of the aforementioned sessions, obtain the electromyographic signals of the pelvic floor muscles during the resting phase, screen out stable segments before abdominal pressure fluctuations, compare the connection state between the active contraction and the resting segment, and obtain the resting reference segment. S3: Based on the resting reference segment, analyze the relationship between the starting point and the ending point of electromyography, screen continuous contraction segments, compare the changes and maintenance states of pelvic floor muscle pressure, determine the attribution of abdominal pressure compensation, and obtain the contraction attribution sequence. S4: Based on the contraction attribution sequence, analyze the time sequence of effective contraction events, select the first-stage target events as tolerance references, compare the deviation positions of the subsequent rising rhythm and the maintenance state, and obtain the fatigue boundary sequence. S5: Based on the fatigue boundary sequence, check the peak time of abdominal pressure, compare the proximity relationship between the pelvic floor muscle contraction coverage area and the leakage trigger record, determine the order of load, fatigue and compensation, and obtain the recovery status level.

2. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The session time sequence includes sampling channel number, event anchor point and frame number; the resting reference segment includes electromyographic baseline segment, signal drift amplitude and fallback stabilization segment; the contraction attribution sequence includes contraction event index, quality classification label and compensation classification label; the fatigue boundary sequence includes tolerance benchmark segment, turning point event location and load cutoff interval; and the rehabilitation status level includes training load level, urinary control time sequence level and compensation interference level.

3. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The specific steps for obtaining the time-series chain of the sessions are as follows: S101: Based on rehabilitation training data, the sampling channel number is attached according to the source of the acquisition device, the time identifier and the order of device upload are checked, and the mis-uploaded records are uploaded according to the frame sequence number and the status mark is continued. The mis-uploaded records are then assigned to the acquisition sequence to obtain the record sequence index. S102: Based on the recorded sequence index, mark the start and end points of the pelvic floor muscle electromyography (EMG) signals, align the start and end points with the abdominal pressure triggering time, and determine the triggering phase according to the sequential relationship between EMG activation, abdominal pressure change, and EMG fallback, thereby obtaining the rise and fall triggering relationship. S103: Based on the aforementioned rise and fall triggering relationship, analyze the training phase state interval into which the urine leakage recording time falls, compare the proximity relationship between the urine leakage recording time and the electromyography interval and the abdominal pressure triggering time, determine the arrangement position of the uploaded misaligned recording within the session, adjust the connection relationship between the event anchor point and the adjacent frame sequence number, and obtain the session time sequence chain.

4. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The specific steps for obtaining the resting reference segment are as follows: S201: Based on the time sequence chain of the training sessions, analyze the connection relationship between the state intervals and frame numbers during the training phase, compare the fluctuation trajectory of electromyography recording with the starting point of changes in abdominal pressure recording, determine the continuous frame segments in which no active contraction occurs and the abdominal pressure recording remains stable, classify the electromyography signals of the pelvic floor muscles into the resting phase, and obtain the resting electromyography segments. S202: Based on the resting electromyography segmentation, analyze the electromyography fluctuation trajectory within the recording interval before the abdominal pressure fluctuation, compare the electromyography amplitude change and the starting point of the abdominal pressure change in adjacent frames, identify the interval where the electromyography fluctuation is continuous and the abdominal pressure recording has not changed, and obtain stable candidate segments. S203: Based on the stable candidate segments, analyze the connection state between the end point of the active contraction segment and the beginning point of the resting segment, compare the electromyographic trajectory after the fall with the stable state of the candidate interval, determine whether the baseline segment belongs to the resting segment or the fall stable segment, adjust the position of the baseline segment, and obtain the resting reference segment.

5. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The specific steps for obtaining the shrinkage attribution sequence are as follows: S301: Based on the resting reference segment, analyze the starting point of the electromyographic fluctuation trajectory after it leaves the resting segment, compare the connection state of the electromyographic trajectory with the resting segment after it falls back to the end point, determine that the starting point and the end point correspond to the same event anchor point, and obtain the relationship between the rising and falling endpoints. S302: Based on the relationship between the rising and falling endpoints, analyze the sequence of the endpoints recorded by the pelvic floor muscle training acquisition device, compare the connection status of the rising segment, holding segment and falling segment in the time sequence of the session, screen out the recording intervals that are not separated by abdominal pressure fluctuations, and obtain the complete contraction interval. S303: Based on the complete contraction interval, analyze the pelvic floor muscle pressure change record and the state of the holding segment, compare the overlap between the pressure change direction and the electromyographic holding interval, determine the active contraction or compensatory segment corresponding to the increase in abdominal pressure, and obtain the contraction assignment sequence.

6. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The specific steps for obtaining the fatigue boundary sequence are as follows: S401: Based on the contraction attribution sequence, analyze the order of effective contraction events, compare the corresponding status of adjacent contraction event indices and quality classification labels, determine the intervention position of compensation classification labels and the position of ineffective contraction discontinuity, adjust the correlation between event anchor points and the temporal position of the event, and obtain event ranking mapping data. S402: Based on the event ranking mapping data, obtain the pre-training contraction events, compare the initial rhythm, the maintenance state and the fallback connection state, filter the quality classification labels and classify them into effective contraction events and the compensation classification labels do not mark the interference events, adjust the continuous assignment interval of the events, and obtain the tolerance reference interval. S403: Based on the tolerance reference segment, obtain the contraction performance in the post-training phase, compare the changes in the rising rhythm, holding state, and event interval relative to the previous reference segment, determine the adjacent event boundaries corresponding to the delayed rising start and the interruption of the holding segment, adjust the load cutoff position, and obtain the fatigue boundary sequence.

7. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The specific steps for obtaining the rehabilitation status level are as follows: S501: Based on the fatigue boundary sequence, analyze the position of the tolerance benchmark segment and the load cutoff segment in the urinary continence training scenario, compare the connection between the abdominal pressure rise segment and the peak segment stay segment, determine the event anchor point corresponding to the abdominal pressure peak segment moment, adjust the correlation between the peak segment moment and the fatigue boundary, and obtain the peak segment boundary index. S502: Based on the peak segment boundary index, compare the relative positions of the pelvic floor muscle contraction initiation, maintenance coverage and fall-off end times to the leakage record, determine the stage of the leakage record's adjacent contraction coverage interval, adjust the association relationship between the contraction event index and the leakage record, and obtain the leakage proximity relationship. S503: Based on the aforementioned proximity relationship of urinary leakage, analyze the order of occurrence of effective load, fatigue transition and compensation attribution in the training sequence, determine the attribution of training load level, urinary control sequence level and compensation interference level, adjust the rehabilitation assessment judgment level, and obtain the rehabilitation status level.

8. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The abdominal pressure triggering moment refers to the time point at which the abdominal pressure acquisition device records the beginning of changes in abdominal pressure. The training phase refers to the intervals in the same training session divided in chronological order: resting preparation, active contraction, relaxation and recovery, abdominal pressure induction, and training end.

9. The method for assessing pelvic floor muscle rehabilitation training for stress urinary incontinence according to claim 1, characterized in that, The resting phase refers to the recording phase in which the patient does not actively contract the pelvic floor muscles and there are no movement-induced changes in abdominal pressure. The abdominal pressure fluctuation refers to the changes in abdominal pressure recorded due to coughing, standing up, bending over, holding one's breath, or exertion.

10. A stress urinary incontinence pelvic floor muscle rehabilitation training assessment system, said system being used to implement the stress urinary incontinence pelvic floor muscle rehabilitation training assessment method as described in any one of claims 1-9, characterized in that, The system includes: The time sequence construction module is based on rehabilitation training data. It analyzes the time markers and upload order of the same training session, compares the electromyographic start and stop points of the pelvic floor muscles with the triggering time of abdominal pressure, determines the training stage to which the urinary leakage record belongs, and obtains the session time sequence chain. The resting reference module acquires the electromyographic signals of the pelvic floor muscles during the resting phase based on the time sequence chain of the field, filters the stable segments before the abdominal pressure fluctuation, compares the connection state between the active contraction and the resting segment, and obtains the resting reference segment. Based on the resting reference segment, the contraction attribution module analyzes the relationship between the starting point and the ending point of electromyography, filters continuous contraction segments, compares the changes and maintenance states of pelvic floor muscle pressure, determines the attribution of abdominal pressure compensation, and obtains the contraction attribution sequence. Based on the contraction attribution sequence, the fatigue localization module analyzes the timing of effective contraction events, selects the first-stage target events as tolerance references, compares the deviation positions of the subsequent rising rhythm and the maintenance state, and obtains the fatigue boundary sequence. The grading module, based on the fatigue boundary sequence, checks the peak time of abdominal pressure, compares the proximity of the pelvic floor muscle contraction coverage area with the leakage trigger record, determines the order of load, fatigue, and compensation, and obtains the recovery status level.

Images