Postpartum pelvic floor dysfunction disease rehabilitation training system based on multi-mode ultrasound

Abstract

The invention provides a rehabilitation training system for postpartum pelvic floor dysfunction diseases based on multi-modal ultrasound, and the rehabilitation training system integrates multi-modal evaluation, personalized graded training and dynamic training scheme adjustment. The method comprises the steps that pelvic floor two-dimensional ultrasonic parameters, pelvic floor three-dimensional ultrasonic parameters and shear wave elastography (SWE) parameters are collected, a target muscle area is automatically positioned through an image recognition model, the average gray value of an area of interest is calculated, if the average gray value is larger than a preset muscle gray threshold value, effective positioning is judged, and an evaluation result is obtained based on effective positioning; the rehabilitation training module is used for grading based on the evaluation result, formulating a corresponding rehabilitation training scheme, executing the rehabilitation training scheme and recording the rehabilitation training scheme in real time at the same time; and quantifying the evaluation result difference through a statistical analysis algorithm, and if the evaluation result is not improved, automatically adjusting the rehabilitation training scheme according to the non-improvement degree.

Description

Technical Field

[0001] This invention relates to the field of medical equipment and rehabilitation training technology, specifically a rehabilitation training system for postpartum pelvic floor dysfunction based on multimodal ultrasound. Background Technology

[0002] Postpartum pelvic floor dysfunction is a common complication after childbirth, mainly including urinary incontinence, pelvic organ prolapse, defecation difficulties, sexual dysfunction, and pelvic floor pain. Its incidence is high and seriously affects women's quality of life and physical and mental health. With the change of fertility concepts and the increasing emphasis on postpartum health, the clinical need for early assessment, precise intervention and long-term management of pelvic floor dysfunction is becoming increasingly urgent. Current clinical methods for assessing pelvic floor function have significant limitations. Traditional assessments often rely on single ultrasound modalities or subjective symptom questionnaires, only obtaining morphological indicators or qualitative symptom descriptions, failing to simultaneously quantify core parameters such as the elasticity of pelvic floor muscles and the integrity of supporting structures. Although shear wave elastography can reflect muscle elasticity and three-dimensional ultrasound can accurately measure the area of ​​the levator ani hiatus, current technologies have not systematically integrated these two with two-dimensional ultrasound, resulting in one-sided assessment results that fail to comprehensively reflect the state of pelvic floor function. Regarding rehabilitation training, existing programs lack personalized grading mechanisms, often employing a uniform training model that cannot tailor plans to the degree of functional impairment in postpartum women. Furthermore, monitoring training effectiveness lacks scientific quantitative standards, relying solely on subjective feelings or data from single-point follow-up examinations. When training results are unsatisfactory, current technologies lack clear criteria for determining no improvement and automated adjustment mechanisms, leading to delayed program adjustments and impacting rehabilitation efficiency. Therefore, a rehabilitation training system that integrates multimodal assessment, personalized grading training, and dynamically adjusted training programs is urgently needed. Summary of the Invention

[0003] In order to solve the technical problems mentioned in the background art, the purpose of this invention is to provide a multimodal ultrasound rehabilitation training system for postpartum pelvic floor dysfunction.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: A rehabilitation training system for postpartum pelvic floor dysfunction based on multimodal ultrasound includes: The multimodal evaluation module collects two-dimensional ultrasound parameters, three-dimensional ultrasound parameters, and SWE parameters of the pelvic floor and automatically locates the target muscle region through an image recognition model. It calculates the average gray value of the region of interest. If the average gray value is greater than the preset muscle gray value threshold, it is determined to be a valid location. The evaluation result is obtained based on the valid location. The rehabilitation training module classifies the assessment results and formulates corresponding rehabilitation training plans, executes the rehabilitation training plans and records them in real time; it quantifies the differences in assessment results through statistical analysis algorithms, and if the assessment results show no improvement, it automatically adjusts the rehabilitation training plan according to the degree of lack of improvement.

[0005] Furthermore, the two-dimensional ultrasound parameters of the pelvic floor include the bladder neck position and the posterior vesicourethral angle in the resting state and the maximum Valsalva state; the three-dimensional ultrasound parameters of the pelvic floor include the area of ​​the levator ani hiatus in the resting state and the maximum Valsalva state; and the SWE parameters include the Young's modulus values ​​of the bilateral puborectalis muscles in the resting state and the maximum anal contraction state.

[0006] Furthermore, the process by which the image recognition model locates the target muscle region is as follows: 1) Key structure detection: The image recognition model automatically identifies three markers in the image: the bladder neck (marked as L1), the pubic symphysis (marked as L2), and the anorectal junction (marked as L3). The relative distance between the markers is calculated to confirm the validity of the image positioning. 2) Muscle region segmentation: Based on marker point localization, the image recognition model outputs the bounding boxes of the bilateral puborectalis muscle regions; 3) Within the segmented muscle region, the model automatically avoids interfering areas and defines a region at the point of highest muscle fiber density. The circular region of interest; The formula for calculating the average gray value of the region of interest is:

[0007] in, This represents the average grayscale value of a single region of interest; n represents the total number of pixels within the region of interest. Indicates the first [unit] within the region of interest. i The grayscale value of each pixel; The preset muscle grayscale threshold ,like If so, the target muscle region localization is deemed effective; if If this is triggered, the image recognition model will automatically adjust the position of the region of interest during the repositioning process.

[0008] Furthermore, the bladder neck position in the maximum Valsalva state minus the bladder neck position in the resting state is obtained as the bladder neck mobility; the bladder urethral posterior angle in the maximum Valsalva state minus the bladder urethral posterior angle in the resting state is obtained as the urethral rotation angle. The assessment results include bladder neck mobility, urethral rotation angle, levator ani hiatus area, and Young's modulus values ​​of the bilateral puborectalis muscles. Bladder neck mobility greater than 15 mm, urethral rotation angle greater than 30 degrees, levator ani hiatus area greater than 25 cm², and Young's modulus values ​​of the bilateral puborectalis muscles less than 15 kPa at rest are considered to be outside the normal range, and the increase in maximum Valsalva state compared to the resting state is greater than 5 kPa.

[0009] Furthermore, based on the assessment results, pelvic floor function is divided into four levels: Normal level: All parameters of the assessment results are within the normal range. Mild impairment level, where one or two parameters in the assessment results exceed the normal range; Moderate impairment level, where two or more parameters in the assessment results exceed the normal range; Severe disability level: Two or more parameters in the assessment results are outside the normal range and are accompanied by pelvic organ prolapse and stress urinary incontinence; The rehabilitation training programs corresponding to each level are as follows: The rehabilitation training program corresponding to the normal level includes home-based self-training, health management and follow-up monitoring; The rehabilitation training program corresponding to mild disability includes equipment training, home-assisted training, and effect monitoring; The rehabilitation training program corresponding to moderate disability level includes intensive equipment training, specialized functional training, home-based intensive training, and dynamic adjustment; The rehabilitation training program corresponding to the severe disability level includes early intervention, low-intensity initiation training, biofeedback, phased upgrades, and long-term management.

[0010] Furthermore, the statistical analysis algorithm calculates the parameter change by measuring the parameter difference between each time point and the baseline through difference quantification, using the following formula:

[0011] in, Indicates the amount of parameter change; express Parameter values ​​at specific times; express Parameter values ​​at specific times; The change in the parameter is converted into a relative improvement rate using the following formula:

[0012] in, express Relative improvement rate at a given time point; The P-value is calculated using repeated measures ANOVA, where P is the ratio of time point variation to random error; P < 0.05 indicates abnormal non-random error.

[0013] Furthermore, at a certain point in time , and The following conditions are used to determine whether there has been any improvement each month. If any of the following conditions are met, it is determined that there has been no improvement: Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. ; Frequency of urinary incontinence ; Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. ; Frequency of urinary incontinence ; Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. ; Frequency of urinary incontinence .

[0014] Furthermore, based on the determination of whether the conditions have been improved, the severity of no improvement is categorized: Mild symptoms showed no improvement; one parameter did not reach the threshold. Moderate improvement was observed, and two parameters did not reach the threshold. There was no improvement in severe cases, and more than two parameters failed to reach the threshold.

[0015] Compared with the prior art, the advantages of the present invention are as follows: 1. This invention integrates pelvic floor two-dimensional ultrasound, pelvic floor three-dimensional ultrasound and SWE technology to construct a comprehensive assessment system that not only captures morphological abnormalities such as cystocele and uterine prolapse, but also quantifies muscle activity indicators, thus solving the one-sided problem of existing assessments that focus on morphology and neglect function. 2. This invention uses multimodal assessment results to clearly classify pelvic floor function into four levels: normal, mild impairment, moderate impairment, and severe impairment, and designs targeted training programs to significantly improve the rehabilitative effectiveness and safety. 3. This invention establishes a full-cycle monitoring system covering 6 weeks, 3 months, 6 months and 12 months postpartum, and automatically adjusts the plan according to the severity of the lack of improvement, without the need for manual intervention, thus achieving timely and scientific optimization of the plan. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the system workflow of the present invention; Figure 2 This is a schematic diagram of the multimodal evaluation module of the present invention; Figure 3 This is a schematic diagram of the rehabilitation training module of the present invention. Detailed Implementation

[0018] To achieve the above objectives, the present invention provides a multimodal ultrasound-based rehabilitation training system for postpartum pelvic floor dysfunction. Please refer to [link to relevant documentation]. Figures 1 to 3 The system includes the following modules: The multimodal evaluation module collects two-dimensional ultrasound parameters, three-dimensional ultrasound parameters, and SWE parameters of the pelvic floor and automatically locates the target muscle region through an image recognition model. It calculates the average gray value of the region of interest. If the average gray value is greater than the preset muscle gray value threshold, it is determined to be a valid location. The evaluation result is obtained based on the valid location. One hour before the collection, the subject should drink 500ml of water to moderately fill the bladder, which will facilitate clear visualization of the bladder neck and urethral structures. The rectum should be emptied 30 minutes before the collection to reduce interference from intestinal gas. The subject should be instructed to practice the maximum Valsalva maneuver and the maximum anal sphincter contraction exercise to ensure that the movements are performed correctly and to ensure data consistency. The maximum Valsalva maneuver involves holding one's breath with force to simulate an increase in abdominal pressure. The maximum anal sphincter contraction exercise involves actively contracting the pelvic floor muscles and holding the contraction for 5 seconds to simulate an active muscle contraction. The subject should be instructed to practice the movements correctly and to ensure data consistency. The device uses a multi-probe setup, acquiring data in the order of two-dimensional ultrasound, then three-dimensional ultrasound, and finally SWE. During the acquisition process, the effectiveness of the subject's movements is monitored in real time by the motion quality control sensor built into the probe. Two-dimensional ultrasound parameters of the pelvic floor were collected, including the position of the bladder neck and the posterior vesicourethral angle in the resting state and the maximum Valsalva state. At the same time, the presence of cystocele, uterine prolapse, anterior rectal wall prolapse and urethral infundibulum formation were recorded, which are typical morphological manifestations of pelvic floor dysfunction. Scan the levator ani hiatus area, from the posterior margin of the pubic symphysis to the center of the levator ani ring behind the anorectal junction, and collect pelvic floor three-dimensional ultrasound parameters. The pelvic floor three-dimensional ultrasound parameters include the area of ​​the levator ani hiatus, which is a core indicator of weak pelvic floor support structure when the area increases. SWE parameters were collected at rest and at maximum anal sphincter contraction. At rest, when the subject was relaxed, [the following parameters were collected]. The circular region of interest was placed in the middle of the muscle fibers of the bilateral puborectalis muscles, avoiding the intermuscular spaces, blood vessels, and adipose tissue. Young's modulus values ​​of the bilateral puborectalis muscles were measured. Under maximal anal contraction, the subject actively contracted the pelvic floor muscles, and the elasticity changes during muscle contraction were recorded. Each muscle was measured three times at rest and under maximal anal contraction, and the average value was taken to ensure the stability of the elasticity values.

[0019] The acquired pelvic floor 2D ultrasound parameters, 3D ultrasound parameters, and SWE parameters were preprocessed. Gaussian filtering and histogram equalization algorithms were used to remove ultrasound artifacts and enhance the grayscale contrast of muscle tissue and surrounding tissues. All ultrasound parameters were converted into a pelvic floor standard coordinate system with the lower edge of the pubic symphysis as the origin, the horizontal x-axis to the right and the vertical y-axis downward. The image resolution was uniformly set to 512×512 pixels to eliminate coordinate deviations caused by different acquisition angles. An image recognition model based on YOLOv8 object detection is used. The process of locating the target muscle region using this image recognition model is as follows: 1) Key structure detection: The image recognition model automatically identifies three markers in the image: the bladder neck (marked as L1), the pubic symphysis (marked as L2), and the anorectal junction (marked as L3). The relative distance between the markers is calculated to confirm the validity of the image positioning. 2) Muscle region segmentation: Based on marker point localization, the image recognition model outputs the bounding boxes of the bilateral puborectalis muscle regions; 3) Within the segmented muscle region, the model automatically avoids areas interfering with blood vessels and fat, defining a region at the point of highest muscle fiber density. The circular region of interest is used as the target region for subsequent calculation of the average gray value; Extract the grayscale values ​​of all pixels within each region of interest. In this embodiment, the grayscale values ​​for muscle tissue are 100-180, for adipose tissue 50-80, and for blood vessels 0-30. Calculate the average grayscale value for each region of interest using the following formula:

[0020] in, This represents the average grayscale value of a single region of interest; n represents the total number of pixels within the region of interest. Indicates the first [unit] within the region of interest. i The grayscale value of each pixel; The preset muscle grayscale threshold In this embodiment, based on the grayscale statistical data of pelvic floor muscles from 500 healthy postpartum women, a 95% confidence interval lower limit method was used to set the limit; if This triggers a repositioning process where the image recognition model automatically adjusts the region of interest and recalculates. until the location is valid; if If so, the target muscle region is determined to be effectively located; Based on the parameter data of effective positioning, the evaluation results are generated according to four core evaluation indicators and compared with the normal range. The four core evaluation indicators include bladder neck mobility, urethral rotation angle, levator ani hiatus area and Young's modulus value of bilateral puborectalis muscles. The bladder neck mobility and urethral rotation angle are calculated based on the two-dimensional ultrasound parameters of the pelvic floor. The bladder neck mobility is the bladder neck position in the maximum Valsalva state minus the bladder neck position in the resting state. The urethral rotation angle is the posterior bladder urethral angle in the maximum Valsalva state minus the posterior bladder urethral angle in the resting state. The assessment results showed that a bladder neck mobility greater than 15 mm and a urethral rotation angle greater than 30 degrees were considered outside the normal range, indicating decreased bladder neck stability and a risk of urinary incontinence. The presence of morphological abnormalities such as cystocele and uterine prolapse was also recorded. A hiatus area greater than 25 cm² was considered outside the normal range, indicating weak levator ani muscle support and serving as an important early warning indicator for pelvic organ prolapse. A resting Young's modulus value of less than 15 kPa for both puborectalis muscles and an increase of less than 5 kPa at maximum Valsalva maneuver were considered outside the normal range, indicating poor muscle elasticity and reduced contractile function. The rehabilitation training module classifies the assessment results and formulates corresponding rehabilitation training plans, executes the rehabilitation training plans and records them in real time; it quantifies the differences in assessment results through statistical analysis algorithms, and if the assessment results show no improvement, it automatically adjusts the rehabilitation training plan according to the degree of lack of improvement.

[0021] Based on the assessment results, pelvic floor function is divided into 4 levels: Normal level: All parameters of the assessment results are within the normal range, there are no symptoms such as pelvic organ prolapse and urinary incontinence, no regular rehabilitation training is required, only routine health education is needed; Mild impairment level: 1 to 2 parameters in the assessment results are outside the normal range. In this example, the bladder neck mobility is 15-20 mm, the levator ani hiatus area is 25-30 cm², there are no obvious clinical symptoms or the symptoms are mild. If there is occasional mild urinary incontinence, it is recommended to start basic rehabilitation training. Moderate impairment level, with two or more parameters exceeding the normal range. In this embodiment, the bladder neck mobility is greater than 20mm, the levator ani hiatus area is greater than 30cm², the Young's modulus value of the bilateral puborectalis muscles is less than 12kPa, accompanied by clear symptoms such as frequent urinary incontinence and pelvic distension. Intensive rehabilitation training is required. Severe disability level: Two or more parameters in the assessment results are severely abnormal and accompanied by pelvic organ prolapse and stress urinary incontinence. In this embodiment, uterine prolapse of grade 2 or above requires comprehensive clinical treatment, with rehabilitation training as an auxiliary means. The rehabilitation training program corresponding to the normal level focuses on prevention and maintenance, requiring no regular equipment training. It only requires self-management through health education: Home self-training, guided Kegel exercises, contracting the pelvic floor muscles for 3 seconds and relaxing for 3 seconds, 15 minutes each time, twice daily, using a mobile app for timer reminders to ensure correct contraction and avoid compensation by abdominal or thigh muscles; Health management includes a daily intake of at least 25g of dietary fiber to prevent constipation and reduce long-term increases in abdominal pressure; and weight management to control BMI within a certain range. To avoid obesity, which is an independent risk factor for pelvic floor dysfunction; follow-up monitoring, with telephone follow-ups every 3 months to monitor symptom changes, and a final multimodal assessment to be completed 12 months postpartum to confirm long-term stability; The rehabilitation training program corresponding to mild impairment level initiates a combined program of basic equipment training and self-training, with biofeedback correction as the core: Equipment training utilizes a PHENIX USB4 electrical stimulation-biofeedback device with a 30mm diameter vaginal electrode; device parameters include electrical stimulation set to low-frequency pulses of 50Hz and a current intensity of 15-20mA, with the threshold being the subjective feeling of pelvic floor muscle contraction without discomfort; and biofeedback settings display real-time electromyographic waveforms to guide the maintenance of peak electromyographic activity during contraction. Avoid excessive exertion; twice a week, 20 minutes each time, including 5 minutes of electrode placement and 15 minutes of training, for a total of 10 sessions; the training position is supine with knees bent at 90°, and the electrode placement depth is 4-5cm. Before each training session, confirm that the bladder is adequately full using two-dimensional ultrasound to avoid overfilling and affecting the contraction sensation; home-assisted training, one Kegel exercise per day, the same as the normal level, and record the contraction quality in conjunction with the training. Clear contraction of the pelvic floor muscles is recorded as effective; effect monitoring, subjective comfort is assessed after every 2 sessions of device training, and a multimodal assessment is performed 3 months postpartum. If the parameters return to normal, switch to the rehabilitation training program corresponding to the normal level. The rehabilitation training program for moderate impairment level employs intensive equipment training, specific feedback, and home-based reinforcement to improve muscle contraction and coordination: Intensive equipment training uses the same equipment as for mild impairment, increasing compensatory muscle activity; equipment parameters are set to low-frequency pulses of 50Hz, with the current intensity increased to 20-25mA, not exceeding 30mA, and the duration of a single training session extended to 25 minutes, including 5 minutes of alternating contraction and relaxation specific training; biofeedback incorporates anti-compensation training, where abdominal electromyography signals are detected to be greater than... When the instrument automatically pauses and prompts for adjustment of the force application method; 3 times a week, with an interval of 1-2 days between each session, for a total of 12 sessions, the first 6 sessions are intensive training, and the last 6 sessions are consolidation; Specific functional training, after each instrument training session, add 5 minutes of maximum anal contraction maintenance training, hold the contraction for 5 seconds and then relax for 3 seconds, repeat 10 sets, combined with SWE real-time monitoring of muscle elasticity changes, the target is that the Young's modulus value of both puborectalis muscles in the maximum Valsalva state increases by more than 5 kPa compared with the resting state; Home strengthening training, 2 Kegel training sessions per day for 20 minutes each, using a home portable biofeedback device to record the duration of each contraction, the target is to gradually increase from 3 seconds to 8 seconds; Dynamic adjustment, assessment at 3 months postpartum, if the bladder neck mobility decreases by more than 5 mm and the levator ani hiatus area decreases by more than 3 cm², maintain the plan until the end of the treatment course; if the improvement is less than 3 mm, increase the current intensity by 2-3 mA but not exceeding 30 mA, increase electrical stimulation and active contraction synchronous training, active contraction during stimulation, enhance neuromuscular linkage; The rehabilitation training program corresponding to the severe disability level follows the principles of clinical collaboration, low-intensity initiation, and gradual progression to avoid aggravating the injury: Initial intervention involves suspending independent rehabilitation training and referring the patient to the obstetrics and gynecology department for clinical evaluation. If necessary, symptomatic treatment such as a uterine pessary and medication is used. Once symptoms subside, in this embodiment, training is initiated after prolapse is reduced to grade 1 and urinary incontinence occurs less than 3 times per week. Low-intensity initiation training uses the same equipment as the moderate disability training program, but a soft vaginal electrode with a diameter of 25mm is selected to reduce the stimulation intensity. Equipment parameters include an initial electrical stimulation current of 8-10mA, ensuring no pain, and each training session lasts 15 minutes with only passive stimulation. Stimulation and gentle active contraction, twice a week; biofeedback is used to prioritize correcting problems of complete inability to contract or overcompensation, with the goal of completing basic anal contraction movements; phased upgrades, with assessments every 4 weeks. If the current intensity is tolerated and parameters do not worsen, the current is gradually increased by 2mA each time, not exceeding 20mA, and the duration is extended by 5 minutes each time, not exceeding 25 minutes, until moderate obstacle training parameters are reached. If pain or parameter deterioration occurs, the treatment is immediately suspended and returned to clinical management; long-term management, with the course extended to 16-20 sessions. A comprehensive assessment in conjunction with obstetrics and gynecology is required at 6 months and 12 months postpartum to develop a long-term maintenance plan; Before each training session, ensure that lochia has stopped, there is no vaginal bleeding or inflammation, and that the bladder is adequately full. For adverse reactions, if vaginal discomfort occurs, reduce the current intensity by 2-3 mA. If pain or bleeding persists, suspend training and investigate the cause. Two-dimensional ultrasound parameters, three-dimensional ultrasound parameters, and SWE parameters of the pelvic floor were extracted from the same subject at 6 weeks, 3 months, 6 months, and 12 months postpartum and preprocessed. The preprocessing steps included removing outliers, standardizing units and formats to convert all parameters to standard units, quantifying symptom data into numerical values, and filling missing values ​​with the median of the parameter from subjects of the same level. Statistical analysis algorithms calculate the magnitude of parameter changes and statistical differences through difference quantification and significance testing; the parameter change is quantified by longitudinal differences, calculating the parameter difference between each time point and the baseline, reflecting the magnitude of change. The formula is:

[0022] in, Indicates the amount of parameter change; express Parameter values ​​at specific times; express Parameter values ​​at specific times; Regarding bladder neck mobility and levator ani hiatus area. Negative values ​​indicate improvement, with smaller values ​​indicating more significant improvement; positive values ​​indicate deterioration. For the Young's modulus value of the puborectalis muscle... Positive values ​​indicate improvement, with higher values ​​indicating better elasticity; negative values ​​indicate deterioration. Regarding the frequency of urinary incontinence, Negative values ​​indicate improvement, while positive values ​​indicate deterioration. The parameter changes are converted into relative improvement rates to eliminate the influence of baseline differences. The formula is as follows:

[0023] in, express The relative improvement rate at a given time point, with the absolute value ensuring a positive result; a larger value indicates a more significant relative improvement. Before calculating the differences in evaluation results, a normality test was performed using the Shapiro-Wilk test. For each evaluation result parameter, the data at four time points were examined to determine if they conformed to a normal distribution. For normally distributed data, a homogeneity of variance test was used to verify the homogeneity of variance across different time points. For data that conformed to both a normal distribution and homogeneity of variance, repeated measures ANOVA was used to calculate the ratio of time point differences to random error. P Values; for parameters that do not conform to a normal distribution, the Friedman test is used, which is applicable to non-normally distributed data. Calculation PThe value is used to determine whether the difference is caused by random error; if P < 0.05, the difference is not caused by random error. Based on the natural recovery process of the pelvic floor after childbirth, different time points The improvement thresholds are different, combined with and If any of the following conditions are met, it is determined that there is no improvement: Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. The rise is less than 3 kPa, or ; Frequency of urinary incontinence The frequency drop is less than 0.5 levels, or ; Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. The rise is less than 5 kPa, or ; Frequency of urinary incontinence The frequency drop is less than 1.0 level, or ; Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. The rise is less than 8 kPa, or ; Frequency of urinary incontinence The frequency drop is less than 1.5 levels, or ; Based on the severity of the lack of improvement, an adjustment plan is automatically generated: Mild improvement, one parameter not reaching the threshold. In this embodiment, When one parameter fails to reach the threshold, based on the relative improvement rate The difference between the training rate and the target improvement rate is used to proportionally increase the training intensity; the electrical stimulation current intensity is adjusted, the training duration is extended by 5 minutes, and biofeedback is used to enhance the target peak value of the electromyographic signal from... Upgraded to This enhances the contraction force; Moderate improvement was observed, and two parameters did not reach the threshold. In this embodiment, When one parameter does not reach the threshold, based on The absolute value of the puborectalis muscle was adjusted, the intensity was doubled, and the training mode was adjusted; the electrical stimulation current intensity was adjusted, and the training mode was optimized, changing from continuous stimulation to intermittent stimulation (10 seconds stimulation, 5 seconds rest), combined with real-time SWE monitoring. Pause stimulation at certain times to strengthen muscle memory; If there is no improvement and more than two parameters fail to reach the threshold, in this embodiment... More than two parameters still did not reach the threshold; based on Long-term differences from baseline trigger clinical co-adjustment, linking parameter adjustments to clinical assessments. Self-training is paused, and after clinical exclusion of comorbidities, training parameters are reset, with the electrical stimulation frequency adjusted from... 50Hz Change to 30-50Hz Frequency conversion enhances nerve sensitivity; the current intensity is adjusted to 1.2 times the original intensity, without exceeding the safety threshold. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A rehabilitation training system for postpartum pelvic floor dysfunction based on multimodal ultrasound, characterized in that, include: The multimodal evaluation module collects two-dimensional ultrasound parameters, three-dimensional ultrasound parameters, and SWE parameters of the pelvic floor and automatically locates the target muscle region through an image recognition model. It calculates the average gray value of the region of interest. If the average gray value is greater than the preset muscle gray value threshold, it is determined to be a valid location. The evaluation result is obtained based on the valid location. The rehabilitation training module classifies the assessment results and formulates corresponding rehabilitation training plans, executes the rehabilitation training plans and records them in real time; it quantifies the differences in assessment results through statistical analysis algorithms, and if the assessment results show no improvement, it automatically adjusts the rehabilitation training plan according to the degree of lack of improvement.

2. The system according to claim 1, characterized in that, The two-dimensional ultrasound parameters of the pelvic floor include the position of the bladder neck and the posterior vesicourethral angle in the resting state and the maximum Valsalva state; the three-dimensional ultrasound parameters of the pelvic floor include the area of ​​the levator ani hiatus in the resting state and the maximum Valsalva state; the SWE parameters include the Young's modulus values ​​of the bilateral puborectalis muscles in the resting state and the maximum anal contraction state.

3. The system according to claim 2, characterized in that, The process by which the image recognition model locates the target muscle region is as follows: 1) Key structure detection: The image recognition model automatically identifies three markers in the image: the bladder neck (marked as L1), the pubic symphysis (marked as L2), and the anorectal junction (marked as L3). The relative distance between the markers is calculated to confirm the validity of the image positioning. 2) Muscle region segmentation: Based on marker point localization, the image recognition model outputs the bounding boxes of the bilateral puborectalis muscle regions; 3) Within the segmented muscle region, the model automatically avoids interfering areas and defines a region at the point of highest muscle fiber density. The circular region of interest; The formula for calculating the average gray value of the region of interest is: in, This represents the average grayscale value of a single region of interest; n represents the total number of pixels within the region of interest. Indicates the first [unit] within the region of interest. i The grayscale value of each pixel; The preset muscle grayscale threshold ,like If so, the target muscle region localization is deemed effective; if If this is triggered, the image recognition model will automatically adjust the position of the region of interest during the repositioning process.

4. The system according to claim 3, characterized in that, The bladder neck mobility is obtained by subtracting the bladder neck position at rest from the bladder neck position at maximum Valsalva state; the urethral rotation angle is obtained by subtracting the posterior urethral angle at rest from the posterior urethral angle at maximum Valsalva state. The assessment results include bladder neck mobility, urethral rotation angle, levator ani hiatus area, and Young's modulus values ​​of the bilateral puborectalis muscles. Bladder neck mobility greater than 15 mm, urethral rotation angle greater than 30 degrees, levator ani hiatus area greater than 25 cm², and Young's modulus values ​​of the bilateral puborectalis muscles less than 15 kPa at rest are considered to be outside the normal range. The increase in maximum Valsalva state compared to the resting state is greater than 5 kPa.

5. The system according to claim 4, characterized in that, Based on the assessment results, pelvic floor function is divided into four levels: Normal level: All parameters of the assessment results are within the normal range. Mild impairment level, where one or two parameters in the assessment results exceed the normal range; Moderate impairment level, where two or more parameters in the assessment results exceed the normal range; Severe disability level: Two or more parameters in the assessment results are outside the normal range and are accompanied by pelvic organ prolapse and stress urinary incontinence; The rehabilitation training programs corresponding to each level are as follows: The rehabilitation training program corresponding to the normal level includes home-based self-training, health management and follow-up monitoring; The rehabilitation training program corresponding to mild disability includes equipment training, home-assisted training, and effect monitoring; The rehabilitation training program corresponding to moderate disability level includes intensive equipment training, specialized functional training, home-based intensive training, and dynamic adjustment; The rehabilitation training program corresponding to the severe disability level includes early intervention, low-intensity initiation training, biofeedback, phased upgrades, and long-term management.

6. The system according to claim 5, characterized in that, The statistical analysis algorithm calculates the parameter change by measuring the difference between the parameter at each time point and the baseline through difference quantification. The formula is as follows: in, Indicates the amount of parameter change; express Parameter values ​​at specific times; express Parameter values ​​at specific times; The change in the parameter is converted into a relative improvement rate using the following formula: in, express Relative improvement rate at a given time point; The P-value is calculated using repeated measures ANOVA, where P is the ratio of time point variation to random error; P < 0.05 indicates abnormal non-random error.

7. The system according to claim 6, characterized in that, At the point of time , and The following conditions are used to determine whether there has been any improvement each month. If any of the following conditions are met, it is determined that there has been no improvement: Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. ; Frequency of urinary incontinence ; Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. ; Frequency of urinary incontinence ; Bladder neck mobility ; Area of ​​the levator ani hiatus ; Young's modulus of the puborectalis muscle. ; Frequency of urinary incontinence .

8. The system according to claim 7, characterized in that, Based on the criteria for determining whether improvement is possible, the severity of no improvement is categorized as follows: Mild symptoms showed no improvement; one parameter did not reach the threshold. Moderate improvement was observed, and two parameters did not reach the threshold. There was no improvement in severe cases, and more than two parameters failed to reach the threshold.

Images