A method and system for pelvic floor dysfunction rehabilitation assessment

Abstract

The application relates to the technical field of intelligent medical treatment, and particularly provides a pelvic floor dysfunction rehabilitation evaluation method and system, which comprises the following steps: collecting electrical signal data of pelvic floor muscles of a current patient to be measured, patient action posture data and pelvic floor image data; inputting the electrical signal data, the patient action posture data and the pelvic floor image data into an adaptive evaluation model to obtain a current node evaluation result; inputting the current node evaluation result and a previous node evaluation result into a forward reconstruction equation to obtain a next node evaluation result; determining a pelvic floor dysfunction rehabilitation evaluation curve based on the current node evaluation result, the previous node evaluation result and the next node evaluation result; and determining a final pelvic floor dysfunction rehabilitation evaluation result based on the pelvic floor dysfunction rehabilitation evaluation curve, so that the application can improve the accuracy of pelvic floor dysfunction rehabilitation evaluation.

Description

Technical Field

[0001] This application relates to the field of intelligent medical technology, and in particular to a method and system for rehabilitation assessment of pelvic floor dysfunction. Background Technology

[0002] Pelvic floor dysfunction is a common condition among middle-aged and elderly people, as well as those who have undergone prostate cancer or low anterior rectal cancer surgery, severely impacting their quality of life and recovery outcomes. Currently, the effectiveness of biofeedback combined with pelvic floor muscle exercises in improving pelvic floor dysfunction and anterior resection syndrome for rectal cancer has been validated by scholars both domestically and internationally. However, current pelvic floor rehabilitation assessments primarily rely on hospital pelvic floor clinics.

[0003] Existing assessment methods have several drawbacks: First, the assessment is highly subjective, relying mainly on the patient's own feelings or simple manual testing, lacking objective data support; second, the accuracy is low, often using single-modal data for assessment, which cannot fully reflect the functional status of the pelvic floor muscles, resulting in significant deviations in assessment results; third, patient compliance is poor, lacking real-time feedback and supervision, making it difficult to judge the standardization of their own training movements, thus affecting the rehabilitation effect.

[0004] In existing technologies, some pelvic floor rehabilitation assessment methods attempt to introduce neural network algorithms, but most are limited to single-modal data input, failing to achieve effective fusion of multimodal data and fully utilize the complementary advantages of different modal data, and the assessment accuracy still needs to be improved. Summary of the Invention

[0005] One objective of this application is to provide a method for assessing pelvic floor dysfunction rehabilitation, at least to address the question of how to improve the accuracy of pelvic floor dysfunction rehabilitation assessment.

[0006] To achieve the above objectives, some embodiments of this application provide the following aspects: In a first aspect, some embodiments of this application also provide a method for rehabilitation assessment of pelvic floor dysfunction, including: The system collects electrical signal data of the pelvic floor muscles, patient movement and posture data, and pelvic floor image data of the patient currently being tested; wherein, the electrical signal data includes bioelectrical signals and pressure signals; The electrical signal data, the patient's movement and posture data, and the pelvic floor image data are input into the adaptive evaluation model to obtain the evaluation result of the current node; The current node evaluation result and the previous node evaluation result are input into the forward reconstruction equation to obtain the next node evaluation result; Based on the current node assessment result, the previous node assessment result, and the next node assessment result, a pelvic floor dysfunction rehabilitation assessment curve is determined. Based on the pelvic floor dysfunction rehabilitation assessment curve, the final pelvic floor dysfunction rehabilitation assessment result is determined.

[0007] Secondly, some embodiments of this application also provide a pelvic floor dysfunction rehabilitation assessment system, the system comprising: The acquisition unit is configured to acquire electrical signal data of the pelvic floor muscles, patient movement and posture data, and pelvic floor image data of the current patient to be tested; wherein the electrical signal data includes bioelectrical signals and pressure signals; The first data processing unit is configured to input the electrical signal data, the patient's movement and posture data, and the pelvic floor image data into the adaptive evaluation model to obtain the current node evaluation result; The second data processing unit is configured to input the current node evaluation result and the previous node evaluation result into the forward reconstruction equation to obtain the next node evaluation result; The third data processing unit is configured to determine the pelvic floor dysfunction rehabilitation assessment curve based on the current node assessment result, the previous node assessment result, and the next node assessment result. The fourth data processing unit is configured to determine the final pelvic floor dysfunction rehabilitation assessment result based on the pelvic floor dysfunction rehabilitation assessment curve.

[0008] Thirdly, some embodiments of this application also provide an electronic device, the electronic device comprising: one or more processors; and a memory storing computer program instructions, which, when executed, cause the processor to perform the steps of the method described above.

[0009] Fourthly, some embodiments of this application also provide a computer-readable medium having computer program instructions stored thereon, which can be executed by a processor to implement the method described above.

[0010] Compared to related technologies, the solution provided in this application provides a method for assessing pelvic floor dysfunction rehabilitation, which collects pelvic floor electromyographic signal data, patient movement and posture data, and pelvic floor image data of the patient currently being tested. The electromyographic signal data includes bioelectrical signals and pressure signals. Bioelectrical signals characterize the electrophysiological activity of the pelvic floor muscles, and pressure signals characterize the pressure changes during pelvic floor muscle contraction. The movement and posture signals monitor the patient's posture adjustments during rehabilitation, reflecting the synergy between pelvic floor muscles and body movements. The pelvic floor image data visually presents the morphology and structural changes of the pelvic floor tissues, providing intuitive morphological support for functional assessment. After signal acquisition, the electromyographic signal data, patient movement and posture data, and pelvic floor image data are synchronously input into a pre-trained adaptive assessment model, outputting the pelvic floor dysfunction rehabilitation assessment result for the current node, thus achieving a quantitative determination of the patient's pelvic floor functional status at the current moment. Subsequently, the assessment result of the current node and the assessment result of the previous node are jointly input into the forward reconstruction equation to reconstruct and output the assessment result of the next node, filling the information gaps between discrete sampling nodes and achieving continuous connection between assessment nodes. Based on this, using time as the horizontal axis and the assessment result as the vertical axis, the assessment results of the previous, current, and next nodes are used as key feature nodes. Through continuous interpolation fitting and local trend constraints, abrupt fluctuations between nodes are eliminated, generating a smooth and coherent pelvic floor dysfunction rehabilitation assessment curve that fully presents the dynamic evolution of the patient's pelvic floor function. Finally, based on this rehabilitation assessment curve, the final pelvic floor dysfunction rehabilitation assessment result is determined. Thus, this invention, through the optimization of the entire process of multimodal signal acquisition, adaptive assessment, temporal reconstruction, and curve fitting, effectively avoids the limitations of single-modal signal assessment, reduces the impact of environmental interference and instantaneous noise on the assessment results, and significantly improves the accuracy of pelvic floor dysfunction rehabilitation assessment. Attached Figure Description

[0011] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0012] Figure 1 A flowchart illustrating a rehabilitation assessment method for pelvic floor dysfunction provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of a pelvic floor dysfunction rehabilitation assessment system according to an embodiment of this application; Figure 3 This is a schematic diagram of an exemplary structure of a processor and a memory according to this application; Figure 4 This is an exemplary structural diagram of an electronic device according to this application. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] See appendix Figure 1 , Figure 1 This is a flowchart illustrating a rehabilitation assessment method for pelvic floor dysfunction provided in an embodiment of this application. Figure 1 As shown, the pelvic floor dysfunction rehabilitation assessment method in this application embodiment includes: Step 100: Collect electrical signal data of the pelvic floor muscles, patient movement and posture data, and pelvic floor image data of the current patient to be tested; wherein, the electrical signal data includes bioelectrical signals and pressure signals; Step 102: Input the electrical signal data, patient movement and posture data, and pelvic floor image data into the adaptive evaluation model to obtain the evaluation result of the current node; Step 104: Input the current node evaluation result and the previous node evaluation result into the forward reconstruction equation to obtain the next node evaluation result; Step 106: Based on the assessment results of the current node, the previous node, and the next node, determine the rehabilitation assessment curve for pelvic floor dysfunction; Step 108: Based on the pelvic floor dysfunction rehabilitation assessment curve, determine the final pelvic floor dysfunction rehabilitation assessment result.

[0015] In this embodiment, pelvic floor electromyographic (EMG) data, patient movement and posture data, and pelvic floor image data are collected from the patient currently being tested. The EMG data includes bioelectrical signals and pressure signals. Bioelectrical signals characterize the electrophysiological activity of the pelvic floor muscles, while pressure signals characterize pressure changes during pelvic floor muscle contraction. Movement and posture signals monitor the patient's posture adjustments during rehabilitation, reflecting the synergy between pelvic floor muscles and body movement. Pelvic floor image data visually presents the morphological and structural changes of the pelvic floor tissues, providing intuitive morphological support for functional assessment. After signal acquisition, the EMG data, patient movement and posture data, and pelvic floor image data are simultaneously input into a pre-trained adaptive assessment model, outputting the pelvic floor dysfunction rehabilitation assessment result for the current node, thus quantitatively determining the patient's pelvic floor functional status at the current moment. Subsequently, the assessment result of the current node and the assessment result of the previous node are input together into the forward reconstruction equation to reconstruct and output the assessment result of the next node, filling the information gaps between discrete sampling nodes and achieving continuous connection between assessment nodes. Based on this, using time as the horizontal axis and the assessment result as the vertical axis, the assessment results of the previous, current, and next nodes are used as key feature nodes. Through continuous interpolation fitting and local trend constraints, abrupt fluctuations between nodes are eliminated, generating a smooth and coherent pelvic floor dysfunction rehabilitation assessment curve that fully presents the dynamic evolution of the patient's pelvic floor function. Finally, based on this rehabilitation assessment curve, the final pelvic floor dysfunction rehabilitation assessment result is determined. Thus, this invention, through the optimization of the entire process of multimodal signal acquisition, adaptive assessment, temporal reconstruction, and curve fitting, effectively avoids the limitations of single-modal signal assessment, reduces the impact of environmental interference and instantaneous noise on the assessment results, and significantly improves the accuracy of pelvic floor dysfunction rehabilitation assessment.

[0016] It should also be noted that the calculation method for the evaluation result of the previous node is the same as that for the current node, and the evaluation result of the previous node can be directly retrieved from the historical records.

[0017] In one implementation, the adaptive evaluation model is obtained by training an improved hybrid model, which includes an input layer, a multimodal feature extraction layer, an environment adaptive allocation layer, a modality feature fusion layer, an LSTM layer, a fully connected layer, and an output layer connected in sequence. The input layer receives electrical signal data, patient posture data, and pelvic floor image data. The multimodal feature extraction layer extracts features from the electrical signal data, patient posture data, and pelvic floor image data. The environment adaptive allocation layer assigns weights to the electrical signal data, patient posture data, and pelvic floor image data based on the environmental data of the corresponding sensors. The modality feature fusion layer fuses the features of the weighted electrical signal data, patient posture data, and pelvic floor image data. The LSTM layer captures the temporal correlation of the fused multimodal features and mines the temporal patterns of pelvic floor muscle contraction and relaxation. The fully connected layer filters and integrates the temporal features, converting them into low-dimensional evaluation parameters suitable for the output. The output layer outputs the evaluation result of the current node.

[0018] In this embodiment, the adaptive assessment model for pelvic floor function evaluation is obtained by systematically training an improved hybrid model. This improved hybrid model consists of an input layer, a multimodal feature extraction layer, an environment adaptive allocation layer, a modal feature fusion layer, an LSTM temporal correlation layer, a fully connected layer, and an output layer connected in series. The input layer receives multi-dimensional raw data, including pelvic floor electromyography signals, patient posture data, and pelvic floor image data. The multimodal feature extraction layer extracts features such as amplitude and frequency of electrical signals, coordination features of posture signals, and morphological features of image data for different types of input signals, achieving preliminary screening and refinement of multi-source data. The environment adaptive allocation layer, as the core module, dynamically allocates the weights of various signals based on real-time environmental data from each sensor. Through a differentiated weighting strategy, it highlights the role of core modal signals while suppressing biases caused by environmental interference, ensuring the accuracy of feature extraction. The modal feature fusion layer deeply fuses the various features after weight allocation, integrating multi-dimensional information and eliminating data redundancy. The LSTM layer focuses on capturing the temporal correlation of fused multimodal features, deeply mining the temporal patterns of pelvic floor muscle contraction and relaxation, and accurately extracting its dynamic change features; the fully connected layer further filters and integrates the temporal features, converting high-dimensional features into low-dimensional evaluation parameters that are suitable for the output, thus completing the dimensionality reduction and optimization of the features; finally, the evaluation result of the current node is output through the output layer.

[0019] In this embodiment, the signal-to-noise ratio (SNR) of a single-mode signal is used as the criterion for fluctuation determination, and a unified determination threshold is set: if the SNR of a mode signal is <20dB, it is determined that there is signal fluctuation in the mode; if the SNR of a mode signal is ≥20dB, it is determined that there is no signal fluctuation in the mode.

[0020] In one implementation, when there are signal fluctuations in the electrical signal data, patient posture data, and pelvic floor image data, the environment adaptive allocation layer performs the following calculation logic:

[0021]

[0022] In the formula, The comprehensive environmental interference impact factor for the i-th modal data. These are modality-specific correction coefficients. This is the environmental sensitivity adjustment coefficient. This is a weighting factor for the environmental disturbance level. These are the normalized environmental observations at the current moment. The Sigmoid smoothing constraint function is used. Here, t is the slope adjustment coefficient, and t is the sampling time. This is the time variable for integration.

[0023] In this embodiment, the formula introduces a time-domain integral term. This expands environmental data from a single moment into a cumulative disturbance assessment over a period of time, while also using disturbance level weighting coefficients. Differentiated processing of strong, medium, and weak interference more accurately reflects the overall impact of the environment on signal quality, avoiding misjudgments caused by instantaneous fluctuations and significantly improving the algorithm's robustness in complex home environments. A dual correction mechanism of mode specificity and smoothing constraints is proposed for the first time to address the difference in anti-interference performance between different modes. The formula uses mode specificity correction coefficients... This method differentiates the inherent anti-interference capabilities of different modalities, such as electrical signals, attitude signals, and image data, overcoming the limitation of traditional methods that apply the same interference evaluation standard to all modalities. Simultaneously, it introduces a Sigmoid smoothing constraint function to nonlinearly smooth interference mutations in extreme environments, avoiding abnormal fluctuations in factor values ​​when the environment changes drastically. This makes the weight allocation process more stable and continuous. Each step has adjustable parameters (such as the environmental sensitivity adjustment coefficient k and the slope adjustment coefficient λ), ensuring both the algorithm's universality and allowing for targeted optimization based on different patients and scenarios. The comprehensive environmental interference impact factor of the output i-th modality data can be directly used for subsequent weight allocation, realizing a direct mapping from environmental data to fused weights. This provides a quantitative basis for adaptive fusion of multimodal data and improves the accuracy of pelvic floor dysfunction rehabilitation assessment.

[0024] In this embodiment, the current environmental normalized observation value refers to the quantized value of the original environmental interference parameters detected by the sensors collecting three types of modal data under the current real-time operating conditions, after being uniformly standardized and mapped to a fixed range. This includes external interference parameters such as the real-time electromagnetic interference intensity at the electrical signal acquisition end, the intensity of limb movement disturbance during posture signal acquisition, and the intensity of light noise in the pelvic floor image acquisition environment. In the weight calculation mechanism of the environmental adaptive allocation layer, the modal feature weights are adjusted in real-time based on the comprehensive environmental interference impact factor of the modal data. When the comprehensive environmental interference impact factor of a certain modality increases, the fusion proportion of that modality is automatically reduced to decrease the negative impact of invalid interference data on the overall evaluation; when the comprehensive environmental interference impact factor decreases and the environmental conditions are favorable, the feature weight of that modality is reasonably increased to fully utilize high-quality and effective features in multimodal fusion calculation.

[0025] In this embodiment, the comprehensive environmental interference influence factor for the i-th modality data is uniformly constrained to a range of 0 to 1. A value closer to 1 indicates a stronger degree of environmental interference affecting the corresponding modality, while a value closer to 0 indicates a stable acquisition environment and weaker interference. The modality-specific correction coefficient is set within a range of 0.6 to 0.95, with values ​​differentiated based on the inherent anti-interference capabilities of the three modalities: pelvic floor electromyography (EMG) signals, movement posture signals, and pelvic floor image data. The EMG signal modality has a lower value, the movement posture signal modality uses a value in the middle range, and the pelvic floor image modality has a higher value. This value is a fixed constant and does not change with environmental conditions. The environmental sensitivity adjustment coefficient is set within a range of 1.0 to 3.0, manually preset as a globally unified fixed parameter. A lower value results in a smoother adjustment effect of environmental interference on the modality factor, while a higher value indicates greater sensitivity to minor environmental fluctuations. A fixed value can be selected based on the anti-interference and response requirements of the actual evaluation scenario. The environmental interference level weighting coefficient is set between 0.5 and 1.2, a globally fixed preset parameter. Lower values ​​are selected for application scenarios with weak environmental interference, while higher values ​​are selected for complex, high-interference scenarios at home. This is used to scale the overall magnitude of the environmental cumulative integral term and adjust the interference level. The slope adjustment coefficient is set between 4 and 10, a globally fixed hyperparameter manually set. A smaller value results in a smoother trend in the nonlinear constraint function, effectively smoothing small fluctuations in environmental noise; a larger value results in a steeper trend in the function, enabling rapid response and constraint correction to sudden, strong environmental interference.

[0026] In one implementation, when there are no signal fluctuations in the electrical signal data, patient posture data, and pelvic floor image data, the environment adaptive allocation layer performs the following calculation logic:

[0027] In the formula, For the dynamic weights of the i-th mode, This is the weighting balance coefficient. For modal prior association weights, The modal signal quality coefficient at the current moment. The cumulative integral mean of signal quality degradation. This represents the individual variability coefficient for patients.

[0028] In this embodiment, traditional methods typically use fixed initial weights, which cannot adapt to dynamic changes in signal quality under different scenarios. This formula uses a weight balancing coefficient γ to associate modal priors with weights. The inherent importance of pelvic floor assessment is combined with a dynamic term based on real-time signal quality. This approach retains inherent medical knowledge while adjusting weights according to the reliability of current data, resolving the contradiction that fixed weights cannot simultaneously consider prior knowledge and real-time conditions. The introduction of the cumulative integral mean of signal quality degradation upgrades the assessment formula from single-point quality to long-term stability. The algorithm calculates the average degradation of signal quality over a period of time through time-domain integration. This ensures that weight allocation not only considers the current data quality but also reflects the long-term stability of the signal, effectively avoiding misjudgments of weights caused by instantaneous data fluctuations and significantly improving the reliability of the algorithm in long-term home monitoring scenarios. Furthermore, by incorporating patient individual difference coefficients, the algorithm bridges the gap between a general population model and a personalized adaptation model. The formula introduces patient individual difference coefficients, and the initial weights are adjusted individually based on characteristics such as the patient's age, disease severity, and post-operative recovery stage.

[0029] In this embodiment, the dynamic weight of the i-th modality is constrained to a range of 0 to 1. It is adaptively solved by the model combining prior modality association weights, the current modality signal quality coefficient, the patient's individual variability coefficient, and environmental interference factors, and is dynamically adjusted in real time according to signal quality, individual characteristics, and environmental conditions. The weight balance coefficient is set between 0.3 and 0.7, a globally fixed preset parameter used to balance the distribution ratio between prior modality association weights and real-time signal quality weights. A larger value emphasizes the inherent importance of the modality, while a smaller value emphasizes the influence of real-time signal quality. The prior modality association weight is set between 0.2 and 0.5, and is manually assigned a prior value based on the inherent correlation between the three modalities (pelvic floor electromyography signals, movement posture signals, and pelvic floor image data) and pelvic floor function rehabilitation assessment. Modalities with high correlation are assigned the upper limit of the range, and modalities with low correlation are assigned the lower limit, and this setting remains constant. The modal signal quality coefficient is normalized to a range of 0 to 1 at the current moment. It is automatically calculated by the model based on the signal-to-noise ratio, integrity, and distortion of the real-time acquired signal. A value closer to 1 indicates excellent modal signal quality, while a value closer to 0 indicates severe signal distortion and a low proportion of effective features. The patient individual difference coefficient is set between 0.1 and 0.4, and is determined based on individual physiological characteristics such as patient age, disease severity, and postoperative recovery stage. As a constant parameter, it is used to personalize the offset correction of each modal weight for different patients, adapting to the differences in pelvic floor function assessment among different populations.

[0030] In one implementation, the forward reconstruction equation is constructed using the following formula:

[0031] In the formula, For the next node evaluation results, The evaluation result for the current node. Based on the evaluation results of the previous node, For time series smoothing coefficients, This is the residual correction factor. To assess the residual at the current moment, Here, t is the trend integral correction coefficient, and t is the sampling time. For integration time variable, The exponential decay coefficient is... As a local mean benchmark, for A function for continuously evaluating results at each time step.

[0032] In this embodiment, the equation dynamically combines the current node assessment result with the second-order linear extrapolation trend. It utilizes second-order differences to capture the basic linear trend of the rehabilitation process while suppressing the influence of instantaneous noise on the curve through a weighting coefficient α, achieving a balance between real-time performance and stability. This addresses the pain points of traditional methods in rehabilitation assessment, such as curve jumps or trend lags. An exponentially decaying local trend integral term is introduced to achieve refined trend capture of the rehabilitation process. Addressing the slow and gradual physiological characteristics of pelvic floor rehabilitation, the equation innovatively incorporates… On the one hand, by using integral operations to capture the cumulative trend of evaluation results over a short period of time, information loss caused by extrapolation of discrete nodes is avoided. On the other hand, by introducing an exponentially decaying weight e-λ(t-τ), a weighting effect of "stronger influence of recent trends and diminishing influence of distant trends" is achieved, which is more in line with the actual law that "recent changes have a greater impact on future states" in the rehabilitation process, significantly improving the continuity and authenticity of curve reconstruction. A dual correction framework of residual correction and trend correction is constructed to realize the error closed-loop suppression equation while introducing a residual correction term. The trend integral correction term: The residual correction term corrects the extrapolation error by using the deviation between the current evaluated value and the reconstructed value, thus preventing the error from accumulating over time; the trend integral correction term smooths out the local fluctuations caused by the residual correction by accumulating the local trend. This dual mechanism of error correction and trend constraint effectively solves the problems of error accumulation and poor robustness in traditional extrapolation algorithms.

[0033] In this embodiment, the temporal smoothing coefficient is set to a range of 0.3 to 0.7, a globally fixed preset parameter. This parameter balances the weight distribution between the current node's evaluation result and the temporal extrapolation reconstruction result. A larger value emphasizes retaining the current evaluation result, while a smaller value strengthens the smoothing constraint effect of temporal trend extrapolation. The residual correction coefficient is set to a range of 0.2 to 0.6, a globally fixed constant parameter. This parameter regulates the correction strength of the evaluation residuals on the forward reconstruction equation. A larger value results in stronger error correction, while a smaller value weakens residual disturbances, ensuring a stable reconstruction result without abrupt changes. The trend integral correction coefficient is set to a range of 0.1 to 0.5, a globally unified preset parameter. This parameter controls the influence of the local trend integral term on the next node's evaluation result. A larger value enhances the ability to capture the changing trend of the rehabilitation time series, while a smaller value reduces the interference caused by integral fluctuations. The exponential decay coefficient is set between 1.0 and 4.0 as a fixed hyperparameter. It is used to achieve the effect of high trend weight in the near time and decay of trend weight in the far time during the time series integration process. If the value is too large, the historical data in the far time will decay and weaken rapidly. If the value is too small, the influence of historical trends over a longer period will be retained, which is suitable for the slow and gradual changes in pelvic floor rehabilitation.

[0034] In this embodiment, The continuous assessment result function at any given time is the pelvic floor dysfunction rehabilitation assessment curve between the assessment result at the current node and the assessment result at the previous node.

[0035] In one embodiment of the present invention, a pelvic floor dysfunction rehabilitation assessment curve is determined based on the assessment results of the current node, the previous node, and the next node, including: The current node assessment result, the previous node assessment result, and the next node assessment result are continuously interpolated and fitted to obtain the pelvic floor dysfunction rehabilitation assessment curve. In the pelvic floor dysfunction rehabilitation assessment curve, the horizontal axis is the time axis, and the vertical axis is the node assessment result axis.

[0036] In this embodiment, three sets of time-series discretized quantitative assessment data are used as baseline nodes. A continuous interpolation fitting algorithm is employed for smoothing optimization, effectively eliminating numerical abrupt changes and local fluctuations between discrete data points, ensuring the continuity and smoothness of data changes. The constructed pelvic floor dysfunction rehabilitation assessment curve uses time as the horizontal axis to characterize the temporal changes in the rehabilitation process; and uses the quantitative node assessment results as the vertical axis to intuitively reflect the strength and weakness of the patient's pelvic floor function at different time periods, fully restoring the dynamic rehabilitation trend of the patient's pelvic floor function over time.

[0037] In one embodiment of the present invention, the final pelvic floor dysfunction rehabilitation assessment result is determined by the following formula:

[0038] In the formula, For the final pelvic floor dysfunction rehabilitation assessment results, The mean of the global integral of the rehabilitation assessment curve. For weighted time series trend features, This is the characteristic quantity of the steady-state fluctuation of the curve. As the first weighting coefficient, This is the second weighting coefficient. This is the third weighting coefficient. For the total duration of the complete evaluation cycle, For integration time variable, This is a trend amplification correction factor. This is the time-series decay coefficient. It is a continuous evaluation function at any point on the rehabilitation assessment curve.

[0039] In this embodiment, firstly, a global time-domain integral operation is introduced to continuously integrate and average the data of the entire rehabilitation assessment curve, providing a complete picture of the patient's pelvic floor function throughout the entire cycle and avoiding the shortcomings of one-sided assessment based on data from a single moment. Secondly, by combining the curve's differential rate of change and exponential decay weighting mechanism, the dynamic evolution trend of rehabilitation is extracted. Based on the differences in time sequence, the interference of long-term ineffective fluctuations is weakened, while the characteristics of recent rehabilitation improvement are strengthened, aligning with the physiological law of gradual recovery of pelvic floor function and achieving a refined quantitative representation of the rehabilitation development trend. Simultaneously, a deviation integral mean square operation is introduced to quantify the degree of fluctuation and dispersion of the rehabilitation curve, objectively reflecting the stability of pelvic floor muscle contraction control and functional status, compensating for the shortcomings of conventional assessments in quantifying physiological homeostasis. Furthermore, three evaluation dimensions—overall rehabilitation level, dynamic recovery trend, and functional stability—are combined, and weighted normalization constraints are set to ensure the objectivity and uniqueness of the assessment results. Compared to traditional static and fragmented assessment methods, this equation upgrades from discrete single-point evaluation to continuous time-domain comprehensive evaluation, taking into account static basic rehabilitation level, dynamic recovery potential and physiological stability, effectively reducing the interference of random noise and transient abnormal data on the final assessment results, and improving the accuracy of pelvic floor dysfunction rehabilitation assessment.

[0040] It should also be noted that the electrical signal data, patient movement and posture data, and pelvic floor image data collected by this invention are all transmitted via Bluetooth devices.

[0041] See appendix Figure 2 , Figure 2 This is a schematic diagram of the structure of a pelvic floor dysfunction rehabilitation assessment system according to an embodiment of this application. Figure 2 As shown, the pelvic floor dysfunction rehabilitation assessment system in this embodiment of the invention mainly includes: The acquisition unit 200 is configured to acquire electrical signal data of the pelvic floor muscles, patient movement and posture data, and pelvic floor image data of the current patient to be tested; wherein, the electrical signal data includes bioelectrical signals and pressure signals; The first data processing unit 202 is configured to input the electrical signal data, the patient's action posture data and the pelvic floor image data into the adaptive evaluation model to obtain the current node evaluation result; The second data processing unit 204 is configured to input the current node evaluation result and the previous node evaluation result into the forward reconstruction equation to obtain the next node evaluation result; The third data processing unit 206 is configured to determine a pelvic floor dysfunction rehabilitation assessment curve based on the current node assessment result, the previous node assessment result, and the next node assessment result. The fourth data processing unit 208 is configured to determine the final pelvic floor dysfunction rehabilitation assessment result based on the pelvic floor dysfunction rehabilitation assessment curve.

[0042] In one embodiment of the present invention, the adaptive evaluation model is obtained by training an improved hybrid model, which includes an input layer, a multimodal feature extraction layer, an environment adaptive allocation layer, a modality feature fusion layer, an LSTM layer, a fully connected layer, and an output layer connected in sequence. The input layer is used to receive the electrical signal data, the patient's posture data, and the pelvic floor image data. The multimodal feature extraction layer is used to extract features from the electrical signal data, the patient's posture data, and the pelvic floor image data. The environment adaptive allocation layer is used to allocate weights to the electrical signal data, the patient's posture data, and the pelvic floor image data according to the environmental data of the corresponding sensor. The modality feature fusion layer is used to fuse the features of the weighted electrical signal data, the patient's posture data, and the pelvic floor image data. The LSTM layer is used to capture the temporal correlation of the fused multimodal features and mine the temporal pattern of pelvic floor muscle contraction and relaxation. The fully connected layer is used to filter and integrate temporal features and convert them into low-dimensional evaluation parameters suitable for output. The output layer is used to output the evaluation result of the current node.

[0043] In one embodiment of the present invention, when there are signal fluctuations in the electrical signal data, the patient's posture data, and the pelvic floor image data, the environment adaptive allocation layer is used to execute the calculation logic of the following formula:

[0044]

[0045] In the formula, The comprehensive environmental interference impact factor for the i-th modal data. These are modality-specific correction coefficients. This is the environmental sensitivity adjustment coefficient. This is a weighting factor for the environmental disturbance level. These are the normalized environmental observations at the current moment. The Sigmoid smoothing constraint function is used. Here, t is the slope adjustment coefficient, and t is the sampling time. This is the time variable for integration.

[0046] In one embodiment of the present invention, when there are no signal fluctuations in the electrical signal data, the patient's posture data, and the pelvic floor image data, the environment adaptive allocation layer is used to execute the calculation logic of the following formula:

[0047] In the formula, For the dynamic weights of the i-th mode, This is the weighting balance coefficient. For modal prior association weights, The modal signal quality coefficient at the current moment. The cumulative integral mean of signal quality degradation. This represents the individual variability coefficient for patients.

[0048] In one embodiment of the present invention, the forward reconstruction equation is constructed using the following formula:

[0049] In the formula, The evaluation result for the next node, The evaluation result for the current node, The evaluation result of the previous node, For time series smoothing coefficients, This is the residual correction factor. To assess the residual at the current moment, Here, t is the trend integral correction coefficient, and t is the sampling time. For integration time variable, The exponential decay coefficient is... As a local mean benchmark, for A function for continuously evaluating results at each time step.

[0050] In one embodiment of the present invention, determining the pelvic floor dysfunction rehabilitation assessment curve based on the current node assessment result, the previous node assessment result, and the next node assessment result includes: The current node evaluation result, the previous node evaluation result, and the next node evaluation result are continuously interpolated and fitted to obtain the pelvic floor dysfunction rehabilitation evaluation curve. In this context, the horizontal axis of the pelvic floor dysfunction rehabilitation assessment curve represents the time axis, and the vertical axis represents the node assessment results axis.

[0051] In one embodiment of the present invention, the final pelvic floor dysfunction rehabilitation assessment result is determined by the following formula:

[0052] In the formula, For the final pelvic floor dysfunction rehabilitation assessment results, The mean of the global integral of the rehabilitation assessment curve. For weighted time series trend features, This is the characteristic quantity of the steady-state fluctuation of the curve. As the first weighting coefficient, This is the second weighting coefficient. This is the third weighting coefficient. For the total duration of the complete evaluation cycle, For integration time variable, This is a trend amplification correction factor. This is the time-series decay coefficient. It is a continuous evaluation function at any point on the rehabilitation assessment curve.

[0053] The steps of the various methods described above are only for clarity. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, but without changing the core design of the algorithm and process, are also within the scope of protection of this patent.

[0054] It is not difficult to see that this embodiment is a method embodiment corresponding to the system embodiment, and this embodiment can be implemented in conjunction with the system embodiment. The relevant technical details mentioned in the system embodiment are still valid in this embodiment, and will not be repeated here to reduce repetition. Correspondingly, the relevant technical details mentioned in this embodiment can also be applied to the system embodiment.

[0055] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this application, this embodiment does not introduce units that are not closely related to solving the technical problem proposed in this application; however, this does not mean that other units are absent from this embodiment.

[0056] Furthermore, some embodiments of this application also provide an electronic device. The electronic device can be various forms of digital computer, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, etc. The electronic device can also be various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices.

[0057] The electronic device includes: one or more processors; and a memory storing computer program instructions, which, when executed, cause the processor to perform the steps of the methods provided in any one or more of the above embodiments. Figure 3As shown, the electronic device includes one or more processors 1101, a memory 1102, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components are interconnected via different buses and can be mounted on a common motherboard or otherwise as required. The processors can process instructions executed within the electronic device, including instructions stored in or on memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some other embodiments, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple electronic devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). The components, their connections and relationships, and their functions shown herein are merely examples and are not intended to limit the implementation of the present application described and / or claimed herein.

[0058] The electronic device may further include an input device 1103 and an output device 1104. The processor 1101, memory 1102, input device 1103, and output device 1104 may be connected via a bus or other means. Figure 4 Taking the example of a connection between China and Israel via a bus.

[0059] Input device 1103 can receive input numerical or character information, and generate key signal inputs related to user settings and function control of the electronic device, such as a touch screen, keypad, mouse, trackpad, touchpad, joystick, one or more mouse buttons, trackball, joystick, etc. Output device 1104 may include a display device, auxiliary lighting device (e.g., LED), and haptic feedback device (e.g., vibration motor). The display device may include, but is not limited to, a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0060] To provide interaction with the user, the electronic device can be a computer. The computer has: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0061] In this embodiment, a computer-readable medium stores a computer program / instructions that, when executed by a processor, implement the steps of the methods provided in any one or more of the above embodiments. This computer-readable medium may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into that device. The aforementioned computer-readable medium carries one or more computer-readable instructions.

[0062] The memory 1102 can serve as a non-transitory computer-readable storage medium, used to store non-transitory software programs, non-transitory computer-executable programs, and modules. The processor 1101 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions, and modules stored in the memory 1102, thereby implementing the program instructions / modules corresponding to the methods provided in any one or more of the embodiments described above in this application.

[0063] The memory 1102 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device. Furthermore, the memory 1102 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 1102 may optionally include memory remotely located relative to the processor 1101, and these remote memories can be connected to the electronic device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0064] It should be noted that the computer-readable medium described in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0065] Computer-readable media include permanent and non-permanent, removable and non-removable media, which can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, read-only optical disc (CD-ROM), digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0066] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0067] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. For example, it can be implemented using an application-specific integrated circuit (ASIC), a general-purpose computer, or any other similar hardware device. In some embodiments, the software program of this application can be executed by a processor to implement the steps or functions described above. Similarly, the software program of this application (including related data structures) can be stored in a computer-readable recording medium, such as RAM memory, magnetic or optical drives, floppy disks, or similar devices. Furthermore, some steps or functions of this application can be implemented in hardware, for example, as circuitry that works with a processor to perform the various steps or functions.

[0068] The computer program product provided in this application includes one or more computer programs / instructions. When executed by a processor, these computer programs / instructions generate, in whole or in part, the processes or functions described in this application. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0069] The flowcharts or block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of devices, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-specific system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0070] The scope of this application is defined by the appended claims rather than the foregoing description, and is therefore intended to encompass all variations falling within the meaning and scope of equivalents of the claims. No reference numerals in the claims should be construed as limiting the scope of the claims. Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices recited in a device claim may also be implemented by a single unit or device in software or hardware. Terms such as "first," "second," etc., are used only for distinguishing descriptions and do not indicate any particular order, nor should they be construed as indicating or implying relative importance.

[0071] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily made by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims, and the above embodiments should be regarded as exemplary and non-limiting.

Claims

1. A method for assessing pelvic floor dysfunction rehabilitation, characterized in that, The method includes: The system collects electrical signal data of the pelvic floor muscles, patient movement and posture data, and pelvic floor image data of the patient currently being tested; wherein, the electrical signal data includes bioelectrical signals and pressure signals; The electrical signal data, the patient's movement and posture data, and the pelvic floor image data are input into the adaptive evaluation model to obtain the evaluation result of the current node; The current node evaluation result and the previous node evaluation result are input into the forward reconstruction equation to obtain the next node evaluation result; Based on the current node assessment result, the previous node assessment result, and the next node assessment result, a pelvic floor dysfunction rehabilitation assessment curve is determined. Based on the pelvic floor dysfunction rehabilitation assessment curve, the final pelvic floor dysfunction rehabilitation assessment result is determined.

2. The method according to claim 1, characterized in that, The adaptive evaluation model is obtained by training an improved hybrid model, which includes an input layer, a multimodal feature extraction layer, an environment adaptive allocation layer, a modality feature fusion layer, an LSTM layer, a fully connected layer, and an output layer connected in sequence. The input layer receives electrical signal data, patient posture data, and pelvic floor image data. The multimodal feature extraction layer extracts features from these data. The environment adaptive allocation layer assigns weights to the electrical signal data, patient posture data, and pelvic floor image data based on the environmental data of the corresponding sensors. The modality feature fusion layer fuses the weighted features of the electrical signal data, patient posture data, and pelvic floor image data. The LSTM layer captures the temporal correlation of the fused multimodal features and uncovers the temporal patterns of pelvic floor muscle contraction and relaxation. The fully connected layer filters and integrates temporal features, converting them into low-dimensional evaluation parameters suitable for the output. The output layer outputs the evaluation result for the current node.

3. The method according to claim 2, characterized in that, When there are signal fluctuations in the electrical signal data, the patient's posture data, and the pelvic floor image data, the environment adaptive allocation layer performs the following calculation logic: In the formula, The comprehensive environmental interference impact factor for the i-th modal data. These are modality-specific correction coefficients. This is the environmental sensitivity adjustment coefficient. This is a weighting factor for the environmental disturbance level. These are the normalized environmental observations at the current moment. The Sigmoid smoothing constraint function is used. Here, t is the slope adjustment coefficient, and t is the sampling time. This is the time variable for integration.

4. The method according to claim 3, characterized in that, When there are no signal fluctuations in the electrical signal data, the patient's posture data, and the pelvic floor image data, the environment adaptive allocation layer performs the following calculation logic: In the formula, For the dynamic weights of the i-th mode, This is the weighting balance coefficient. For modal prior association weights, The modal signal quality coefficient at the current moment. The cumulative integral mean of signal quality degradation. This represents the individual variability coefficient for patients.

5. The method according to claim 1, characterized in that, The positive reconstruction equation is constructed using the following formula: In the formula, The evaluation result for the next node, The evaluation result for the current node, The evaluation result of the previous node, For time series smoothing coefficients, This is the residual correction factor. To assess the residual at the current moment, Here, t is the trend integral correction coefficient, and t is the sampling time. For integration time variable, The exponential decay coefficient is... As a local mean benchmark, for A function for continuously evaluating results at each time step.

6. The method according to claim 1, characterized in that, The process of determining the pelvic floor dysfunction rehabilitation assessment curve based on the current node assessment result, the previous node assessment result, and the next node assessment result includes: The current node evaluation result, the previous node evaluation result, and the next node evaluation result are continuously interpolated and fitted to obtain the pelvic floor dysfunction rehabilitation evaluation curve. In this context, the horizontal axis of the pelvic floor dysfunction rehabilitation assessment curve represents the time axis, and the vertical axis represents the node assessment results axis.

7. The method according to claim 5, characterized in that, The final pelvic floor dysfunction rehabilitation assessment result is determined by the following formula: In the formula, For the final pelvic floor dysfunction rehabilitation assessment results, The mean of the global integral of the rehabilitation assessment curve. For weighted time series trend features, This is the characteristic quantity of the steady-state fluctuation of the curve. As the first weighting coefficient, This is the second weighting coefficient. This is the third weighting coefficient. For the total duration of the complete evaluation cycle, For integration time variable, This is a trend amplification correction factor. This is the time-series decay coefficient. It is a continuous evaluation function at any point on the rehabilitation assessment curve.

8. A pelvic floor dysfunction rehabilitation assessment system, characterized in that, The system, applicable to the method of any one of claims 1-7, comprises: The acquisition unit is configured to acquire electrical signal data of the pelvic floor muscles, patient movement and posture data, and pelvic floor image data of the current patient to be tested; wherein the electrical signal data includes bioelectrical signals and pressure signals; The first data processing unit is configured to input the electrical signal data, the patient's movement and posture data, and the pelvic floor image data into the adaptive evaluation model to obtain the current node evaluation result; The second data processing unit is configured to input the current node evaluation result and the previous node evaluation result into the forward reconstruction equation to obtain the next node evaluation result; The third data processing unit is configured to determine the pelvic floor dysfunction rehabilitation assessment curve based on the current node assessment result, the previous node assessment result, and the next node assessment result. The fourth data processing unit is configured to determine the final pelvic floor dysfunction rehabilitation assessment result based on the pelvic floor dysfunction rehabilitation assessment curve.

9. An electronic device, characterized in that, The device includes: One or more processors; and A memory storing computer program instructions, which, when executed, cause the processor to perform the method of any one of claims 1-7.

10. A computer-readable medium having stored thereon computer program instructions that can be executed by a processor to implement the method as described in any one of claims 1-7.

Images