Pelvic floor muscle repair instrument, pelvic floor electromyographic signal interference removal method and related equipment

Abstract

The invention relates to the technical field of postpartum rehabilitation equipment, and provides a pelvic floor muscle repairing instrument, a pelvic floor electromyographic signal interference removing method and related equipment. The pelvic floor muscle electromyographic acquisition module is configured to acquire pelvic floor muscle surface electromyographic signals containing adjacent muscle group interference to obtain target electromyographic signals; the interference muscle group monitoring module is configured to collect physiological activity signals of muscle groups adjacent to pelvic floor muscles so as to obtain interference reference signals; the posture sensing module is configured to collect a body posture time sequence signal related to pelvic floor muscle activity so as to obtain a scene auxiliary signal; and the signal processing module is configured to perform suppression processing on adjacent muscle group interference in the target electromyographic signal based on the interference reference signal and the scene auxiliary signal, and output a purified pelvic floor electromyographic signal. Pelvic floor muscle real signals and dynamic interference are distinguished through the scene auxiliary signals obtained by the posture sensing module, and the purity and reliability of output signals of the pelvic floor muscle repairing instrument are improved.

Description

Technical Field

[0001] This application relates to the field of postpartum rehabilitation equipment technology, specifically to a pelvic floor muscle repair device, a method for removing pelvic floor muscle electrical signal interference, and related equipment. Background Technology

[0002] In the field of medical rehabilitation equipment technology, pelvic floor muscle electromyography (sEMG) has important application value in areas such as female pelvic floor function assessment. However, its detection process is easily interfered with by the contraction activity of adjacent muscle groups such as the rectus abdominis. Especially when an individual is actively contracting or exercising, the electrical signal of the rectus abdominis muscle will be mixed with the pelvic floor muscle electromyography signal, resulting in the distortion of the collected pelvic floor muscle signal and seriously affecting the accurate analysis of pelvic floor function in the future.

[0003] However, current pelvic floor electromyography (EMG) signal processing techniques have several shortcomings. Firstly, the pelvic floor muscles are anatomically very close to adjacent muscle groups such as the rectus abdominis and quadratus lumborum. Pelvic floor muscle activity easily triggers coordinated contractions of these adjacent muscles, resulting in a natural mixture of electrical activity from neighboring muscles into the acquired EMG signals, creating a mixed signal. Current techniques cannot distinguish this mixture, lacking clear criteria for locating interfering components. Secondly, pelvic floor muscle activity is closely related to body posture. Changes in posture alter the intensity and signal characteristics of adjacent muscle groups, causing interference to exhibit dynamic fluctuations. Current techniques focus solely on acquiring the EMG signal itself, lacking contextual criteria for interference assessment. Ultimately, this results in processed EMG signals either containing residual interference or having inadvertently deleted effective components, leading to insufficient signal reliability and severely hindering the precision development of pelvic floor function assessment and rehabilitation. Therefore, current techniques suffer from poor interference suppression, resulting in insufficient purity and reliability of the output signal, making it difficult to meet the clinical needs of pelvic floor function assessment.

[0004] The preceding description is intended to provide general background information and does not necessarily constitute prior art. Summary of the Invention

[0005] This application provides a pelvic floor muscle repair device, a method for removing interference from pelvic floor muscle electromyography signals, and related equipment. By using scene auxiliary signals obtained through a posture perception module, the device distinguishes between real pelvic floor muscle signals and dynamic interference, thereby improving the purity and reliability of the output signal of the pelvic floor muscle repair device and providing high-quality data support for accurate assessment and rehabilitation of pelvic floor function.

[0006] In a first aspect, this application provides a pelvic floor muscle repair device, the pelvic floor muscle repair device comprising:

[0007] case;

[0008] A pelvic floor muscle electromyography (EMG) acquisition module is connected to the housing and configured to acquire EMG signals from the surface of the pelvic floor muscles, including interference from adjacent muscle groups, in order to obtain the target EMG signal.

[0009] The interference muscle group monitoring module is configured to collect physiological activity signals of adjacent muscle groups of the pelvic floor muscles in order to obtain interference reference signals.

[0010] An attitude perception module, connected to the housing, is configured to collect temporal signals of body posture related to pelvic floor muscle activity in order to obtain scene auxiliary signals.

[0011] In addition, the signal processing module is electrically connected to the pelvic floor muscle electromyography acquisition module, the interfering muscle group monitoring module, and the posture perception module; the signal processing module is configured to: based on the interference reference signal and the scene auxiliary signal, suppress the interference of adjacent muscle groups in the target electromyography signal, and output the purified pelvic floor electromyography signal.

[0012] Secondly, embodiments of this application provide a method for removing interference from pelvic floor electromyographic signals, including:

[0013] Electromyographic signals of the pelvic floor muscles, including those affected by adjacent muscle groups, were collected to obtain the target electromyographic signal.

[0014] Physiological activity signals of adjacent muscle groups of the pelvic floor muscles were collected to obtain interference reference signals;

[0015] Collect temporal signals of body posture related to pelvic floor muscle activity to obtain scene auxiliary signals;

[0016] Based on the interference reference signal and the scene auxiliary signal, the interference from adjacent muscle groups in the target electromyographic signal is suppressed, and the purified pelvic floor electromyographic signal is output.

[0017] Thirdly, embodiments of this application provide a device for removing interference from pelvic floor electromyography signals, including:

[0018] The first acquisition module is used to acquire electromyographic signals on the surface of the pelvic floor muscles, which are subject to interference from adjacent muscle groups, in order to obtain the target electromyographic signal.

[0019] The second acquisition module is used to acquire physiological activity signals of the muscle groups adjacent to the pelvic floor muscles in order to obtain interference reference signals.

[0020] The third acquisition module is used to acquire body posture time-series signals related to pelvic floor muscle activity in order to obtain scene auxiliary signals.

[0021] The signal purification module is used to suppress interference from adjacent muscle groups in the target electromyographic signal based on the interference reference signal and the scene auxiliary signal, and output the purified pelvic floor electromyographic signal.

[0022] In a fourth aspect, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the pelvic floor muscle electrical signal interference removal method according to the second aspect when executing the computer program.

[0023] In a fifth aspect, a storage medium is provided, which stores a computer program capable of being loaded and executed by a processor to implement the pelvic floor muscle electrical signal interference removal method according to the second aspect.

[0024] The present application provides a pelvic floor muscle repair instrument, a pelvic floor muscle electrical signal interference removal method, a device, an apparatus, and a storage medium. The pelvic floor muscle repair instrument includes a shell, a pelvic floor muscle electromyography acquisition module connected to the shell and configured to acquire a pelvic floor muscle surface electromyography signal containing adjacent muscle group interference to obtain a target electromyography signal, an interference muscle group monitoring module configured to acquire a physiological activity signal of the adjacent muscle group of the pelvic floor muscle to obtain an interference reference signal, a posture sensing module connected to the shell and configured to acquire a body posture time sequence signal related to the activity of the pelvic floor muscle to obtain a scene auxiliary signal, and a signal processing module electrically connected to the pelvic floor muscle electromyography acquisition module, the interference muscle group monitoring module, and the posture sensing module. The signal processing module is configured to suppress the adjacent muscle group interference in the target electromyography signal based on the interference reference signal and the scene auxiliary signal, and output a purified pelvic floor muscle electrical signal. The pelvic floor muscle repair instrument provided by the present application synchronously acquires the interference reference signal through the pelvic floor muscle electromyography acquisition module and the interference muscle group monitoring module, provides a clear feature reference for the adjacent muscle group interference in the target electromyography signal, accurately locates the interference component, and avoids the blindness of interference differentiation caused by the lack of reference in traditional technologies. Then, the body posture time sequence signal related to the activity of the pelvic floor muscle is acquired through the posture sensing module to obtain the scene auxiliary signal, the dynamic law of the interference changing with the posture is captured, a scene-based judgment basis is provided for the interference suppression, and the interference fluctuation characteristics in different body positions are adapted. Finally, the signal processing module jointly suppresses the interference based on the interference reference signal and the scene auxiliary signal, completely removes the adjacent muscle group interference in the target electromyography signal, maximally retains the physiological characteristics of the effective signal of the pelvic floor muscle, and thus improves the purity and reliability of the output signal. It can be seen that the scene auxiliary signal obtained by the posture sensing module distinguishes the real signal of the pelvic floor muscle from the dynamic interference, accurately removes the adjacent muscle group interference in the target electromyography signal, completely retains the effective signal characteristics of the pelvic floor muscle, improves the purity and reliability of the output signal, provides high-quality data support for accurate evaluation and rehabilitation diagnosis and treatment of the pelvic function, meets the clinical needs of the pelvic function evaluation, and the like. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description only represent some of the embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

[0026] Figure 1 is a structural schematic diagram of a pelvic floor muscle repair instrument provided by the embodiments of the present application;

[0027] Figure 2 is an application environment diagram of a pelvic floor muscle electrical signal interference removal method provided by the embodiments of the present application;

[0028] Figure 3 is a flowchart of a pelvic floor muscle electrical signal interference removal method provided by the embodiments of the present application;

[0029] Figure 4 is another flowchart of a pelvic floor muscle electrical signal interference removal method provided by the embodiments of the present application;

[0030] Figure 5 is a flowchart of feature fusion provided by the embodiments of the present application;

[0031] Figure 6 is a flowchart of signal purification provided by the embodiments of the present application;

[0032] Figure 7 is a structural schematic diagram of a pelvic floor muscle electrical signal interference removal device provided by the embodiments of the present application;

[0033] Figure 8 is a structural schematic diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0034] The exemplary embodiments will be described in detail herein with reference to the drawings. Unless otherwise specified, the same numbers in different drawings indicate the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all the embodiments consistent with the present application. Rather, they are merely examples of the systems and methods consistent with some aspects of the present application, as detailed in the appended claims, or as otherwise described in this application.

[0035] It should be noted that, in this document, the terms "comprise", "comprising", or any other variant thereof are intended to cover a non-exclusive inclusion, such that processes, methods, articles, or apparatuses that comprise a list of elements are not limited to those elements, but can also include other elements not expressly listed, or inherent to such processes, methods, articles, or apparatuses. Without more limitations, an element defined by the statement "comprising a" does not exclude the existence of additional identical elements in the process, method, article, or apparatus that includes the element. In addition, components, features, elements with the same name in different embodiments of the present application can have the same meaning or different meanings, and the specific meaning thereof should be determined in combination with the explanation thereof in the specific embodiment or further in combination with the context in the specific embodiment.

[0036] It should be understood that the specific embodiments described herein are merely intended to explain the present application and are not intended to limit the present application.

[0037] In the following description, the suffixes used for elements such as "module", "part", or "unit" are used only to facilitate explanation of the present application, and have no specific meaning by themselves. Therefore, "module", "part", or "unit" can be used interchangeably.

[0038] In order to solve the above technical problems and break through the defects of the prior art, the embodiments of the present application provide a pelvic floor muscle repair instrument, a pelvic floor muscle electrical signal interference removal method and related equipment, which can accurately distinguish between real pelvic floor muscle signals and dynamic interference, improve the purity and reliability of the output signal, and provide high-quality data support for accurate evaluation and rehabilitation diagnosis and treatment of pelvic function.

[0039] As shown in Figure 1 The pelvic floor muscle repair instrument provided by the embodiments of the present application comprises:

[0040] a shell 10;

[0041] Specifically, the shell 10 serves as the physical bearing structure of the pelvic floor muscle repair instrument, providing a mounting and fixing basis for other functional modules, ensuring that the modules maintain a relatively stable positional relationship in a clinical use scenario (such as when the patient's body position is changed or moved), and guaranteeing the continuity and reliability of signal acquisition. The material of the shell includes, but is not limited to, one of medical-grade ABS plastic, medical-grade polycarbonate (PC), and medical-grade silicone composite material. Medical-grade ABS plastic has good biocompatibility and no sensitization, and can be directly contacted with the human skin. At the same time, it has high mechanical strength, can stably fix various functional modules, and has a smooth surface that is easy to clean and disinfect, which is suitable for the frequent cleaning requirements in the medical field and is a commonly used material for small and medium-sized medical device shells. Medical-grade polycarbonate (PC) has good light transmission and excellent impact resistance, and can withstand a wide range of high and low temperatures, which can adapt to temperature changes in different clinical environments. Moreover, it also meets the medical-grade biocompatibility standards and is suitable for devices with high structural strength and environmental adaptability requirements. If some areas of the shell need to be attached to the human body surface (such as in a wearable design scenario), a composite structure of silicone and hard plastic can be used. The edge area attached to the human body is made of medical silicone (soft, skin-friendly, and slip-resistant), and the main bearing part is made of ABS or PC plastic (to ensure structural stability), which balances comfort and functionality.

[0042] The pelvic floor muscle electromyography acquisition module 20 is connected with the shell and is configured to acquire a pelvic floor muscle surface electromyography signal containing interference of adjacent muscle groups to obtain a target electromyography signal.

[0043] Specifically, the pelvic floor muscle electromyography acquisition module is connected with the shell, and the core function is to acquire a pelvic floor muscle surface electromyography signal, which naturally contains interference components of adjacent muscle groups (such as the rectus abdominis muscle and the lumbar quadratus muscle), and finally outputs a target electromyography signal. As the basis for the device to obtain core physiological indicators, the signal acquired by the module is directly related to the core data of the pelvic floor muscle function evaluation.

[0044] The interference muscle group monitoring module 30 is configured to acquire a physiological activity signal of the adjacent muscle groups of the pelvic floor muscle to obtain an interference reference signal.

[0045] Specifically, the interference muscle group monitoring module 30 is responsible for the acquisition task of the physiological activity signal of the adjacent muscle groups of the pelvic floor muscle, and outputs an interference reference signal. Its role is to actively capture the signal characteristics of the interference source, and to provide a direct reference benchmark for subsequent differentiation between the true signal of the pelvic floor muscle and the interference signal.

[0046] The posture sensing module 40 is connected with the shell and is configured to acquire a body posture time sequence signal related to the activity of the pelvic floor muscle to obtain a scene auxiliary signal.

[0047] Specifically, the posture perception module 40, connected to the housing, is responsible for collecting temporal signals of body posture related to pelvic floor muscle activity and outputting scene auxiliary signals. This module, by sensing changes in body posture, provides data support for determining the context of interference, adapting to dynamically changing acquisition scenarios in clinical settings.

[0048] In addition, the signal processing module 50 is electrically connected to the pelvic floor muscle electromyography acquisition module, the interfering muscle group monitoring module, and the posture perception module; the signal processing module is configured to: based on the interference reference signal and the scene auxiliary signal, suppress the interference of adjacent muscle groups in the target electromyography signal, and output the purified pelvic floor electromyography signal.

[0049] Specifically, the signal processing module 50, as the core control and processing unit of the device, is electrically connected to all three acquisition modules mentioned above, and can receive and process the signals output by each module. Its core configuration is as follows: based on the interference reference signal and the scene auxiliary signal, it suppresses the interference of adjacent muscle groups in the target electromyographic signal, and finally outputs the purified pelvic floor electromyographic signal.

[0050] In the pelvic floor muscle repair device of this embodiment, each hardware module and the signal processing module form a closed-loop collaboration: the pelvic floor muscle electromyography acquisition module acquires the core target signal containing interference, the interference muscle group monitoring module simultaneously captures the characteristics of the interference source, and the posture perception module provides scene context information; after the three transmit the signal to the signal processing module, the processing module uses the interference reference signal (locating interference characteristics) and the scene auxiliary signal (associating with the interference scene) as dual criteria to specifically suppress the interference components in the target electromyography signal, thereby achieving signal purification.

[0051] Taking the postpartum female pelvic floor function assessment scenario as an example, the shell of the pelvic floor muscle repair device is fixed next to the assessment bed or on a wearable carrier for the patient. The electrodes of the pelvic floor muscle electromyography (EMG) acquisition module are attached to the area below the pubic symphysis (the surface projection area of ​​the pelvic floor muscles), the electrodes of the interference muscle group monitoring module are attached to the surface area of ​​the rectus abdominis muscle, and the posture sensing module is fixed at the midpoint of the line connecting the anterior superior iliac spines of the patient's pelvis. When the patient performs pelvic floor muscle contraction movements as instructed by the doctor, the pelvic floor muscle EMG acquisition module acquires pelvic floor EMG signals containing interference from the rectus abdominis muscle (target EMG signal); simultaneously, the interference muscle group monitoring module acquires the EMG signal generated by the contraction of the rectus abdominis muscle (interference reference signal); the posture sensing module synchronously acquires the changes in the pitch angle of the pelvis and the timing signals (scenario auxiliary signals) during the patient's contraction. After receiving three types of signals, the signal processing module identifies the characteristics of rectus abdominis interference (such as signal amplitude and frequency distribution) based on the interference reference signal. It combines the scene auxiliary signal to determine whether the current pitch angle change corresponds to active contraction of the rectus abdominis muscle. By combining the two, it accurately suppresses the rectus abdominis muscle interference in the target electromyographic signal and finally outputs the purified pelvic floor electromyographic signal, providing reliable data for assessing the patient's pelvic floor muscle strength level.

[0052] The embodiment can synchronously collect the pelvic floor muscle signal containing interference, the interference signal of adjacent muscle groups and the body posture signal through the integration of the pelvic floor muscle electromyographic acquisition module, the interference muscle group monitoring module, the posture sensing module and the signal processing module, and extract and fuse the multi-dimensional features through the signal processing module, so that the adjacent muscle group interference is accurately suppressed, the purified pelvic floor muscle electrical signal is output, and the signal quality is significantly improved to provide reliable data for pelvic floor function evaluation.

[0053] Further, in some embodiments, the signal processing module is configured to: extract multi-dimensional features based on the target muscle electrical signal, the interference reference signal and the scene auxiliary signal, the multi-dimensional features including time domain features, frequency domain features and high-order statistical features extracted from the target muscle electrical signal and the interference reference signal, and posture features extracted from the scene auxiliary signal; perform feature fusion on the target muscle electrical signal, the interference reference signal, the scene auxiliary signal and the multi-dimensional features to generate integrated features; and perform suppression processing on the adjacent muscle group interference in the target muscle electrical signal based on the integrated features, and output the purified pelvic floor muscle electrical signal.

[0054] Specifically, the signal processing module is the core operation unit of the device, and is electrically connected with the pelvic floor muscle electromyographic acquisition module, the interference muscle group monitoring module and the posture sensing module, receives the target muscle electrical signal, the interference reference signal and the scene auxiliary signal output by the three types of modules, and completes signal processing. The specific signal processing process is as follows: based on the received target muscle electrical signal, interference reference signal and scene auxiliary signal, multi-dimensional features covering physiological signal features and scene signal features are actively extracted. Among them, time domain features, frequency domain features and high-order statistical features are extracted from the target muscle electrical signal and the interference reference signal; posture features are extracted from the scene auxiliary signal. The three types of original collected signals (target muscle electrical signal, interference reference signal and scene auxiliary signal) are fused with the extracted multi-dimensional features to generate integrated features. This step is not a simple signal superposition, but through the operation logic of the device layer, the real-time characteristics of the original signal are associated and integrated with the artificially extracted key features to form a comprehensive data carrier that can fully represent. Based on the generated integrated features, the adjacent muscle group interference in the target muscle electrical signal is suppressed, and the purified pelvic floor muscle electrical signal is finally output. The pelvic floor muscle electromyographic acquisition module, the interference muscle group monitoring module and the posture sensing module are responsible for the synchronous collection of multi-source signals to provide original data input for the signal processing module; the signal processing module converts the original signal into integrated features with strong discrimination through internal refined operation logic, and then completes interference removal based on the integrated features. In the whole process, the device layer closed loop formed by the collection module and the processing module ensures the pertinence and reliability of signal processing.

[0055] Take the abdominal rectus muscle interference suppression scenario in the postpartum female pelvic floor muscle rehabilitation evaluation as an example, the working process of the pelvic floor muscle repair instrument: according to the clinical operation specification, the electrodes of the pelvic floor muscle electromyography acquisition module are attached to the perineal region of the patient (the surface projection area of the pelvic floor muscle), the electrodes of the interference muscle group monitoring module are attached to the surface of the middle and lower segments of the abdominal rectus muscle, and the posture sensing module is fixed to the pelvic bone region. Each module is electrically connected to the signal processing module (circuit board integrated in the device shell) through a wire; the patient completes the "5-second sustained pelvic floor muscle contraction" action as instructed, at which time: the pelvic floor muscle electromyography acquisition module collects pelvic floor muscle signals containing abdominal rectus muscle interference (target electromyography signals, such as time-domain signals with amplitude fluctuations of 0.1-0.5 mV); the interference muscle group monitoring module collects electromyography signals generated by the contraction of the abdominal rectus muscle (interference reference signals, such as frequency-domain signals with a dominant frequency of 20-30 Hz); the posture sensing module collects the posture timing signal of the patient's pelvic bone slightly leaning forward (scenario auxiliary signal, such as timing data with a pitch angle change of 5°-10°). Through the operation circuit of the signal processing module, time-domain features, frequency-domain features, and high-order statistical features are extracted from the target electromyography signals and the interference reference signals; posture features are extracted from the scenario auxiliary signal to form a multi-dimensional feature set. The signal processing module fuses the original three types of signals with the extracted multi-dimensional features to generate integrated features, for example, the 35-45 Hz dominant frequency feature of the target electromyography signal, the 20-30 Hz dominant frequency feature of the interference reference signal, and the 0.5° / s pitch angle change rate are associated and integrated to form comprehensive data that can clearly distinguish between pelvic floor muscle contraction signals and abdominal rectus muscle interference signals. Based on the integrated features, the signal processing module identifies the component with a dominant frequency of 20-30 Hz in the target electromyography signal that fluctuates synchronously with the pitch angle change rate as abdominal rectus muscle interference. Through device-level filtering and signal subtraction operations, the interference component is removed, and the purified pelvic floor muscle electromyography signal (such as a time-domain signal with a stable amplitude and a dominant frequency concentrated in 35-45 Hz) is finally output. The device transmits the purified pelvic floor muscle electromyography signal to the display terminal through the output interface of the signal processing module, which can be directly used by the clinician to evaluate the patient's pelvic floor muscle contraction strength and coordination, and provide accurate data support for developing rehabilitation programs.

[0056] The present embodiment extracts time-domain, frequency-domain, high-order statistical features, and posture features, and fuses the original signals with multi-dimensional features to generate integrated features, thereby more accurately suppressing the interference of adjacent muscle groups in the pelvic floor muscle electromyography signal, and further improving the accuracy and reliability of signal processing.

[0057] Further, in some embodiments, the pelvic floor muscle electromyography acquisition module includes at least one pair of surface electromyography electrodes attached to the surface projection area of the pelvic floor muscle, which includes the area below the pubic symphysis, the perineal region, or the sacrococcygeal region.

[0058] Specifically, the pelvic floor muscle electromyography collection module in the embodiment includes at least one pair of surface electromyography electrodes. A pair of electrodes is a basic hardware configuration for collecting electromyography signals. Usually, it includes a positive electrode and a negative electrode. The weak electrical signals generated by the contraction of the pelvic floor muscle are captured through the potential difference between the two electrodes, avoiding the problem that a single electrode cannot form a potential difference and cannot effectively collect signals. The number of electrodes can be increased according to clinical needs (such as multi-channel collection and improvement of signal redundancy) to adapt to the collection needs in different scenarios. Through the close contact between the electrode and the skin, the impedance between the skin and the electrode is reduced, and the attenuation and noise introduction in the signal transmission process are reduced. The hardware installation position of the electrode ranges from the surface projection area of the pelvic floor muscle, avoiding the problem that the incorrect position of the electrode (such as being attached to the surface of the non-pelvic floor muscle) leads to the collection of invalid signals or too high a proportion of interference signals. In addition, the surface projection area of the pelvic floor muscle includes the area below the pubic symphysis, the perineum, or the sacrococcygeal bone. These three areas are the accurate projection positions of the pelvic floor muscle on the surface verified by anatomy in clinical practice, providing an operable hardware positioning standard for electrode installation, ensuring that different operators can attach the electrodes to the area that can effectively collect the pelvic floor muscle electromyography signals when using the device, and guaranteeing the consistency and reliability of the device collection.

[0059] For example, the operator takes out the pelvic floor muscle repair instrument, confirms that a pair of surface electromyography electrodes (positive electrode and negative electrode) of the pelvic floor muscle electromyography collection module are not damaged, and the conductive gel on the surface of the electrode is in a normal state. According to the body type of the patient and the clinical collection standard, the perineum is selected as the electrode attachment area (for postpartum women, this area is convenient to operate and can effectively collect the electrical signals of the levator ani muscle in the pelvic floor muscle). The operator helps the patient to take a supine position, cleans the skin around the perineum, and then attaches the positive electrode to the left side of the perineal midline by 2 cm and the negative electrode to the right side of the midline by 2 cm, ensuring that the two electrodes are closely attached to the skin (without bubbles or deviation). Then the device is started, and the pelvic floor muscle electromyography collection module starts to capture the electrical signals generated by the contraction of the pelvic floor muscle of the patient through the potential difference between the two electrodes. At this time, because the abdominal rectus muscle of the patient has not completely recovered after childbirth, the collected electrical signals naturally contain the interference components generated by the slight contraction of the abdominal rectus muscle, forming the target electromyography signals meeting the requirements.

[0060] In the embodiment, the surface electromyography electrodes are attached to the surface projection area of the pelvic floor muscle to collect signals, ensuring that the surface electromyography signals of the pelvic floor muscle can be accurately obtained, and providing high-quality raw data for subsequent signal processing and interference suppression.

[0061] Further, in some embodiments, the interference muscle group monitoring module includes at least one pair of surface electromyography electrodes, the surface electromyography electrodes are attached to the surface area of the abdominal rectus muscle and other muscle groups around the pelvic floor muscle, and the interference reference signal includes the surface electromyography signal of the abdominal rectus muscle.

[0062] Specifically, the interfering muscle group monitoring module includes at least one pair of surface electromyography electrodes, a basic configuration of a pair of electrodes capturing muscle group electrical signals, through the potential difference between the two electrodes, the weak electromyography signals generated by the contraction of the adjacent muscle group can be effectively collected, avoiding the problem that a single electrode cannot form a potential difference and cannot accurately capture the interference signal; for multiple types of adjacent interfering muscle groups such as rectus abdominis or quadratus lumborum, the number of electrodes can be increased to realize synchronous monitoring of multiple interference sources, and adapt to the differences in the types of interfering muscle groups of different patients in clinical practice. Through the close contact of the electrodes with the surface of the interfering muscle group, the attenuation of the signal by the skin impedance is reduced, and it is ensured that the collected interference signal truly reflects the activity state of the muscle group; the hardware coverage range of the interference monitoring includes the rectus abdominis and other muscle groups around it, including the most common rectus abdominis interference source in clinical practice, as well as other adjacent muscle groups such as quadratus lumborum and iliolumbar muscle that may generate interference, avoiding the lack of interference reference due to the omission of the monitoring range, and ensuring that the interference reference signal collected by the device can accurately match the most common interference scene in clinical practice, and provide targeted reference data for subsequent interference suppression.

[0063] For example, the operator takes out the pelvic floor muscle repair instrument, checks that the pair of surface electromyography electrodes (positive and negative electrodes) of the interfering muscle group monitoring module have normal conductivity, and that the electrode sheets are not aged; according to the clinical assessment requirements, it is determined to take the rectus abdominis as the core interference monitoring object, and the lower segment of the rectus abdominis (the region 2-4 cm below the navel) is selected as the electrode attachment position, which is the part of the rectus abdominis of postpartum women that is most likely to be activated in linkage with the contraction of the pelvic floor muscles, and the interference signal is the most typical. Assist the patient to take a supine position, clean the skin on the surface of the lower segment of the rectus abdominis, and then attach the positive electrode of the interfering muscle group monitoring module to the left edge of the rectus abdominis (3 cm away from the midline), and the negative electrode to the right edge (symmetrical position), ensuring that the electrodes are tightly attached and not displaced; then start the pelvic floor muscle repair instrument, and when the patient contracts the pelvic floor muscles as instructed, the rectus abdominis generates electromyography signals due to the linkage contraction caused by postpartum separation, and the interfering muscle group monitoring module captures the signals through the potential difference between the two electrodes to generate surface electromyography signals of the rectus abdominis as interference reference signals.

[0064] The present embodiment collects interference reference signals by attaching surface electromyography electrodes to the surface area of the rectus abdominis and other muscle groups around the pelvic floor muscles, and can directly obtain the physiological activity signals of the interference source, providing a direct basis for accurately identifying and suppressing interference.

[0065] Further, in some embodiments, the posture sensing module includes at least one of an inertial measurement unit, an optical posture sensor, or a magnetic sensing posture module, the posture sensing module is fixed to the pelvic region, the scene auxiliary signal includes acceleration data and angular velocity data, and the pitch angle time sequence signal and the roll angle time sequence signal obtained according to the acceleration data and the angular velocity data.

[0066] Specifically, the posture perception module includes at least one of an inertial measurement unit, an optical posture sensor, or a magnetic sensing posture module, covering the current mainstream posture acquisition technology (such as an inertial measurement unit IMU, which can directly collect acceleration and angular velocity data) and compatible with high-precision posture sensing schemes that may be applied in the future (such as an optical posture sensor that locates the posture through image recognition, and a magnetic sensing posture module that calculates the posture through magnetic field changes); the appropriate hardware can be selected according to the clinical scene requirements (such as selecting a low-cost IMU for a home scene and selecting a high-precision optical sensor for a professional diagnosis and treatment scene), taking into account practicality and expandability. Through fixing methods such as pasting and buckling, the sensor is ensured to move synchronously with the pelvic bone, avoiding distortion of the posture data caused by displacement of the sensor; the core position of the hardware installation is the pelvic bone area, because the pelvic bone is a key skeletal structure to which the pelvic floor muscles are attached, and its posture change is directly related to the activity of the pelvic floor muscles (such as the pelvic floor muscles being prone to tension when the pelvic bone is tilted forward, which may also cause interference from adjacent muscle groups), and the posture signal collected at this position can most accurately reflect the scene background of the pelvic floor muscle activity. The scene auxiliary signal includes acceleration data and angular velocity data, as well as pitch angle time series signal and roll angle time series signal calculated from the two. The original data reserves space for subsequent complex processing, and the derived data can be directly used to judge the posture change trend (such as an increase in the pitch angle indicating forward flexion of the trunk, which may be accompanied by an increase in abdominal straight muscle interference), ensuring that the scene auxiliary signal can directly serve the judgment of the interference scene and improving the efficiency and relevance of signal processing. In this embodiment, the inertial measurement unit for biological motion monitoring can integrate a three-axis accelerometer and a three-axis gyroscope; when in use, it is fixed to a landmark position of the pelvic bone (such as the midpoint of the anterior superior iliac spine connecting line) through medical tape or elastic bandage, ensuring that the inertial measurement unit is rigidly connected to the pelvic bone and avoiding distortion of the posture signal caused by relative sliding; after being fixed, a brief posture calibration (such as remaining supine for 3-5 seconds) is required to eliminate the initial zero offset error; in addition, the inertial measurement unit can also be connected to the processor through wireless connection, and the posture signal collected by the wireless inertial measurement unit is sent to the processor; whether the inertial measurement unit is integrated in the pelvic floor muscle repair instrument is not specifically limited herein.

[0067] For example, the operator takes out the pelvic floor muscle repair instrument, selects the inertial measurement unit (IMU) as the core hardware of the posture sensing module; fixes the IMU at the midpoint of the iliac spine of the patient's pelvis through a medical waterproof sticker (this position can most accurately reflect the change of the pelvis posture), ensures that the sticker is firmly pasted and the IMU is not loose; checks that the electrical connection between the IMU and the signal processing module is normal, and confirms that the device can receive posture data in real time. During the evaluation process, the patient needs to complete the "supine → sit up → forward leaning" three position changing actions according to the doctor's advice, and keep each action for 3 seconds and synchronously contract the pelvic floor muscle: when supine, the IMU collects acceleration data (vertical acceleration about 1g), angular velocity data (close to 0 rad / s), and solves to get the pitch angle about 0°, generating the scene auxiliary signal of the static supine scene; when sitting up, the IMU collects angular velocity data (trunk rotation angular velocity about 0.5 rad / s) and vertical acceleration fluctuation (0.8-1.2g), and solves to get the pitch angle from 0° to 30°, generating the scene auxiliary signal of the dynamic sitting up scene; when leaning forward, the IMU collects horizontal acceleration (about 0.3g), and solves to get the pitch angle to 45°, generating the scene auxiliary signal of the forward leaning effort scene.

[0068] In this embodiment, the inertial measurement unit, optical attitude sensor or magnetic sensing attitude module is used to collect the posture signal related to the activity of the pelvic floor muscle. By introducing the posture sensing, the posture sensing module can provide the time sequence data of the body posture, which provides auxiliary information for the signal processing module to judge the source of interference and dynamic changes in complex scenes.

[0069] Further, in some embodiments, the signal processing module is configured to: respectively pre-process the target electromyographic signal, the interference reference signal and the scene auxiliary signal; based on the pre-processed target electromyographic signal, the interference reference signal and the scene auxiliary signal, feature extraction is performed to obtain multi-dimensional features.

[0070] Specifically, the signal processing module needs to preprocess the target electromyographic signal, the interference reference signal and the scene auxiliary signal respectively. Due to the problems such as noise (such as electrode contact noise, environmental electromagnetic interference) and inconsistent signal format (such as sampling rate difference) in the clinical collection scene, the preprocessing can eliminate invalid noise and unify the signal format through the internal hardware circuit (such as filter circuit) or basic operation unit of the device, so as to ensure that the subsequent feature extraction can be based on clean and regular signal data, and avoid feature extraction deviation caused by original signal defects. After the preprocessing based on the feature extraction signal processing module is completed, the feature extraction needs to be carried out based on the preprocessed target electromyographic signal, interference reference signal and scene auxiliary signal, and multi-dimensional features are obtained. The optimized original signal is converted into feature data that can be used for interference judgment. Through a special feature extraction operation unit (such as an algorithm module in an embedded processor), key features (such as time domain, frequency domain and posture features) that can distinguish the target signal of the pelvic floor muscle from the interference of the adjacent muscle group are mined from the three types of preprocessed signals, and a multi-dimensional feature set is formed.

[0071] For example, during the evaluation, the pelvic floor muscle electromyographic acquisition module (attached to the perineum) acquires a target electromyographic signal containing rectus abdominis interference and environmental electromagnetic noise (such as 50Hz power frequency noise generated by the evaluation room equipment); the interference muscle group monitoring module (attached to the rectus abdominis muscle) acquires a rectus abdominis interference reference signal containing slight contact noise; and the posture sensing module (fixed to the pelvis) acquires a scene auxiliary signal (acceleration and angular velocity data) containing sensor jitter noise. The three types of original signals are transmitted synchronously through internal wires in the device to the signal processing module.

[0072] For the target electromyographic signal and the interference reference signal, the 50Hz power frequency noise and high-frequency clutter are filtered out through the band-pass filter circuit (hardware implementation) built in the device, and the effective frequency range (such as 20-500Hz) of the electromyographic signal is retained. For the scene auxiliary signal, the signal calibration unit of the device eliminates abnormal data points caused by sensor jitter to ensure the stability of the acceleration and angular velocity data.

[0073] After the preprocessing is completed, the feature extraction operation unit of the signal processing module starts to work, extracts time domain features (such as signal peak value and contraction duration) and frequency domain features (such as signal main frequency distribution) from the preprocessed target electromyographic signal and interference reference signal, and extracts posture features (such as pelvis pitch angle change rate) from the preprocessed scene auxiliary signal. The above features are integrated to form a multi-dimensional feature set, which is ready for subsequent fusion with the original signal and realization of interference suppression.

[0074] The pre-processing operations of the three types of signals in this embodiment eliminate noise, format differences and other defects in the original collected signals through device hardware, avoid the influence of invalid interference on subsequent feature extraction, and ensure that the input signal of the feature extraction link is clean and regular. On the other hand, based on the pre-processed signal, multi-dimensional features are extracted, which enables the feature extraction to focus on the core effective information of the signal, reduces the feature deviation caused by the defects of the original signal, and ensures that the generated multi-dimensional features can accurately reflect the differences, and the data source quality of the subsequent feature fusion and interference suppression link is guaranteed.

[0075] Further, in some embodiments, the signal processing module is configured to: perform band-pass filtering processing on the target electromyographic signal and the interference reference signal to obtain filtered target electromyographic signal and interference reference signal; perform attitude solving and linear interpolation processing on the scene auxiliary signal to make the sampling rate of the interpolated scene auxiliary signal consistent with the sampling rate of the target electromyographic signal; and perform normalization processing on the filtered target electromyographic signal, the interference reference signal and the interpolated scene auxiliary signal respectively to obtain pre-processed multi-source signals.

[0076] Specifically, for the pre-processing of the target electromyographic signal and the interference reference signal, the signal processing module needs to perform filtering processing on the target electromyographic signal and the interference reference signal through the built-in band-pass filter circuit (such as an active band-pass filter module based on an operational amplifier). The effective frequency range of electromyographic signal is usually 20-500Hz, while 50Hz power frequency noise (such as power supply equipment interference in the evaluation room) and high-frequency clutter (such as electrode contact noise) are easily mixed in the clinical collection scene. The band-pass filter circuit can accurately retain the electromyographic signal in the effective frequency range, filter out invalid noise components, avoid the interference of noise on subsequent feature extraction (such as frequency domain feature analysis), and ensure the purity of the two types of electromyographic signals.

[0077] For the pre-processing of the scene auxiliary signal, the attitude solving and linear interpolation signal processing module needs to perform solving on the scene auxiliary signal (acceleration, angular velocity data) through a dedicated attitude solving unit (such as a hardware module of an attitude solving algorithm in an embedded processor) to obtain pitch angle and roll angle data that can directly reflect the posture of the pelvic bone. At the same time, since the sampling rate of the scene auxiliary signal may not be consistent with the sampling rate of the target electromyographic signal, the device needs to adjust the sampling rate of the solved attitude signal through a linear interpolation hardware module to make the sampling rate of the interpolated scene auxiliary signal completely synchronized with the target electromyographic signal. This operation can eliminate the time dimension misalignment of multi-source signals and ensure that the subsequent feature extraction can be based on time-aligned signal data.

[0078] For the unified preprocessing of the three types of signals, the normalization signal processing module needs to perform normalization processing on the filtered target electromyographic signal, the interference reference signal, and the interpolated scene auxiliary signal respectively through a normalization operation unit (such as an amplitude normalization module based on a digital signal processor (DSP)). 2 The original amplitude ranges of the three types of signals are quite different (for example, the amplitude of the electromyographic signal is usually 0.1-5 mV, and the amplitude of the acceleration signal is usually 0-10 m / s 2 ). Normalization can uniformly map the amplitudes of all signals to the same range (such as [0, 1] or [-1, 1]), eliminate the feature weight imbalance problem caused by the difference in signal amplitudes, and ensure that all types of signal features can be considered equally during subsequent feature extraction.

[0079] The present embodiment includes band-pass filtering of the target electromyographic signal and the interference reference signal, attitude solving and interpolation processing of the scene auxiliary signal, and normalization processing. These preprocessing steps can effectively improve the signal quality, remove noise, and unify the sampling rate, providing a high-quality data basis for subsequent feature extraction and fusion

[0080] Further, in some embodiments, the signal processing module is configured to generate a corresponding first feature map based on the target electromyographic signal, the interference reference signal, and the scene auxiliary signal; generate a corresponding second feature map based on the multi-dimensional features; and splice the first feature map and the second feature map to generate an integrated feature.

[0081] Specifically, the signal processing module needs to generate a corresponding first feature map based on the target electromyographic signal, the interference reference signal and the scene auxiliary signal through the built-in feature map conversion unit (such as the imaging operation module in the embedded processor). The first feature map is a hardware data carrier in which a one-dimensional time sequence signal is converted into a two-dimensional matrix form. For example, the target electromyographic signal is divided into multiple segments according to a time window, and the amplitude, frequency and other information of each segment correspond to a pixel point or a region of the feature map, so that the time-space characteristics of the original signal are presented in a visualized form (hardware level: data matrix), which facilitates the subsequent capture of the correlation between signals by the device. The signal processing module needs to generate a corresponding second feature map based on the multi-dimensional features through the same feature map conversion unit. Unlike the first feature map, the core of the second feature map is to present the correlation characteristics of the artificially extracted features, such as the peak value in the time domain feature, the main frequency in the frequency domain feature, and the angle change rate in the posture feature. These discrete features are arranged in a two-dimensional matrix according to logical correlation, so that the difference characteristics of multi-dimensional features can be more intuitively recognized by the device, avoiding the judgment deviation caused by the isolation of a single feature. The signal processing module needs to generate integrated features by splicing the first feature map and the second feature map through a feature map splicing unit (such as a matrix splicing hardware module in a digital signal processor (DSP)). The purpose is to deeply integrate the time-space characteristics of the original signal and the difference characteristics of the artificial features, so that the integrated features can cover both the original form of the signal and the interference judgment basis, providing more comprehensive hardware data support for subsequent interference suppression.

[0082] The first feature map and the second feature map are generated and spliced to generate integrated features, so that the pelvic floor muscle rehabilitation instrument can deeply integrate the time-space characteristics of the original signal and the difference characteristics of the multi-dimensional features. The generated integrated features can comprehensively represent the correlation between the target signal of the pelvic floor muscles and the adjacent muscle group interference, avoiding the limitations of traditional single signal or single feature judgment, and providing high-quality hardware data support for subsequent precise interference suppression.

[0083] Further, in some embodiments, the signal processing module is configured to: calculate an adjacent muscle group interference component prediction value based on the integrated features; determine a preliminary de-interference signal based on the difference between the target electromyographic signal and the adjacent muscle group interference component prediction value; and perform high-frequency interference removal processing on the preliminary de-interference signal to generate a purified pelvic floor electromyographic signal.

[0084] Specifically, the signal processing module calculates the adjacent muscle group interference component prediction value based on the integrated features through the built-in interference prediction operation unit (such as a prediction module based on an embedded neural network chip or a digital signal processor (DSP)). The integrated features are data carriers generated by the device, which contain the spatial and temporal characteristics of the original signal and the differences in multi-dimensional features. The interference prediction operation unit extracts features related to adjacent muscle group interference (such as the frequency and amplitude variation of the interference reference signal) from the integrated features through hardware algorithm logic (such as signal feature matching and interference pattern recognition), and then predicts the specific component data of the adjacent muscle group interference in the target electromyographic signal (such as the amplitude and waveform of the interference signal within a certain time window), providing a quantitative basis for precise positioning of interference. The signal processing module determines the preliminary interference-removed signal based on the difference between the target electromyographic signal and the adjacent muscle group interference component prediction value through the signal subtraction operation unit (analog or digital subtraction circuit at the hardware level). Since the target electromyographic signal is the superposition of the pelvic floor muscle effective signal and the adjacent muscle group interference signal, by subtracting the corresponding interference component prediction value from the amplitude and timing data of the target electromyographic signal through the hardware circuit, most of the interference components can be directly stripped, obtaining the preliminary interference-removed signal mainly composed of the pelvic floor muscle effective signal. The signal processing module generates the purified pelvic floor electromyographic signal by removing the high-frequency interference through the high-frequency interference filter unit (such as a hardware low-pass filter circuit or an adaptive filter module), and finally generates the purified pelvic floor electromyographic signal. A small amount of high-frequency noise (such as electronic noise and high-frequency electromagnetic interference introduced during signal acquisition or operation) may be left in the preliminary interference-removed signal. This kind of noise does not belong to the effective frequency range of the pelvic floor electromyographic signal (usually 20-500Hz), and can be accurately filtered out through the hardware filter unit while maximizing the preservation of the physiological characteristics of the pelvic floor muscle effective signal (such as contraction peak value and duration), ensuring that the output signal meets the needs of high purity and high integrity for clinical evaluation.

[0085] The interference suppression operation of the present embodiment can accurately locate and remove the adjacent muscle group interference in the target electromyographic signal, filter out residual high-frequency noise, maximize the preservation of the physiological characteristics of the pelvic floor muscle effective signal, avoid the problems of incomplete noise removal or damage to the effective signal in traditional methods, and finally output high-purity pelvic floor electromyographic signals that meet the clinical evaluation standards, effectively improving the accuracy of device interference suppression and the quality of signal output.

[0086] Figure 2 An application environment diagram for the pelvic floor electromyographic signal interference removal method in one embodiment. Refer to Figure 2The pelvic floor muscle electrical signal interference removal method is applied to an abdominal rectus muscle interference removal system. The abdominal rectus muscle interference removal system includes a terminal 110 and a server 120. The terminal 110 and the server 120 are connected through a network, and the terminal 110 can be a desktop terminal or a mobile terminal, and the mobile terminal can be at least one of a mobile phone, a tablet computer, a notebook computer, etc. The server 120 can be implemented by an independent server or a server cluster composed of multiple servers. The server 120 is configured to execute the above-mentioned pelvic floor muscle electrical signal interference removal method, which includes: collecting a pelvic floor muscle surface electromyogram signal containing adjacent muscle group interference to obtain a target electromyogram signal; collecting a physiological activity signal of the adjacent muscle group of the pelvic floor muscle to obtain an interference reference signal; collecting a body posture time sequence signal related to the activity of the pelvic floor muscle to obtain a scene auxiliary signal; and based on the interference reference signal and the scene auxiliary signal, performing suppression processing on the adjacent muscle group interference in the target electromyogram signal to output a purified pelvic floor muscle electrical signal.

[0087] Please refer to Figure 3 , Figure 3 is a flowchart of the pelvic floor muscle electrical signal interference removal method provided by an embodiment of the present application. The embodiment mainly takes the application of the pelvic floor muscle electrical signal interference removal method to a computer device as an example for illustration. The pelvic floor muscle electrical signal interference removal method provided by an embodiment of the present application can specifically include the following steps:

[0088] S1. Collecting a pelvic floor muscle surface electromyogram signal containing adjacent muscle group interference to obtain a target electromyogram signal;

[0089] Specifically, for step S1, the pelvic floor muscle electromyogram acquisition module of the electromyogram purification device is used to complete the collection of the pelvic floor muscle surface electromyogram signal containing adjacent muscle group interference, and finally the target electromyogram signal is obtained. The pelvic floor muscle electromyogram acquisition module is a core acquisition unit specially designed by the device to capture the physiological electrical signal of the pelvic floor muscle. Its hardware configuration (such as electrode type, signal amplification circuit) needs to be adapted to the characteristics of the pelvic floor muscle electrical signal to ensure that the electrical signal generated during the contraction of the pelvic floor muscle can be effectively captured. The pelvic floor muscle surface electromyogram signal containing adjacent muscle group interference is collected. Because the pelvic floor muscle is adjacent to the adjacent muscle groups such as the abdominal rectus muscle and the lumbar quadratus muscle in the clinical scene, the pelvic floor muscle contraction is easy to be accompanied by the linkage of the adjacent muscle groups, which leads to the natural mixing of interference components in the collected signal. The signal containing interference is the core processing object of subsequent interference suppression, and there is no need to filter the interference in advance. Finally, the target electromyogram signal is output. This signal needs to be directly related to the functional state of the pelvic floor muscle and can reflect the key physiological information such as the activation intensity and contraction regularity of the pelvic floor muscle, thereby providing basic data for subsequent judgment of the pelvic floor muscle function.

[0090] S2. Collecting a physiological activity signal of the adjacent muscle group of the pelvic floor muscle to obtain an interference reference signal;

[0091] Specifically, in step S2, the interference muscle group monitoring module of the electromyography (EMG) signal purification device collects physiological activity signals of adjacent muscle groups of the pelvic floor muscles and ultimately outputs an interference reference signal. This module is a dedicated unit designed specifically for capturing interference source signals. Its hardware parameters (such as electrode sensitivity and signal detection frequency band) are adapted to the EMG signal characteristics of adjacent muscle groups (such as the rectus abdominis), enabling precise capture of the physiological signals generated during the activity of the interference muscle groups. Focusing on the physiological activity of adjacent muscle groups of the pelvic floor muscles, which are anatomically close to the pelvic floor muscles and easily contract in conjunction with them during pelvic floor muscle activity (such as the rectus abdominis and quadratus lumborum), the physiological activity signals are mainly the EMG signals generated when these muscle groups contract, which can be directly used as a characteristic benchmark for identifying interference. The core function of the finally obtained interference reference signal is to provide the signal characteristics of the interference source (such as amplitude, frequency, and temporal changes), providing a direct comparative basis for distinguishing the interference components from the effective components of the pelvic floor muscles from the target EMG signal.

[0092] S3. Collect body posture time-series signals related to pelvic floor muscle activity to obtain scene auxiliary signals;

[0093] Specifically, in step S3, the posture perception module of the electromyography (EMG) signal purification device collects temporal signals of body posture related to pelvic floor muscle activity, ultimately outputting a scene assistance signal. Collecting posture signals related to pelvic floor muscle activity is crucial, as posture changes directly affect the state of pelvic floor muscle activity or the probability of interference. The temporal signal refers to posture data collected continuously in chronological order, reflecting the dynamic changes in posture over time. The core function of the final scene assistance signal is to provide the context in which interference occurs. By analyzing posture changes (such as sudden changes in pitch angle or fluctuations in angular velocity), it helps determine whether interference in the target EMG signal is related to a specific posture, improving the accuracy of interference identification.

[0094] S4. Based on the interference reference signal and the scene auxiliary signal, the interference of adjacent muscle groups in the target electromyographic signal is suppressed, and the purified pelvic floor electromyographic signal is output.

[0095] Specifically, for step S4, the characteristics (such as amplitude fluctuation rule, frequency distribution, timing synchronization) of the target electromyographic signal and the interference reference signal are compared. If the signal characteristics of a certain frequency band (such as 25-30 Hz) and a certain time period in the target electromyographic signal are highly consistent with the interference reference signal, it is preliminarily determined that this part is adjacent muscle group interference. Combined with the scene auxiliary signal, it is verified whether the interference component determined in the above preliminary determination is related to the posture change. For example, when the 25-30 Hz component in the target electromyographic signal is enhanced, the scene auxiliary signal shows that the pelvic bone pitch angle is increased, which further confirms the authenticity of the interference and excludes the case that the pelvic floor muscle effective signal and the interference characteristics are accidentally overlapped. For the confirmed interference component, signal subtraction, adaptive filtering and other methods are used to remove the adjacent muscle group interference in the target electromyographic signal. The physiological characteristics of the pelvic floor muscle effective signal after processing are retained, such as the main frequency 30-50 Hz of the pelvic floor electromyographic signal and the contraction peak value. Finally, the purified pelvic floor electromyographic signal is output.

[0096] The embodiment synchronously collects the target electromyographic signal, the interference reference signal and the scene auxiliary signal, takes the interference characteristic reference and the scene correlation verification as double bases, can accurately identify and remove the adjacent muscle group interference in the target electromyographic signal, maximally retains the physiological characteristics of the pelvic floor muscle effective signal, solves the problems of inaccurate interference suppression and effective signal misdeletion caused by lack of interference reference and scene basis in the traditional method, and finally outputs the purified signal with high purity and strong reliability, which meets the demand of high-quality electromyographic signal for clinical pelvic floor function evaluation (such as muscle strength classification and contraction coordination analysis).

[0097] Further, in some embodiments, as shown in Figure 4 Step S4 “based on the interference reference signal and the scene auxiliary signal, the adjacent muscle group interference in the target electromyographic signal is suppressed and processed, and the purified pelvic floor electromyographic signal is output”, specifically can include:

[0098] S41. Based on the target electromyographic signal, the interference reference signal and the scene auxiliary signal, multi-dimensional features are extracted, and the multi-dimensional features include time domain features, frequency domain features and high-order statistical features extracted from the target electromyographic signal and the interference reference signal, and posture features extracted from the scene auxiliary signal;

[0099] Specifically, for step S41, the features extracted from the target electromyographic signal and the interference reference signal, both of which are physiological signals of the electromyographic type, have consistent feature extraction logic, facilitating subsequent comparative analysis: time domain features are used to reflect the intuitive characteristics of signals changing over time, such as the amplitude range (maximum value, minimum value, average value) of the signal, the peak value (the maximum amplitude of a single contraction), the contraction duration (the time length from the signal rising to the peak value and then falling to the baseline), and the rising / falling slope (reflecting the speed of muscle activation and relaxation). Such features can directly reflect the intensity and time regularity of muscle activity, for example, the peak value of the target signal during pelvic floor muscle contraction is usually higher than that of the interference signal, and the duration is synchronized with the contraction action. Frequency domain features are used to reflect the frequency distribution regularity of the signal, which is obtained through Fourier transform and other methods, such as the main frequency (the frequency with the most concentrated energy) of the signal, the frequency bandwidth (the frequency range of the effective signal), and the energy proportion of a specific frequency band (such as the 20-50Hz frequency band energy proportion). The frequency domain features of electromyographic signals of different muscle groups differ (such as the main frequency of pelvic floor electromyographic signals is usually 30-50Hz, and the main frequency of abdominal rectus electromyographic signals is usually 20-30Hz), and the frequency domain features are the core basis for distinguishing target signals from interference signals. High-order statistical features are used to reflect the complex fluctuation characteristics and non-Gaussian distribution regularity of the signal, such as the variance (reflecting the amplitude dispersion degree) of the signal, the skewness (reflecting the asymmetry of the signal waveform), the kurtosis (reflecting the sharpness of the signal peak), and the autocorrelation coefficient (reflecting the similarity of the signal at different time points). Such features can capture subtle differences that time domain and frequency domain features cannot cover, for example, the skewness and kurtosis of the interference signal are usually significantly different from those of the target signal due to the influence of posture fluctuations.

[0100] The features extracted from the scene auxiliary signal include: posture angle features and posture dynamic features. The posture angle features, such as the real-time values and variation ranges of the pitch angle (forward and backward inclination angle) and roll angle (left and right inclination angle) of the pelvic bone, reflect the static state of the body posture. The posture dynamic features, such as the change rate (angle change amount per unit time) of the pitch angle and roll angle, and the posture stability duration (time of maintaining a certain posture), reflect the dynamic change trend of the body posture. Such features can be associated with the scene of interference occurrence (such as when the pitch angle change rate increases, the probability of abdominal rectus muscle interference significantly increases).

[0101] S42. Feature fusion is performed on the target electromyographic signal, the interference reference signal, the scene auxiliary signal, and the multi-dimensional features to generate integrated features.

[0102] Specifically, for step S43, the three types of original signals (target electromyographic signal, interference reference signal, and scene auxiliary signal) collected and the multi-dimensional features extracted are fused to generate integrated features. Through logical correlation and data structuring processing, scattered information is integrated into a comprehensive data carrier that can represent the essence of the signal, interference characteristics, and scene background at the same time.

[0103] S43. Based on the integrated features, the adjacent muscle group interference in the target electromyographic signal is suppressed to obtain a purified pelvic floor electromyographic signal;

[0104] Specifically, for step S43, the generated integrated features are used as the basis for judgment to perform targeted suppression processing on the adjacent muscle group interference in the collected target electromyographic signal, and finally obtain the purified pelvic floor electromyographic signal. The specific processing procedure is as follows: based on the correlation in the integrated features, the components in the target electromyographic signal that belong to interference are identified. For the located interference components, signal subtraction, filtering and other methods are used for removal. The signal after removing the interference is verified to ensure that the effective components of the target signal are not mistakenly deleted (such as by comparing the time domain peak value and contraction duration of the target signal before and after processing, if the change rate is less than 10%, it is determined to be effective), if there is a mistake, adjust the suppression parameters (such as reduce the filtering strength), and finally output the purified pelvic floor electromyographic signal.

[0105] The embodiment effectively distinguishes the pelvic floor electromyographic signal from the adjacent muscle group interference, removes the interference while retaining the physiological characteristics of the effective pelvic floor muscle signal to the greatest extent, solves the problem of blind interference judgment and incomplete noise removal in traditional methods, provides a high-quality purified signal for pelvic floor function evaluation, and significantly improves the accuracy and reliability of the evaluation results.

[0106] Further, in some embodiments, step S41 "extracting multi-dimensional features based on the target electromyographic signal, the interference reference signal and the scene auxiliary signal" can specifically include:

[0107] S411. Signal preprocessing is performed on the target electromyographic signal, the interference reference signal and the scene auxiliary signal, respectively;

[0108] S412. Feature extraction is performed based on the preprocessed target electromyographic signal, the interference reference signal and the scene auxiliary signal to obtain multi-dimensional features.

[0109] Specifically, signal preprocessing operations are performed on the collected target electromyographic signal, interference reference signal and scene auxiliary signal. The purpose of preprocessing is to eliminate invalid components (such as noise, format differences) in the original signal that affect feature extraction, unify the signal quality and format, and ensure that the features extracted later can accurately reflect the nature of the signal (such as pelvic floor muscle effective signal features, interference features, posture features), and avoid feature deviation caused by defects in the original signal. Because the types (electromyographic signal, posture time sequence signal) and characteristics (amplitude range, frequency distribution, data format) of the three types of signals are different, appropriate preprocessing operations need to be designed for the characteristics of each type of signal, rather than using a unified processing method. For example, electromyographic signals need to be filtered to remove noise, and posture signals need to be adjusted to adjust the timing format.

[0110] The pre-processed target electromyographic signal, the interference reference signal, and the scene auxiliary signal are taken as inputs to extract multi-dimensional features, and finally multi-dimensional features are obtained. The pre-processed signals have eliminated noise and unified format, and can more accurately reflect the real characteristics of the signals. The multi-dimensional features extracted on this basis can effectively distinguish the effective signal of the pelvic floor muscle, the interference signal of the adjacent muscle group, and the posture scene related characteristics. The time domain features, frequency domain features, and high-order statistical features extracted from the pre-processed target electromyographic signal and interference reference signal, and the posture features extracted from the pre-processed scene auxiliary signal, ensure that the features can fully characterize the signal differences and scene correlation.

[0111] The embodiment can fully exploit the key information in the signal by pre-processing the collected signal and extracting time domain, frequency domain, high-order statistical features and posture features from the pre-processed signal, thereby providing rich data support for feature fusion and interference suppression.

[0112] Further, in some embodiments, the step S411 of respectively pre-processing the target electromyographic signal, the interference reference signal, and the scene auxiliary signal can specifically include:

[0113] S4111. Perform band-pass filtering on the target electromyographic signal and the interference reference signal to obtain filtered target electromyographic signal and interference reference signal;

[0114] S4112. Perform posture solving and linear interpolation processing on the scene auxiliary signal to make the sampling rate of the interpolated scene auxiliary signal consistent with the sampling rate of the target electromyographic signal;

[0115] S4113. Perform normalization processing on the filtered target electromyographic signal, the interference reference signal, and the interpolated scene auxiliary signal respectively to obtain pre-processed multi-source signals.

[0116] Specifically, band-pass filtering is performed on the collected target electromyographic signal (including pelvic floor surface electromyographic signal with adjacent muscle group interference) and the interference reference signal (physiological activity signal of the pelvic floor muscle adjacent muscle group), and finally filtered target electromyographic signal and interference reference signal are obtained. The effective frequency range of electromyographic signal is usually 20-500Hz, while 50Hz power frequency noise (such as power supply equipment interference in the evaluation room) and high-frequency clutter (such as electrode contact noise and environmental electromagnetic interference) are easily mixed in the clinical collection scene. Band-pass filtering can accurately retain electromyographic signals within the effective frequency range, filter out invalid noise components, and avoid interference of noise on subsequent feature extraction (such as frequency domain feature analysis). Since the target electromyographic signal and the interference reference signal both belong to electromyographic physiological signals, their effective frequency ranges are similar, so the same band-pass filtering parameters (such as passband 20-500Hz) are used to ensure that the filtering standards of the two types of signals are consistent, and the features extracted subsequently are comparable.

[0117] For the collected scene auxiliary signal (body posture timing signal related to pelvic floor muscle activity), posture solving and linear interpolation operation are performed in turn, and finally the interpolated scene auxiliary signal is obtained, and the sampling rate of the signal is consistent with the sampling rate of the target electromyographic signal. The original data of the scene auxiliary signal is usually basic motion data such as acceleration and angular velocity (such as 3-axis acceleration and 3-axis angular velocity collected by an inertial measurement unit), which needs to be converted into parameters that can directly reflect the body posture (such as pitch angle and roll angle) through posture solving, so that the signal can be directly related to the scene background of pelvic floor muscle activity (such as an increase in pitch angle indicating forward bending of the trunk, which may be accompanied by increased interference from the rectus abdominis muscle). Because the sampling rate of the scene auxiliary signal (such as the commonly used 100Hz of the inertial measurement unit) usually differs from the sampling rate of the target electromyographic signal (such as the commonly used 1000Hz of electromyographic collection), linear interpolation can supplement data points through hardware or algorithm to increase the sampling rate of the scene auxiliary signal to be consistent with the target electromyographic signal, eliminate the time dimension misalignment, and ensure that the three types of signals can be accurately corresponded according to the time stamp during subsequent feature extraction.

[0118] For the filtered target electromyographic signal, the interference reference signal, and the interpolated scene auxiliary signal, normalization processing is performed respectively, and finally the preprocessed multi-source signal is obtained. The original amplitude range of the three types of signals is significantly different (such as the amplitude of the electromyographic signal is usually 0.1-5mV, the amplitude of the acceleration signal is usually 0-10m / s 2 , and the amplitude of the angle signal is usually 0-90°), and normalization can map the amplitudes of all signals to the same range (such as [0, 1] or [-1, 1]), eliminating the feature weight imbalance problem caused by amplitude difference—such as avoiding the feature of the acceleration signal with larger amplitude “covering” the feature of the electromyographic signal with smaller amplitude during subsequent feature extraction, ensuring that all types of signal features can be considered equally. Normalize the three types of signals respectively, calculate the normalization coefficient based on the maximum and minimum values of the amplitude of each type of signal, and ensure that the amplitude variation of a single type of signal can be retained after normalization (such as the contraction peak value of the target electromyographic signal can still reflect the relative size after normalization).

[0119] This embodiment can effectively remove noise and interference components in the signal, unify the sampling rate and eliminate the dimension difference, thereby improving the signal quality and ensuring the accuracy of subsequent feature extraction and processing, through band-pass filtering, posture solving, linear interpolation and normalization processing.

[0120] Further, in some embodiments, the step S412 “performing feature extraction based on the preprocessed target electromyographic signal, the interference reference signal and the scene auxiliary signal to obtain multi-dimensional features” can specifically include:

[0121] S4121. Extracting time domain features, frequency domain features and high order statistical features from the preprocessed target electromyographic signal and the interference reference signal;

[0122] S4122. Extract the posture feature from the pre-processed scene auxiliary signal;

[0123] S4123. Perform splicing processing on the time domain feature, frequency domain feature, high-order statistical feature and posture feature to generate multi-dimensional features.

[0124] Specifically, taking the pre-processed target electromyographic signal (pelvic floor surface electromyographic signal containing adjacent muscle group interference after band-pass filtering and normalization) and the interference reference signal (pelvic floor adjacent muscle group physiological activity signal after band-pass filtering and normalization) as inputs, three types of features are extracted respectively:

[0125] The time domain feature is used to reflect the intuitive characteristics of the signal changing with time, such as the amplitude statistics (maximum value, minimum value, average value, peak value) of the signal, the time parameters (contraction duration, rise time, fall time), the waveform characteristics (signal duty cycle, zero-crossing rate) and the like. Such features can directly reflect the intensity and time law of muscle activity (such as the target signal peak value being higher than the interference signal during pelvic floor muscle contraction, and the duration being synchronized with the contraction action).

[0126] The frequency domain feature is used to reflect the law of signal frequency distribution, which is obtained by Fourier transform, power spectrum analysis and the like, such as the main frequency (energy most concentrated frequency) of the signal, the frequency bandwidth (effective signal frequency range), the energy proportion of a specific frequency band (such as the 20-50Hz frequency band energy proportion), the spectral gravity center (center frequency after frequency weighting average) and the like. The frequency domain features of electromyographic signals of different muscle groups are different (such as the main frequency of pelvic floor electromyographic signal is usually 30-50Hz, and the main frequency of rectus abdominis electromyographic signal is usually 20-30Hz), which is the core basis for distinguishing target signal and interference signal.

[0127] The high-order statistical feature is used to reflect the complex fluctuation and non-Gaussian distribution characteristics of the signal, such as the variance (reflecting the amplitude dispersion degree) of the signal, the skewness (reflecting the signal waveform asymmetry), the kurtosis (reflecting the signal peak sharpness), the autocorrelation coefficient (reflecting the similarity of the signal at different time points) and the like. Such features can capture subtle differences that cannot be covered by time domain and frequency domain features (such as the interference signal is usually different from the target signal in skewness and kurtosis due to the influence of posture fluctuation).

[0128] The pre-processed scene auxiliary signal (the body posture time sequence signal related to the pelvic floor muscle activity after attitude solution, linear interpolation, and normalization) is taken as an input to extract posture features. The posture feature types include posture static features (such as real-time values of the pitch angle and roll angle of the pelvis, the angle range during posture stabilization) and posture dynamic features (such as the change rate of the pitch angle / roll angle, the peak value of the angle change, the number of posture mutations, and the length of posture stabilization). Such features can be associated with the scene in which the interference occurs (for example, when the pitch angle change rate increases, the probability of rectus abdominis interference significantly increases; when the length of posture stabilization is short, the interference signal fluctuation is more obvious).

[0129] The extracted target electromyographic signal time domain features, target electromyographic signal frequency domain features, target electromyographic signal high-order statistical features, interference reference signal time domain features, interference reference signal frequency domain features, interference reference signal high-order statistical features, and posture features are spliced to form a unified multi-dimensional feature set. The splicing is structured in the order of “signal type-feature category”, for example, the three types of features of the target electromyographic signal are arranged first, the three types of features of the interference reference signal are arranged second, and the posture features are arranged last. Each feature is presented in the form of a numerical value to form a one-dimensional feature vector or a two-dimensional feature matrix, ensuring the order and interpretability of the features. The dispersed electromyographic signal features and posture features are integrated, and the multi-dimensional features can simultaneously cover the essence of the pelvic floor muscle signal, the interference signal features, and the scene-related information, providing a complete feature basis for the subsequent feature association.

[0130] The embodiment can comprehensively reflect the pelvic floor electromyographic signal and its interference by extracting time domain, frequency domain, and high-order statistical features from the target electromyographic signal and the interference reference signal, extracting posture features from the scene auxiliary signal, and generating multi-dimensional features by splicing these features, thereby providing comprehensive feature support for subsequent interference suppression.

[0131] Further, in some embodiments, as shown in FIG. 42, the step S42 “performing feature fusion on the target electromyographic signal, the interference reference signal, the scene auxiliary signal, and the multi-dimensional features to generate integrated features” can specifically include: Figure 5

[0132] S421. generating a corresponding first feature map based on the target electromyographic signal, the interference reference signal, and the scene auxiliary signal;

[0133] S422. generating a corresponding second feature map based on the multi-dimensional features;

[0134] S423. splicing the first feature map and the second feature map to generate integrated features.

[0135] ​Specifically, the collected target electromyography signal (containing the surface electromyography signal of the pelvic floor muscle with interference from the adjacent muscle group), the interference reference signal (the physiological activity signal of the adjacent muscle group of the pelvic floor muscle), and the scene auxiliary signal (the body posture timing signal related to the pelvic floor muscle activity) are taken as inputs to generate a corresponding first feature map. The first feature map is the result of converting one-dimensional timing signal into a two-dimensional structured data carrier (such as a matrix or image data). Specifically, the timing data of the three types of signals is divided into multiple segments according to a fixed time window (such as 100 ms / window), and the key parameters (such as the amplitude of the target electromyography signal, the frequency of the interference reference signal, and the angle of the scene auxiliary signal) of each segment correspond to a pixel point or data unit of the first feature map, so that the time and signal value dynamic change relationship of the original signal is presented in a two-dimensional form.

[0136] The extracted multi-dimensional features are taken as inputs to generate a corresponding second feature map. The second feature map is the result of converting discrete multi-dimensional features into a two-dimensional structured data carrier. According to the time window, the multi-dimensional features extracted in each window (such as the peak value of the target electromyography signal, the main frequency of the interference reference signal, and the angle change rate of the scene auxiliary signal in a certain window) are arranged into a two-dimensional matrix according to the feature category, for example, the rows of the matrix correspond to the time window, the columns correspond to different feature types, and each matrix element is the specific value of a certain feature in the corresponding window.

[0137] The present embodiment realizes the deep fusion of the original signal features and multi-dimensional artificial features by generating the first feature map and the second feature map and splicing the two to generate the integrated features, which provides key support for subsequent efficient and accurate interference suppression.

[0138] Further, in some embodiments, as shown in Figure 6 Step S44 “based on the integrated features, suppressing the adjacent muscle group interference in the target electromyography signal to obtain the purified pelvic floor electromyography signal” can specifically include:

[0139] S441. Based on the integrated features, calculate the adjacent muscle group interference component prediction value;

[0140] S442. Based on the difference between the target electromyography signal and the adjacent muscle group interference component prediction value, determine the preliminary de-interference signal;

[0141] S443. Perform high-frequency interference removal processing on the preliminary de-interference signal to generate the purified pelvic floor electromyography signal.

[0142] Specifically, the generated integrated features (including target myoelectric signal, interference reference signal, scene auxiliary signal, and fusion data of multi-dimensional features) are taken as input, and through feature matching, rule analysis, etc., the specific component data of the target myoelectric signal belonging to the adjacent muscle group interference, i.e., the adjacent muscle group interference component prediction value, is calculated. The integrated features include the correlation between the reference signal features (such as the dominant frequency and amplitude of the rectus abdominis myoelectric signal), the scene auxiliary signal features (such as the posture change rate), and the target myoelectric signal features. Based on these correlations, the interference can be located. For example, if the integrated features show that in a certain time window, the frequency domain component (such as 28 Hz) of the target myoelectric signal is consistent with the dominant frequency (28 Hz) of the interference reference signal, and the posture change rate (1° / s) of the scene auxiliary signal meets the scene characteristics of the interference, it is determined that the 28 Hz component with the corresponding amplitude in the target signal in this window belongs to the interference, and then through signal amplitude ratio, frequency matching degree, etc. Parameters, the specific value of the interference component (such as amplitude 0.3 mV, duration 100 ms) is calculated to form the interference component prediction value.

[0143] The collected original target myoelectric signal is subtracted from the time-aligned adjacent muscle group interference component prediction value to obtain a signal after preliminary removal of the main interference, i.e., a preliminary interference-removed signal, through the target myoelectric signal amplitude and the interference component prediction value amplitude of the corresponding time window. The target myoelectric signal is essentially the superposition of the pelvic floor muscle effective signal and the adjacent muscle group interference signal. Under the premise that the integrated features have accurately located and quantified the interference component, the main interference component in the target signal can be directly stripped through amplitude subtraction at each time point. For example, the amplitude of the target myoelectric signal at a certain time point is 0.6 mV, and the corresponding interference component prediction value is 0.3 mV. After subtraction, the signal amplitude at this time point is 0.3 mV, which is approximately the amplitude of the pelvic floor muscle effective signal. After full-time sequence operation, a preliminary interference-removed signal is obtained, which is mainly composed of effective signals and only contains a small amount of subtle interference.

[0144] For the obtained preliminary interference-removed signal, filtering (such as low-pass filtering, adaptive filtering) is used to remove the residual high-frequency interference components (such as high-frequency noise introduced in the signal acquisition / operation process) in the signal, and finally a purified pelvic floor myoelectric signal is generated. Although the preliminary interference-removed signal has stripped the main interference of the adjacent muscle group, it may still contain high-frequency noise (such as electromagnetic interference with a frequency > 500 Hz). Such noise does not belong to the effective frequency range of the pelvic floor myoelectric signal (usually 20-500 Hz) and will affect the smoothness of the signal and the accuracy of clinical evaluation. Through filtering processing (such as setting 500 Hz as the low-pass filter cutoff frequency), high-frequency noise can be accurately filtered out, while the time and frequency domain features of the pelvic floor muscle effective signal (such as the dominant frequency of 30-50 Hz and the contraction peak value) are preserved, making the signal more suitable for the needs of high-purity and high-integrity in clinical evaluation.

[0145] The embodiment can effectively remove interference components by calculating the adjacent muscle group interference component prediction value and performing difference operation with the target muscle electrical signal, combining high-frequency interference removal processing, generate high-quality purified pelvic floor muscle electrical signal, significantly improve the purity and reliability of the signal.

[0146] It should be understood that, although Figure 3 The steps in the flowchart of the method for removing interference from a pelvic floor muscle electrical signal are shown in sequence according to the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, Figure 3 At least some of the steps in the method for removing interference from a pelvic floor muscle electrical signal can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be alternately or alternately executed with at least part of other steps or sub-steps or stages of other steps.

[0147] To better implement the method for removing interference from a pelvic floor muscle electrical signal, the embodiment of the present application also provides a pelvic floor muscle electrical signal interference removal device based on the above-mentioned method for removing interference from a pelvic floor muscle electrical signal. The meanings of the terms are the same as in the above-mentioned method for removing interference from a pelvic floor muscle electrical signal, and the specific implementation details can be referred to the description in the method embodiment.

[0148] Please refer to Figure 7 , Figure 7 The structure diagram of the pelvic floor muscle electrical signal interference removal device provided by the embodiment of the present application, which can specifically include a first acquisition module 201, a second acquisition module 202, a third acquisition module 203 and a signal purification module 204, and can be specifically as follows:

[0149] The first acquisition module 201 is configured to acquire a pelvic floor muscle surface electromyography signal containing adjacent muscle group interference to obtain a target muscle electrical signal.

[0150] The second acquisition module 202 is configured to acquire a physiological activity signal of the adjacent muscle group of the pelvic floor muscle to obtain an interference reference signal.

[0151] The third acquisition module 203 is configured to acquire a body posture timing signal related to the activity of the pelvic floor muscle to obtain a scene auxiliary signal.

[0152] The signal purification module 204 is configured to suppress the adjacent muscle group interference in the target muscle electrical signal based on the interference reference signal and the scene auxiliary signal, and output a purified pelvic floor muscle electrical signal.

[0153] Further, in some embodiments, the signal purification module 204 can specifically include:

[0154] a feature extraction unit, configured to extract multi-dimensional features based on the target electromyography signal, the interference reference signal, and the scene auxiliary signal, the multi-dimensional features including time-domain features, frequency-domain features, and high-order statistical features extracted from the target electromyography signal and the interference reference signal, and posture features extracted from the scene auxiliary signal;

[0155] a feature fusion unit, configured to perform feature fusion on the target electromyography signal, the interference reference signal, the scene auxiliary signal, and the multi-dimensional features, to generate integrated features;

[0156] an interference suppression unit, configured to perform suppression processing on adjacent muscle group interference in the target electromyography signal based on the integrated features, to obtain a purified pelvic floor electromyography signal.

[0157] Further, in some embodiments, the feature extraction unit is specifically configured to:

[0158] perform signal preprocessing on the target electromyography signal, the interference reference signal, and the scene auxiliary signal, respectively;

[0159] extract multi-dimensional features based on the preprocessed target electromyography signal, the interference reference signal, and the scene auxiliary signal.

[0160] Further, in some embodiments, the feature extraction unit is specifically further configured to:

[0161] perform band-pass filtering processing on the target electromyography signal and the interference reference signal, to obtain filtered target electromyography signal and interference reference signal;

[0162] perform posture solving and linear interpolation processing on the scene auxiliary signal, so that the sampling rate of the interpolated scene auxiliary signal is consistent with the sampling rate of the target electromyography signal;

[0163] perform normalization processing on the filtered target electromyography signal, the interference reference signal, and the interpolated scene auxiliary signal, respectively, to obtain preprocessed multi-source signals.

[0164] Further, in some embodiments, the feature extraction unit is specifically further configured to:

[0165] extract time-domain features, frequency-domain features, and high-order statistical features from the preprocessed target electromyography signal and the interference reference signal;

[0166] extract posture features from the preprocessed scene auxiliary signal;

[0167] perform splicing processing on the time-domain features, the frequency-domain features, the high-order statistical features, and the posture features, to generate multi-dimensional features.

[0168] Further, in some embodiments, the feature fusion unit is specifically configured to:

[0169] generate a corresponding first feature map based on the target electromyography signal, the interference reference signal, and the scene auxiliary signal;

[0170] generate a corresponding second feature map based on the multi-dimensional features;

[0171] splice the first feature map and the second feature map to generate integrated features.

[0172] Further, in some embodiments, the interference suppression unit is specifically configured to:

[0173] calculate a proximal muscle group interference component prediction value based on the integrated features;

[0174] determine a preliminary interference-removed signal based on a difference between the target electromyography signal and the proximal muscle group interference component prediction value;

[0175] perform high-frequency interference removal processing on the preliminary interference-removed signal to generate a purified pelvic floor electromyography signal.

[0176] For specific limitations of the pelvic floor electromyography signal interference removal device, refer to the limitations of the pelvic floor electromyography signal interference removal method in the foregoing, which will not be repeated here. Each module in the pelvic floor electromyography signal interference removal device described above can be realized by software, hardware, and combinations thereof, in whole or in part. The above modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0177] The pelvic floor electromyography signal interference removal device provided in the embodiment provides an explicit feature reference for the proximal muscle group interference in the target electromyography signal by synchronously collecting the interference reference signal, accurately locates the interference component, and avoids the blindness of interference differentiation caused by the lack of reference in traditional technologies. Then, by collecting the scene auxiliary signal, the dynamic law of the interference with the change in posture is captured, and scene-based judgment basis is provided for interference suppression to adapt to the interference fluctuation characteristics under different body positions. Finally, interference suppression is performed based on the interference reference signal and the scene auxiliary signal, the proximal muscle group interference in the target electromyography signal is completely removed, and the physiological characteristics of the effective signal of the pelvic floor muscle are maximized, thereby improving the purity and reliability of the output signal.

[0178] In addition, the embodiment of the present application also provides an electronic device, as shown in Figure 8 The structure schematic diagram of the electronic device related to the embodiment of the present application is shown, specifically:

[0179] The electronic device can include a processor 301 having one or more processing cores, a memory 302 having one or more computer-readable storage media, a power supply 303, and an input unit 304, etc. Those skilled in the art can understand that Figure 8 The electronic device structure shown in the figure is not a limitation on the electronic device, and can include more or fewer components than shown, or combine certain components, or different component arrangements. Among them:

[0180] The processor 301 is the control center of the electronic device, which connects various parts of the entire electronic device through various interfaces and lines, and performs various functions of the electronic device and processes data by running or executing software programs and / or modules stored in the memory 302, and calling data stored in the memory 302, thereby overall monitoring the electronic device. Optionally, the processor 301 can include one or more processing cores; preferably, the processor 301 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application program, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 301.

[0181] The memory 302 can be used to store software programs and modules, and the processor 301 executes various functions and pelvic floor muscle electrical signal interference removal methods by running the software programs and modules stored in the memory 302. The memory 302 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc.; the data storage area can store data created according to the use of the electronic device, etc. In addition, the memory 302 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state memory device. Accordingly, the memory 302 can also include a memory controller to provide access for the processor 301 to the memory 302.

[0182] The electronic device also includes a power supply 303 for powering various components, and preferably the power supply 303 can be logically connected to the processor 301 through a power management system, so as to realize the functions of managing charging, discharging, and power consumption management, etc. through the power management system. The power supply 303 can also include one or more direct current or alternating current power supplies, a recharging system, a power supply failure detection circuit, a power supply converter or inverter, a power supply state indicator, etc. any component.

[0183] The electronic device can further include an input unit 304 which can be used to receive inputted digital or character information, and to generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function controls.

[0184] Although not shown, the electronic device can further include a display unit or the like, which will not be described here. In particular in the present embodiment, the processor 301 in the electronic device will load the executable file corresponding to the process of one or more application programs into the memory 302 according to the following instructions, and run the application program stored in the memory 302 by the processor 301, thereby realizing various functions, as follows:

[0185] Collect the surface electromyogram of the pelvic floor muscle containing adjacent muscle group interference to obtain a target electromyogram; collect the physiological activity signal of the adjacent muscle group of the pelvic floor muscle to obtain an interference reference signal; collect a body posture timing signal related to the activity of the pelvic floor muscle to obtain a scene auxiliary signal; based on the interference reference signal and the scene auxiliary signal, suppress the adjacent muscle group interference in the target electromyogram, and output a purified pelvic floor electromyogram.

[0186] The specific implementation of each operation can refer to the foregoing embodiments, which will not be described here.

[0187] The embodiments of the present application can distinguish the real signal of the pelvic floor muscle from dynamic interference through the scene auxiliary signal obtained by the posture sensing module, accurately remove the adjacent muscle group interference in the target electromyogram, completely retain the effective signal characteristics of the pelvic floor muscle, improve the purity and reliability of the output signal, provide high-quality data support for accurate assessment and rehabilitation diagnosis and treatment of the pelvic floor function, and meet the clinical needs of the pelvic floor function assessment.

[0188] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by related hardware controlled by the instructions, which can be stored in a computer readable storage medium and loaded and executed by a processor.

[0189] To this end, the embodiments of the present application provide a storage medium, which stores a plurality of instructions, the instructions can be loaded by a processor to execute the steps in any pelvic floor electromyogram interference removal method provided by the embodiments of the present application. For example, the instructions can execute the following steps:

[0190] The surface electromyography signals of the pelvic floor muscles containing adjacent muscle group interference are collected to obtain target muscle electrical signals; the physiological activity signals of the adjacent muscle group of the pelvic floor muscles are collected to obtain interference reference signals; the body posture timing signals related to the activity of the pelvic floor muscles are collected to obtain scene auxiliary signals; the adjacent muscle group interference in the target muscle electrical signals is suppressed based on the interference reference signals and the scene auxiliary signals, and the purified pelvic floor muscle electrical signals are output.

[0191] The specific implementation of each operation can refer to the foregoing embodiments, which will not be described here.

[0192] The storage medium can include a read-only memory (ROM), a random access memory (RAM), a magnetic disk, an optical disk, and the like.

[0193] Due to the instructions stored in the storage medium, the steps of any of the pelvic floor muscle electrical signal interference removal methods provided in the embodiments of the present application can be executed, and thus the beneficial effects of any of the pelvic floor muscle electrical signal interference removal methods provided in the embodiments of the present application can be achieved. Details are described in the foregoing embodiments, which will not be described here.

[0194] The above describes the pelvic floor muscle repair instrument, the pelvic floor muscle electrical signal interference removal method, the device, the equipment and the storage medium in detail, and the principle and the implementation mode of the present application are described by applying specific examples in this paper. The above embodiment is only used to help understand the method and the core idea of the present application; meanwhile, for the person skilled in the art, according to the idea of the present application, the specific implementation mode and the application range will be changed, and the above description should not be understood as the limitation of the present application.

Claims

1. A pelvic floor muscle repair device, characterized in that, The pelvic floor muscle repair device includes: case; A pelvic floor muscle electromyography (EMG) acquisition module is connected to the housing and configured to acquire EMG signals from the surface of the pelvic floor muscles, including interference from adjacent muscle groups, in order to obtain the target EMG signal. The interference muscle group monitoring module is configured to collect physiological activity signals of adjacent muscle groups of the pelvic floor muscles in order to obtain interference reference signals. An attitude perception module, connected to the housing, is configured to collect temporal signals of body posture related to pelvic floor muscle activity in order to obtain scene auxiliary signals. In addition, the signal processing module is electrically connected to the pelvic floor muscle electromyography acquisition module, the interfering muscle group monitoring module, and the posture perception module; the signal processing module is configured to: based on the interference reference signal and the scene auxiliary signal, suppress the interference of adjacent muscle groups in the target electromyography signal, and output the purified pelvic floor electromyography signal.

2. The pelvic floor muscle repair device according to claim 1, characterized in that, The signal processing module is configured to extract multi-dimensional features based on the target electromyography signal, interference reference signal and scene auxiliary signal. The multi-dimensional features include time-domain features, frequency-domain features and higher-order statistical features extracted from the target electromyography signal and interference reference signal, as well as posture features extracted from the scene auxiliary signal. The target electromyographic signal, the interference reference signal, the scene auxiliary signal, and the multi-dimensional features are fused to generate integrated features; Based on the integrated features, interference from adjacent muscle groups in the target electromyographic signal is suppressed, and a purified pelvic floor electromyographic signal is output.

3. The pelvic floor muscle repair device according to claim 1, characterized in that, The pelvic floor muscle electromyography (EMG) acquisition module includes at least one pair of surface EMG electrodes, which are attached to the surface projection area of ​​the pelvic floor muscles. The surface projection area includes the area below the pubic symphysis, around the perineum, or near the sacrum and coccyx.

4. The pelvic floor muscle repair device according to claim 1, characterized in that, The interference muscle group monitoring module includes at least one pair of surface electromyography electrodes, which are attached to the body surface area of ​​the rectus abdominis and other muscle groups around the pelvic floor muscles. The interference reference signal includes the surface electromyography signal of the rectus abdominis.

5. The pelvic floor muscle repair device according to claim 1, characterized in that, The attitude sensing module includes at least one of an inertial measurement unit, an optical attitude sensor, or a magnetic sensing attitude module. The attitude sensing module is fixed to the pelvic region. The scene auxiliary signals include acceleration data and angular velocity data, as well as pitch angle timing signals and roll angle timing signals calculated based on the acceleration data and angular velocity data.

6. The pelvic floor muscle repair device according to claim 2, characterized in that, The signal processing module is configured to: perform signal preprocessing on the target electromyographic signal, the interference reference signal, and the scene auxiliary signal respectively; and extract features based on the preprocessed target electromyographic signal, interference reference signal, and scene auxiliary signal to obtain the multi-dimensional features.

7. The pelvic floor muscle repair device according to claim 6, characterized in that, The signal processing module is configured to: perform bandpass filtering on the target electromyography (EMG) signal and the interference reference signal to obtain filtered target EMG signal and interference reference signal; perform attitude calculation and linear interpolation on the scene auxiliary signal to make the sampling rate of the interpolated scene auxiliary signal consistent with the sampling rate of the target EMG signal; and perform normalization processing on the filtered target EMG signal, the interference reference signal and the interpolated scene auxiliary signal respectively to obtain preprocessed multi-source signals.

8. The pelvic floor muscle repair device according to claim 2, characterized in that, The signal processing module is configured to: generate a corresponding first feature map based on the target electromyographic signal, the interference reference signal, and the scene auxiliary signal; generate a corresponding second feature map based on the multi-dimensional features; and concatenate the first feature map and the second feature map to generate the integrated feature.

9. The pelvic floor muscle repair device according to claim 2, characterized in that, The signal processing module is configured to: calculate the predicted value of the neighboring muscle group interference component based on the integrated features; Based on the difference between the target electromyographic signal and the predicted value of the interference component of the adjacent muscle group, a preliminary de-interference signal is determined; the preliminary de-interference signal is subjected to high-frequency interference removal processing to generate the purified pelvic floor electromyographic signal.

10. A method for removing interference from pelvic floor electromyographic signals, characterized in that, include: Electromyographic signals of the pelvic floor muscles, including those affected by adjacent muscle groups, were collected to obtain the target electromyographic signal. Physiological activity signals of adjacent muscle groups of the pelvic floor muscles were collected to obtain interference reference signals; Collect temporal signals of body posture related to pelvic floor muscle activity to obtain scene auxiliary signals; Based on the interference reference signal and the scene auxiliary signal, the interference from adjacent muscle groups in the target electromyographic signal is suppressed, and the purified pelvic floor electromyographic signal is output.

11. The method for removing pelvic floor electromyographic signal interference according to claim 10, characterized in that, The step of suppressing interference from adjacent muscle groups in the target electromyographic signal based on the interference reference signal and the scene auxiliary signal, and outputting a purified pelvic floor electromyographic signal, includes: Based on the target electromyography signal, the interference reference signal, and the scene auxiliary signal, multi-dimensional features are extracted. The multi-dimensional features include time-domain features, frequency-domain features, and higher-order statistical features extracted from the target electromyography signal and the interference reference signal, as well as posture features extracted from the scene auxiliary signal. The target electromyographic signal, the interference reference signal, the scene auxiliary signal, and the multi-dimensional features are fused to generate integrated features; Based on the integrated features, interference from adjacent muscle groups in the target electromyographic signal is suppressed to obtain a purified pelvic floor electromyographic signal.

12. The method for removing pelvic floor electromyographic signal interference according to claim 11, characterized in that, The extraction of multi-dimensional features based on the target electromyographic signal, interference reference signal, and scene auxiliary signal includes: Signal preprocessing is performed on the target electromyographic signal, the interference reference signal, and the scene auxiliary signal, respectively. Feature extraction is performed based on the preprocessed target electromyographic signal, interference reference signal, and scene auxiliary signal to obtain the multi-dimensional features.

13. The method for removing pelvic floor electromyographic signal interference according to claim 12, characterized in that, The signal preprocessing for the target electromyographic signal, the interference reference signal, and the scene auxiliary signal includes: The target electromyography signal and the interference reference signal are subjected to bandpass filtering to obtain the filtered target electromyography signal and interference reference signal. The scene auxiliary signal is subjected to attitude calculation and linear interpolation to make the sampling rate of the interpolated scene auxiliary signal consistent with the sampling rate of the target electromyographic signal; The filtered target electromyography signal, the interference reference signal, and the interpolated scene auxiliary signal are normalized respectively to obtain the preprocessed multi-source signal.

14. The method for removing pelvic floor electromyographic signal interference according to claim 12, characterized in that, The feature extraction based on the preprocessed target electromyographic signal, interference reference signal, and scene auxiliary signal yields multi-dimensional features, including: Time-domain features, frequency-domain features, and higher-order statistical features are extracted from the preprocessed target electromyography signal and the interference reference signal; Extract attitude features from preprocessed scene auxiliary signals; The time-domain features, frequency-domain features, higher-order statistical features, and attitude features are concatenated to generate the multi-dimensional features.

15. The method for removing pelvic floor electromyographic signal interference according to claim 10, characterized in that, The step of fusing the target electromyographic signal, the interference reference signal, the scene auxiliary signal, and the multi-dimensional features to generate integrated features includes: Based on the target electromyographic signal, the interference reference signal, and the scene auxiliary signal, a corresponding first feature map is generated; A second feature map is generated based on the multi-dimensional features; The first feature map and the second feature map are concatenated to generate the integrated feature.

16. The method for removing pelvic floor electromyographic signal interference according to claim 10, characterized in that, Based on the integrated features, the interference from adjacent muscle groups in the target electromyographic signal is suppressed to obtain a purified pelvic floor electromyographic signal, including: Based on the integrated features, the predicted values ​​of the interference components of adjacent muscle groups are calculated; Based on the difference between the target electromyographic signal and the predicted value of the interference component of the adjacent muscle group, a preliminary interference removal signal is determined; The preliminary interference removal signal is subjected to high-frequency interference removal processing to generate the purified pelvic floor electromyography signal.

17. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the pelvic floor electromyographic signal interference removal method as described in any one of claims 10 to 16.

18. A storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 10 to 16 for the removal of pelvic floor electromyographic signal interference.

Images