An intraoperative real-time feedback system and method for sacral neuromodulation

Abstract

The present application relates to the technical field of nerve monitoring, in particular to an intraoperative real-time feedback system and method for sacral nerve stimulation, the system comprising: a biological motion sensing unit for real-time monitoring of the body surface contraction movement of the perineum and perianal area and foot movement through a sensor and generating corresponding multi-modal sensing signals; a processing unit for receiving the multi-modal sensing signals and performing synchronization and feature extraction, inputting the extracted features into a pre-trained machine learning model to obtain a judgment result representing the effect of sacral nerve electrode implantation target point and the accuracy of the target point position; a feedback unit for receiving the judgment result and converting it into feedback information for unified display. The present application realizes objective quantitative evaluation through multi-modal signal synchronous perception and fusion, as well as integrated machine learning model, provides intelligent and predictive feedback beyond simple threshold judgment, and realizes precise guidance of electrode implantation.

Description

Technical Field

[0001] This invention relates to the field of neuromonitoring technology, and more specifically, to an intraoperative real-time feedback system and method for sacral nerve stimulation. Background Technology

[0002] Sacral nerve modulation is an effective treatment for various lower urinary tract dysfunctions. It mainly involves surgically implanting electrodes into the sacral nerve and conducting in vitro testing. Once the therapeutic outcome is satisfactory, a permanent stimulator is surgically implanted.

[0003] The surgical procedure involves the patient lying prone for X-ray fluoroscopy to locate the S3 foramen. The lower abdomen and calves are slightly elevated to flex the knees and ensure the toes are suspended. During the procedure, electrodes are implanted at the S3 foramen, and motor and sensory responses are tested at the electrode target. S3 is the preferred target for sacral nerve modulation, and clinically, an effective voltage of no more than 2V is generally preferred. During the procedure, the surgeon adjusts the intensity parameters of the external stimulator, visually observes for bellows-like movements in the perianal and perineal areas, verbally inquires about perineal stimulation, and the surgical assistant observes plantar flexion of the big toe. The presence of these signs indicates a good electrode target; conversely, the absence of these signs indicates a poor target and requires readjustment.

[0004] Compared to sensory responses alone during surgery, motor responses (contraction of the levator ani muscle and plantar flexion of the big toe) are a better predictor of successful sacral nerve neuromodulation therapy. Figure 6 As shown in Table 1.

[0005] Table 1. S2-S4 Responses of Sacral Nerves to Electrical Stimulation

[0006] The above steps clearly demonstrate numerous limitations: (1) The feedback on surgical target points relies solely on the surgeon's visual observation of the perianal area, perineum, and soles of the feet; (2) Verbally ask the patient about perineal sensation, local anesthesia and the presence of tension during the operation. Many patients are even unable to give a clear answer as to whether there is stimulation in the perineal area. (3) During the process, it is impossible to quantify the relevant motor and sensory indicators simultaneously. Most of the assessments are qualitative, and different surgeons have different standards for evaluating and judging the phenomena. Summary of the Invention

[0007] In view of this, the present invention proposes an intraoperative real-time feedback system and method for sacral nerve stimulation surgery, aiming to solve the problems of traditional sacral nerve stimulation surgery relying on visual observation of perineal, perianal and foot movements, which cannot be synchronously quantified and is inconvenient to operate.

[0008] This invention proposes an intraoperative real-time feedback system for sacral nerve stimulation surgery, comprising: The bio-motion sensing unit is used to non-invasively monitor the body surface contraction movements and foot movements in the perineum and perianal area in real time through sensors, sense the mechanical motion response triggered by sacral nerve stimulation, and generate corresponding multimodal sensing signals. The processing unit is communicatively connected to the biomotion sensing unit and is used to receive the multimodal sensing signals and perform synchronization and feature extraction. The extracted features are input into a pre-trained machine learning model to obtain a judgment result characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location. The feedback unit is communicatively connected to the processing unit and is used to receive the judgment result and convert it into feedback information for unified display.

[0009] Preferably, the bio-motion sensing unit includes: The perineal and perianal sensing module is used to sense the surface contraction movements of the perineal and perianal measurement areas by measuring at least one of the following surface acquisition methods: pressure, electromyography, optical deformation, and inertial parameters. A foot motion sensing module for sensing foot motion by measuring at least one of pressure distribution, inertial parameters, and visual markers.

[0010] Preferably, the processing unit includes: The signal processing and fusion subunit is used to synchronize and extract features from the sensor signals acquired by the perineal and perianal sensing module and the foot motion sensing module. The neural response judgment subunit is used to input the features extracted by the signal processing and fusion subunit into a pre-trained machine learning model, infer and output the perineal, perianal and foot movement results, thereby generating judgment results that characterize the effect of sacral nerve electrode implantation target and the accuracy of target location.

[0011] Preferably, the perineal and perianal sensing module acquires the pressure signal generated in the perineal and perianal areas when the sacral nerve is electrically stimulated through a body surface thin-film pressure sensor array; the foot motion sensing module acquires the motion signal generated in the foot when the sacral nerve is electrically stimulated through a miniature inertial measurement unit.

[0012] Preferably, the signal processing and fusion subunit calculates the normalized contraction intensity feature based on the multimodal sensing signals acquired by the biomotion sensing unit. Normalized motion angle characteristics and temporal synergy features ; ; ; ; in, This is the normalization function; For amplitude feature extraction; This is a perineal / perianal pressure signal; For angle feature extraction; S represents the plantar flexion angle of the foot; S represents the cross-correlation peak value, coherence coefficient, or distance metric based on dynamic time warping between the two signals. The neural response judgment subunit incorporates a pre-trained lightweight edge convolutional neural network model and integrates decision functions. The classification labels characterizing the implantation effect of sacral nerve electrodes and the target confidence scores characterizing the accuracy of target location are output as the judgment results. ; ; ; ; in, Target confidence score; Category tags; For function parameters; These are the weighting coefficients for each feature; To determine the threshold, the score is compared with... The preset comparison relationship is used to output category labels.

[0013] Preferably, the perineal and perianal sensing module acquires motion signals generated in the perineal and perianal regions when the sacral nerve is electrically stimulated via a miniature inertial measurement unit; the foot motion sensing module acquires motion signals generated in the foot when the sacral nerve is electrically stimulated via a miniature inertial measurement unit.

[0014] Preferably, the signal processing and fusion subunit calculates the node motion feature vector based on the multimodal sensing signals acquired by the bio-motion sensing unit, as follows: For each node i, extract its motion feature vector. ; ; Where i = (A, B), node A is the perineum, and node B is the foot; Characteristic of angular velocity It exhibits linear acceleration characteristics; This refers to attitude angle features; The motion relationship between two nodes is quantified using the pattern contrast function D and the correlation metric function R. ; ; Where d represents the pattern difference degree; r represents the pattern correlation degree; The neural response judgment subunit incorporates a pre-trained lightweight edge convolutional neural network model and integrates d, r, and the original features. As input, the sacral nerve stimulation segment, i.e., the classification label and the corresponding confidence score, are identified through the classification function C, as follows: ; ; in, The segment for sacral nerve stimulation is the classification label; Here, C represents the classifier parameters, where C is a support vector machine, random forest, or edge CNN; S2, S3, and S4 are the neural apertures, and A... typ This is an atypical reaction; NR indicates no reaction.

[0015] Preferably, the perineal and perianal sensing module acquires electromyographic signals generated in the perineal and perianal regions when the sacral nerve is electrically stimulated via surface electromyography electrodes; the foot motion sensing module acquires visual motion signals generated in the foot when the sacral nerve is electrically stimulated via vision-based motion capture.

[0016] Preferably, the signal processing and fusion subunit transforms the temporally continuous electromyographic signals and spatially discrete visual motion signals into a feature set for joint evaluation based on the multimodal sensing signals acquired by the bio-motion sensing unit, as follows: Electromyographic activation event feature set ; Visual motion event feature set ; in, Peak amplitude; This refers to the potential latency period; Duration of the outbreak; Frequency response; To estimate the fitting coefficients; For motion smoothness; Peak speed; The likelihood assessment function evaluates the probability that the observed feature set {E,V} under the current stimulus is generated by an effective neural response, as follows: ; Where L is the likelihood value, and the higher the value, the more realistic and effective the reaction is; H1 represents the function parameters; H1 represents the effective neural response. The neural response judgment subunit incorporates a support vector machine model. It takes E, V, and L as input and outputs the sacral nerve stimulation segment (classification label) and corresponding confidence score through a decision function f, as detailed below: FB= ; Where FB is the feedback unit function; The segment stimulating the sacral nerve is the classification label; C in This represents the confidence level of this judgment.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Achieve objective quantitative assessment: Transform the traditional visual qualitative observation into quantitative measurement based on sensor signals, providing an objective and unified quantitative standard for electrode implantation target assessment and reducing human error.

[0018] (2) Realize multimodal signal synchronous perception and fusion: The system can synchronously capture the motion response of several key parts of the perineum / anal area and the sole of the foot, and accurately align them on the time axis to achieve synchronization, which facilitates the surgeon to perform correlation analysis and comprehensive judgment, and overcomes the problem of asynchronous observation in traditional methods.

[0019] (3) Improved surgical efficiency and convenience: The surgeon does not need to frequently visually inspect the perineum / perianal area or bend over to examine the soles of the feet, nor does he / she need to rely entirely on the patient's verbal description. All judgment results are centrally displayed through the feedback unit, which simplifies the surgical procedure, reduces the workload of assistants, and allows the surgeon to focus more on electrode implantation and parameter adjustment.

[0020] (4) Introducing intelligent judgment to enhance decision support: By integrating machine learning models, the system can learn the complex mapping relationship between features and ideal targets in a large amount of clinical data, providing intelligent and predictive feedback that goes beyond simple threshold judgment, assisting surgeons to make more accurate decisions, especially helpful in handling complex cases with atypical reactions or weak signals.

[0021] On the other hand, this application also provides an intraoperative real-time feedback method for sacral nerve stimulation surgery, including: Step 1: Monitor the patient's perineal and perianal areas for real-time contraction movements and foot movements, and generate corresponding multimodal sensor signals; Step 2: Receive the multimodal sensing signal from Step 1, perform synchronization and feature extraction, input the extracted features into the pre-trained machine learning model, and obtain the judgment results characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location. Step 3: Receive the judgment result from Step 2 and convert it into feedback information for unified display.

[0022] It is understood that the intraoperative real-time feedback method for sacral nerve stimulation in this embodiment has the same beneficial effects as an intraoperative real-time feedback system for sacral nerve stimulation, and will not be described in detail here. Attached Figure Description

[0023] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A schematic diagram of the perineal and perianal sensing module provided in an embodiment of the present invention; Figure 2 A schematic diagram illustrating the application of the perineal and perianal sensing module provided in an embodiment of the present invention; Figure 3 A schematic diagram of perineum and pressure sensing / inertial measurement provided for an embodiment of the present invention; Figure 4 This is a schematic diagram of electromyography sensing and visual capture provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the perineal and perianal measurement area provided in an embodiment of the present invention; Figure 6 A schematic diagram of the S1 to S4 sacral nerves provided in an embodiment of the present invention.

[0024] In the diagram: 101, perineal and perianal sensing module; 1011, opening area; 102, foot motion sensing module; 1021, optical marker point. Detailed Implementation

[0025] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] Example 1 This embodiment provides an intraoperative real-time feedback system for sacral nerve stimulation surgery, including a biomotion sensing unit, a processing unit, and a feedback unit.

[0027] The bio-motion sensing unit is used for non-invasive, synchronous, and real-time monitoring of surface contraction movements and foot movements in the perineum and perianal region. It senses the mechanical movement response induced by sacral nerve stimulation and generates corresponding multimodal sensing signals. The generation and processing of multimodal sensing signals can integrate information from different dimensions, improving the accuracy and application value of monitoring.

[0028] Combination Figures 1 to 5 As shown, the bio-motion sensing unit includes a perineal and perianal sensing module 101 and a foot motion sensing module 102. The perineal and perianal sensing module 101 is used to non-invasively sense surface contraction movements in the perineal and perianal measurement areas by measuring at least one of the following surface acquisition methods: pressure, electromyography, optical deformation, and inertial parameters. The perineal and perianal sensing module 101 has an opening region 1011, which can be, for example, a surface thin-film pressure sensor array, surface electromyography electrodes, optical sensors, or a miniature inertial measurement unit. The foot motion sensing module 102 is used to sense foot movements by measuring at least one of the following: pressure distribution, inertial parameters, and visual markers. For example, it can be a pressure distribution sensing insole, a strap integrating an inertial measurement unit, or visual markers attached to the toes and a matching visual system. The bio-motion sensing unit utilizes multimodal sensor fusion technology to achieve accurate perception of the motion state of the monitored area, providing objective data support for the assessment of motor implantation target points during sacral nerve stimulation surgery and reducing human error.

[0029] The processing unit is communicatively connected to the biomotion sensing unit to receive multimodal sensing signals and perform synchronization and feature extraction. The extracted features are then input into a pre-trained machine learning model to obtain a judgment result characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location.

[0030] The processing unit extracts features from these heterogeneous sensor signals, transforming the raw data into discriminative feature vectors. These features are then input into a pre-trained machine learning model. Based on pre-learned patterns, the model comprehensively assesses the target effectiveness (e.g., stimulation effectiveness) and target location accuracy of sacral nerve electrode implantation, outputting quantitative or qualitative evaluation results. This provides an objective and real-time auxiliary evaluation method for sacral nerve electrode implantation surgery, helping to improve surgical success rates and patient treatment outcomes.

[0031] The processing unit includes a signal processing and fusion subunit and a neural response judgment subunit. The signal processing and fusion subunit is used to synchronize, filter, denoise, perform Fourier transform, amplify, and extract features from the sensor signals acquired by the perineal and perianal sensing module 101 and the foot motion sensing module 102. The extracted features include, but are not limited to, motion amplitude, frequency, waveform, duration, and foot motion angle. The neural response judgment subunit is used to input the features extracted by the signal processing and fusion subunit into a pre-trained machine learning model, infer and output the perineal, perianal, and foot motion results, i.e., to generate judgment results characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location. The neural response judgment subunit can directly and quickly output classification or regression judgment results based on real-time input feature data.

[0032] In this embodiment, the training data for the machine learning model comes from multimodal sensor data synchronously recorded during successful sacral nerve surgeries and electrode target location labels that have been clinically validated as ideal. The training goal is to enable the model to learn to identify feature patterns corresponding to the ideal target from real-time monitored multimodal sensor data. It is understood that the specific architecture of the model and the training method can be selected and optimized according to the actual data conditions, which does not deviate from the core idea of ​​this invention.

[0033] The feedback unit communicates with the processing unit to receive judgment results and convert them into feedback information for unified display, allowing the surgeon to understand intuitively whether the electrode was implanted at the appropriate target or missed the target. The feedback unit includes a graphical user interface (GUI) and a prompting device. The GUI displays quantitative or graphical feedback information on perineal and anal contraction and foot movement simultaneously, such as quantitative data and dynamic curves. The prompting device includes audiovisual and / or mechanical prompting devices, which trigger different audiovisual and / or tactile prompting signals based on the judgment results output by the processing unit.

[0034] Combination Figure 1 , Figure 2 As shown, in this embodiment, the perineal and perianal sensing module 101 acquires the pressure signals generated in the perineal and perianal regions when the sacral nerve is electrically stimulated through a surface thin-film pressure sensor array. The surface thin-film pressure sensor array is preferably a disposable medical thin-film pressure sensor array, which is placed in close contact with the patient's perineum and perianal surface. When the sacral nerve is electrically stimulated, causing contraction of pelvic floor muscles such as the levator ani muscle, the surface thin-film pressure sensor array detects the change in pressure distribution and outputs the corresponding pressure signal characteristics.

[0035] Combination Figure 3As shown, the foot motion sensing module 102 acquires motion signals generated in the foot when the sacral nerve is electrically stimulated via a miniature inertial measurement unit. For example, the miniature inertial measurement unit is integrated onto a soft bandage, which is then fixed to the patient's foot. The miniature inertial measurement unit can measure in real time the angular velocity, acceleration, and angle changes of the foot (especially the big toe) during plantar flexion under electrical stimulation.

[0036] The signal processing and fusion subunit first timestamps and synchronizes the two signals from the perineal and perianal sensing module 101 and the foot motion sensing module 102. It then filters the pressure signal to eliminate aliasing from respiration, heart rate, etc., and calculates the average pressure change amplitude of the entire array or a specific area as the "contraction intensity." The subunit performs attitude calculation on the signals from the miniature inertial measurement unit to obtain the real-time plantar flexion angular velocity, acceleration, and angle change of the big toe.

[0037] Specifically, the normalized shrinkage strength characteristic is calculated. Normalized motion angle characteristics and temporal synergy features The calculation formula is as follows: ; ; ; in, This is the normalization function; For amplitude feature extraction; This is a perineal / perianal pressure signal; For angle feature extraction; S represents the plantar flexion angle of the foot; S represents the cross-correlation peak value, coherence coefficient, or distance metric based on dynamic time warping between the two signals.

[0038] The neural response judgment subunit incorporates a pre-trained lightweight edge convolutional neural network model. This model uses synchronized contraction intensity time series and plantar flexion angle, velocity, and acceleration time series as input features, and integrates them through a decision function. The output is a classification label characterizing the effect of sacral nerve electrode implantation and a target confidence score characterizing the accuracy of target location, which are used as the judgment result. ; ; ; ; in, Target confidence score; For classification labels, such as: typical S3 reaction, atypical reaction, no reaction, S4 reaction, S2 reaction; For function parameters; These are the weighting coefficients for each feature; To determine the threshold, the score is compared with... The preset comparison relationship is used to output category labels.

[0039] The graphical user interface, for example, runs on a monitor next to the operating room or in an AR or VR display unit. The interface displays the pressure gradient map of the thin-film pressure sensor array in real time, and the curves showing the changes in plantar flexion angle, angular velocity, and acceleration over time. The interface prominently displays the real-time contraction intensity n1% and the real-time plantar flexion angle n2° in numerical form, and highlights "Target confidence: 0.95 (typical S3 response)," which is the judgment result.

[0040] The prompting device triggers different audiovisual and / or tactile prompting signals based on the target confidence score in the judgment result. For example, when the target confidence score is higher than 0.80, the system emits a short, soft prompting sound while the interface flashes. When the target confidence score is lower than 0.6, it emits a continuous low-frequency beep and the interface displays red, prompting the operator that the electrode position may need to be adjusted.

[0041] Example 2 Combination Figure 2 and Figure 3 As shown, in this embodiment, based on Embodiment 1, the perineal and perianal sensing module 101 acquires motion signals generated in the perineal and perianal regions when the sacral nerve is electrically stimulated through a miniature inertial measurement unit. For example, a surface-mounted miniature inertial measurement device is used, which is placed on the patient's perineum and perianal surface. When the sacral nerve is electrically stimulated, causing the pelvic floor muscles such as the levator ani to contract, the miniature inertial measurement unit can capture the micro-motion characteristics of the muscles and detect the minute linear acceleration and angular velocity changes generated by the muscle contraction in this area.

[0042] The foot motion sensing module 102 acquires motion signals generated in the foot when the sacral nerve is electrically stimulated via a miniature inertial measurement unit. For example, a surface-mounted miniature inertial measurement device is attached to the patient's foot. The miniature inertial measurement unit can measure in real time the angular velocity, acceleration, and angle changes of the foot (especially the big toe) during plantar flexion under electrical stimulation.

[0043] The signal processing and fusion subunit receives inertial motion parameters from the motion signals of the perineal and perianal sensing unit and the foot motion sensing module 102, and performs time stamp synchronization and attitude calculation on the two signals to obtain the motion parameters of the perineum / perianal area, the real-time plantar flexion angular velocity, acceleration and angle change of the big toe.

[0044] Specifically, the signal processing and fusion subunit calculates the node motion feature vector based on the multimodal sensing signals acquired by the bio-motion sensing unit, using the following formula: For each node i, extract its motion feature vector. ; ; Where i = (A, B), node A is the perineum, and node B is the foot; Characteristic of angular velocity It exhibits linear acceleration characteristics; This refers to attitude angle features; The motion relationship between two nodes is quantified using the pattern contrast function D and the correlation metric function R. ; ; Where d represents the pattern difference degree and r represents the pattern correlation degree.

[0045] The neural response judgment subunit incorporates a pre-trained lightweight edge convolutional neural network model. This model uses synchronized muscle motion inertial sequences and plantar flexion angle, velocity, and acceleration sequences as input features, i.e., d, r, and the original features. As input, the sacral nerve stimulation segment, i.e., the classification label and the corresponding confidence score, are identified through the classification function C, as follows: ; ; in, The segment for sacral nerve stimulation is the classification label; Here, C represents the classifier parameters, where C is a support vector machine, random forest, or edge CNN; S2, S3, and S4 are neural apertures, representing the S2 response, typical S3 response, and S4 response, respectively; A typ This is an atypical reaction; NR indicates no reaction.

[0046] The graphical user interface runs on a monitor next to the operating room or in an AR or VR display unit. The interface displays a real-time illustration of perineal / anal contraction movements, and displays the curves of plantar flexion angle, angular velocity, and acceleration changing over time. It highlights "Target confidence: 0.95 (typical S3 response)", which is the judgment result.

[0047] The prompting device triggers different audiovisual and / or tactile prompting signals based on the target confidence score in the judgment result. For example, when the target confidence score is higher than 0.80, the system emits a short, soft prompting sound while the interface flashes. When the target confidence score is lower than 0.6, it emits a continuous low-frequency beep and the interface displays red, prompting the operator that the electrode position may need to be adjusted.

[0048] Example 3 Combination Figure 2 and Figure 4 As shown, in this embodiment, based on Embodiment 1, the perineal and perianal sensing module 101 acquires electromyographic signals generated in the perineal and perianal regions when the sacral nerve is electrically stimulated via surface electromyographic electrodes. For example, one or two pairs of surface electromyographic electrodes are attached to the body surface on both sides of the perianal / perineal area to collect electromyographic signals generated when sacral nerve stimulation induces pelvic floor muscle contraction.

[0049] The foot motion sensing module 102 acquires visual motion signals generated in the foot when the sacral nerve is electrically stimulated through vision-based motion capture. For example, optical markers 1021 are attached to the patient's big toe and heel, and captured by a small infrared camera facing the sole of the foot. The position of the markers is calculated in real time through image processing algorithms, thereby obtaining the trajectory and angle of plantar flexion movement.

[0050] The signal processing and fusion subunit simultaneously acquires electromyographic (EMG) signals and visual-motor data. The EMG signals are rectified and smoothed, and their integrated EMG values ​​are extracted as an indicator of activity intensity. Real-time plantar flexion angles are calculated from the visual data.

[0051] Specifically, the signal processing and fusion subunit transforms the temporally continuous electromyographic signals and spatially discrete visual motion signals acquired by the bio-motion sensing unit into a feature set for joint evaluation, calculated as follows: Electromyographic activation event feature set ; Visual motion event feature set ; in, Peak amplitude; This refers to the potential latency period; Duration of the outbreak; Frequency response; To estimate the fitting coefficients; For motion smoothness; Peak speed; The likelihood assessment function evaluates the probability that the observed feature set {E,V} under the current stimulus is generated by an effective neural response, as follows: ; Where L is the likelihood value, and the higher the value, the more realistic and effective the reaction is; H1 represents the function parameter; H1 represents the effective neural response.

[0052] The neural response judgment subunit incorporates a support vector machine model. Input features include electromyographic activity intensity, delay time of electromyographic bursts, peak plantar flexion angle, and correlation coefficients between parameters. Specifically, E, V, and L are taken as input, and the decision function f outputs the sacral nerve stimulation segment, i.e., the classification label and corresponding confidence score, as detailed below: FB= ; Where FB is the feedback unit function; The segment of sacral nerve stimulation is categorized by label, for example: S2 response, typical S3 response, S4 response, no response, atypical response; C in This represents the confidence level of this judgment.

[0053] The graphical user interface displays the original waveform and processed envelope of the electromyographic signal in columns, as well as the dynamic simplified skeletal animation of foot movement. The model output judgment result is displayed in a prominent position, or the judgment result is broadcast through the speech synthesis module.

[0054] Example 4 This embodiment provides an intraoperative real-time feedback method for sacral nerve stimulation surgery, applied to an intraoperative real-time feedback system for sacral nerve stimulation surgery as described in Embodiment 1, Embodiment 2, or Embodiment 3, comprising: Step 1: Monitor the patient's perineal and perianal areas for real-time contraction movements and foot movements, and generate corresponding multimodal sensor signals; Step 2: Receive the multimodal sensing signal from Step 1, perform synchronization and feature extraction, input the extracted features into the pre-trained machine learning model, and obtain the judgment results characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location. Step 3: Receive the judgment result from Step 2 and convert it into feedback information for unified display.

[0055] It is understood that the intraoperative real-time feedback method for sacral nerve stimulation in this embodiment has the same beneficial effects as an intraoperative real-time feedback system for sacral nerve stimulation, and will not be described in detail here.

[0056] Embodiments of this application may be provided as methods, systems, or computer program goods. Therefore, this application may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application may take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0057] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0058] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0059] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A real-time intraoperative feedback system for sacral nerve stimulation surgery, characterized in that, include: The bio-motion sensing unit is used to non-invasively monitor the body surface contraction movements and foot movements in the perineum and perianal area in real time through sensors, sense the mechanical motion response triggered by sacral nerve stimulation, and generate corresponding multimodal sensing signals. The processing unit is communicatively connected to the biomotion sensing unit and is used to receive the multimodal sensing signals and perform synchronization and feature extraction. The extracted features are input into a pre-trained machine learning model to obtain a judgment result characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location. The feedback unit is communicatively connected to the processing unit and is used to receive the judgment result and convert it into feedback information for unified display.

2. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 1, characterized in that, The bio-motion sensing unit includes: The perineal and perianal sensing module (101) is used to sense the surface contraction motion of the perineal and perianal measurement areas by measuring at least one of the following surface acquisition methods: pressure, electromyography, optical deformation and inertial parameters. A foot motion sensing module (102) is used to sense foot motion by measuring at least one of pressure distribution, inertial parameters and visual markers.

3. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 2, characterized in that, The processing unit includes: The signal processing and fusion subunit is used to synchronize and extract features from the sensor signals acquired by the perineal and perianal sensing module (101) and the foot motion sensing module (102); The neural response judgment subunit is used to input the features extracted by the signal processing and fusion subunit into a pre-trained machine learning model, infer and output the perineal, perianal and foot movement results, thereby generating judgment results that characterize the effect of sacral nerve electrode implantation target and the accuracy of target location.

4. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 3, characterized in that, The perineal and perianal sensing module (101) acquires the pressure signal generated in the perineal and perianal areas when the sacral nerve is electrically stimulated through a body surface thin film pressure sensor array; the foot motion sensing module (102) acquires the motion signal generated in the foot when the sacral nerve is electrically stimulated through a miniature inertial measurement unit.

5. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 4, characterized in that, The signal processing and fusion subunit calculates the normalized contraction intensity feature based on the multimodal sensing signals acquired by the biomotion sensing unit. Normalized motion angle characteristics and temporal synergy features ; ； ； ； in, This is the normalization function; For amplitude feature extraction; This is a perineal / perianal pressure signal; For angle feature extraction; S represents the plantar flexion angle of the foot; S represents the cross-correlation peak value, coherence coefficient, or distance metric based on dynamic time warping between the two signals. The neural response judgment subunit incorporates a pre-trained lightweight edge convolutional neural network model and integrates decision functions. The classification labels characterizing the implantation effect of sacral nerve electrodes and the target confidence scores characterizing the accuracy of target location are output as the judgment results. ； ； ； ； in, Target confidence score; Category tags; For function parameters; These are the weighting coefficients for each feature; To determine the threshold, the score is compared with... The preset comparison relationship is used to output category labels.

6. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 3, characterized in that, The perineal and perianal sensing module (101) acquires motion signals generated in the perineal and perianal areas when the sacral nerve is electrically stimulated through a miniature inertial measurement unit; the foot motion sensing module (102) acquires motion signals generated in the foot when the sacral nerve is electrically stimulated through a miniature inertial measurement unit.

7. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 6, characterized in that, The signal processing and fusion subunit calculates the node motion feature vector based on the multimodal sensing signals acquired by the bio-motion sensing unit, as follows: For each node i, extract its motion feature vector. ; ； Where i = (A, B), node A is the perineum, and node B is the foot; Characteristic of angular velocity It exhibits linear acceleration characteristics; This refers to attitude angle features; The motion relationship between two nodes is quantified using the pattern contrast function D and the correlation metric function R. ； ； Where d represents the pattern difference degree; r represents the pattern correlation degree; The neural response judgment subunit incorporates a pre-trained lightweight edge convolutional neural network model and integrates d, r, and the original features. As input, the sacral nerve stimulation segment, i.e., the classification label and the corresponding confidence score, are identified through the classification function C, as follows: ； ； in, The segment for sacral nerve stimulation is the classification label; Here, C represents the classifier parameters, where C is a support vector machine, random forest, or edge CNN; S2, S3, and S4 are the neural apertures, and A... typ This is an atypical reaction; NR indicates no reaction.

8. The intraoperative real-time feedback system for sacral nerve stimulation according to claim 3, characterized in that, The perineal and perianal sensing module (101) acquires electromyographic signals generated in the perineal and perianal areas when the sacral nerve is electrically stimulated through surface electromyographic electrodes; the foot motion sensing module (102) acquires visual motion signals generated in the foot when the sacral nerve is electrically stimulated through vision-based motion capture.

9. A real-time intraoperative feedback system for sacral nerve stimulation according to claim 8, characterized in that, The signal processing and fusion subunit, based on the multimodal sensing signals acquired by the biomotion sensing unit, transforms the temporally continuous electromyographic signals and spatially discrete visual motion signals into a feature set for joint evaluation, as follows: Electromyographic activation event feature set ; Visual motion event feature set ; in, Peak amplitude; This refers to the potential latency period; Duration of the outbreak; Frequency response; To estimate the fitting coefficients; For motion smoothness; Peak speed; The likelihood assessment function evaluates the probability that the observed feature set {E,V} under the current stimulus is generated by an effective neural response, as follows: ； Where L is the likelihood value, and the higher the value, the more realistic and effective the reaction is; H1 represents the function parameters; H1 represents the effective neural response. The neural response judgment subunit incorporates a support vector machine model. It takes E, V, and L as input and outputs the sacral nerve stimulation segment (classification label) and corresponding confidence score through a decision function f, as detailed below: FB= ; Where FB is the feedback unit function; The segment stimulating the sacral nerve is the classification label; C in This represents the confidence level of this judgment.

10. A real-time intraoperative feedback method for sacral nerve stimulation surgery, characterized in that, The system is applied to an intraoperative real-time feedback system for sacral nerve stimulation as described in any one of claims 1-9, comprising: Step 1: Monitor the patient's perineal and perianal areas for real-time contraction movements and foot movements, and generate corresponding multimodal sensor signals; Step 2: Receive the multimodal sensing signal from Step 1, perform synchronization and feature extraction, input the extracted features into the pre-trained machine learning model, and obtain the judgment results characterizing the effect of sacral nerve electrode implantation target point and the accuracy of target point location. Step 3: Receive the judgment result from Step 2 and convert it into feedback information for unified display.

Images