Intelligent diagnosis and treatment system for maxillary muscle dysfunction

Abstract

The invention discloses an intelligent diagnosis and treatment system for stomatognathic muscle dysfunction, and the system comprises a flexible wearable sensing module which comprises an electromyographic signal collection unit, a strain sensing unit, and a pressure sensing unit; the data preprocessing and communication module comprises a data preprocessing sub-module and a wireless communication sub-module; and the in-vitro intelligent analysis module is wirelessly connected with the data preprocessing and communication module and comprises at least one of the following items: an application program arranged on a mobile terminal, a cloud end or a doctor end platform, and a machine learning algorithm module is arranged in the in-vitro intelligent analysis module and is used for displaying, analyzing, identifying, monitoring, evaluating or diagnosing the information. According to the scheme, the method is used for dynamically monitoring the stomatognathic muscle function state in real time, a multi-mode physiological signal analysis algorithm is combined, a non-intrusive and comfortable wearing physiological data acquisition mode is provided, dynamic visualization of the facial muscle function can be achieved, the early discovery rate of stomatognathic muscle dysfunction is increased through AI model recognition and early warning and other means, and the method is suitable for popularization and application. Doctors are assisted in objective diagnosis.

Description

Technical Field

[0001] The present invention relates to the field of oral medical technology, and in particular to an intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction. Background Art

[0002] Orofacial myofunctional disorders (OMDs) are a comprehensive group of disorders affecting maxillofacial muscle function, commonly affecting basic oral behaviors such as chewing, swallowing, speech, and breathing. They have significant clinical implications in orthodontics, occlusal reconstruction, sleep apnea, and speech delay. However, current clinical diagnosis relies heavily on subjective judgment and static examinations, lacking dynamic, objective, and physiological data.

[0003] With the development of flexible electronics and wearable device technologies, myoelectric, biostrain, pressure and other sensors have been widely used in clinical health monitoring. However, there is no device that can integrate facial multimodal signal acquisition and real-time analysis for dynamic assessment of the functional status of oral and mandibular muscles. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed to provide an intelligent diagnosis and treatment system for oral and mandibular muscle dysfunction that overcomes the above problems or at least partially solves the above problems.

[0005] The present invention provides an intelligent diagnosis and treatment system for oromandibular muscle dysfunction, the intelligent diagnosis and treatment system for oromandibular muscle dysfunction comprising:

[0006] The flexible wearable sensing module includes an electromyographic signal acquisition unit, a strain sensing unit, and a pressure sensing unit. The electromyographic signal acquisition unit is used to acquire electromyographic signals of the oral and maxillofacial muscles, the strain sensing unit is used to acquire facial dynamic displacement signals, and the pressure sensing unit is used to acquire maxillofacial pressure distribution signals.

[0007] The data preprocessing and communication module includes a data preprocessing submodule and a wireless communication submodule. The data preprocessing submodule is used to preprocess the signal data obtained from the flexible wearable sensor module, and the wireless communication submodule outputs the preprocessed data wirelessly.

[0008] The off-body intelligent analysis module is wirelessly connected to the data preprocessing and communication module and includes at least one of the following: an application program provided on a mobile terminal, a cloud platform or a doctor-side platform, and a built-in machine learning algorithm module for displaying, analyzing, identifying, monitoring, evaluating or assisting clinical diagnosis of information.

[0009] In some embodiments, the electromyographic signal acquisition unit is used to monitor the surface electromyographic signals of the oral and maxillofacial muscles to reflect the activation pattern of the muscles during chewing, swallowing, or speaking. The electromyographic signal acquisition unit uses flexible conductive materials as electrodes that can be attached to the skin, and the electrodes are configured to: formulate a layout plan based on the anatomy of the maxillofacial muscles and signal source modeling, and introduce a data-driven algorithm to optimize the electrode arrangement;

[0010] The strain sensing unit or pressure sensing unit is used to sense facial epidermal deformation or mechanical stress changes caused by opening and closing the mouth, chewing or tongue movement, and includes a sensing layer constructed of graphene, Mxene, carbon nanotubes, super-folded carbon materials, metal-based conductive materials, liquid metal and / or PDMS materials, and the sensing layer adopts an island bridge structure or a serpentine interconnection structure.

[0011] In some embodiments, the flexible wearable sensing module and the data preprocessing and wireless communication module are arranged on a flexible material body; the flexible material body is worn on the human head, craniofacial area and / or neck; or, the flexible material body is attached to the surface of the oral and maxillofacial muscles.

[0012] In some embodiments, the electromyographic signal acquisition unit, the strain sensing unit, and the pressure sensing unit are integrated on a flexible printed circuit board.

[0013] In some embodiments, the data preprocessing submodule is disposed on a flexible circuit board and includes a micro MCU and at least one of the following modules: a multi-channel digital-to-analog converter, a filter, a Fourier transform module, a data compression module, or a data cache module.

[0014] In some embodiments, the wireless communication submodule is configured to do at least one of the following:

[0015] Support BLE or WiFi communication protocols, taking into account both short-range fast synchronization and medium- and long-range stable communication;

[0016] Integrated antenna layout optimization algorithm improves signal penetration and stability, adapting to complex wearing scenarios around the mouth and jaw;

[0017] Design a lightweight custom communication protocol stack that supports data packet packaging, timestamp marking, packet loss detection and / or breakpoint resuming;

[0018] Optimize BLE multi-channel data exchange rate to improve real-time performance and system energy efficiency;

[0019] Implement a standard data docking interface with mobile terminal applications, cloud or doctor-side platforms, and support two-way interaction.

[0020] In some embodiments, the data preprocessing and communication module further includes a battery module, and the battery module is configured as at least one of the following:

[0021] Adopt high-density flexible lithium polymer battery to meet the needs of long-term work;

[0022] Integrated Type-C rechargeable module, supporting intelligent fast charging, battery protection and overvoltage and overcurrent control;

[0023] Supports wireless charging expansion to achieve a more flexible user experience;

[0024] Utilize the multi-level sleep mode of the data pre-processing submodule to achieve automatic switching during non-collection periods;

[0025] Set up a dynamic wake-up mechanism for the sensor unit to extend standby battery life;

[0026] The system power consumption control target is maintained at less than 15mW to ensure that the energy burden and heat accumulation of daily wear are minimized.

[0027] In some embodiments, the data preprocessing and communication module is configured to do at least one of the following:

[0028] Establish a unified signal synchronization sampling mechanism to ensure that the discharge signals of facial oromandibular muscles, facial dynamic displacement signals, and maxillofacial pressure distribution signals have a unified time base;

[0029] Achieve denoising, normalization or feature enhancement of each channel signal to improve the consistency and interpretability of algorithm input data;

[0030] Build a self-calibration algorithm that can adapt to different wearing positions to reduce the risk of misjudgment caused by signal offset.

[0031] In some embodiments, the machine learning algorithm module is configured to perform at least one of the following:

[0032] A hybrid model based on a convolutional neural network and a long short-term memory network was constructed. The convolutional neural network module was used to extract local temporal variation features, and the long short-term memory network module was used to capture the dynamic dependencies of continuous action sequences. The output layer used a softmax or sigmoid classifier to identify the activity categories of facial muscle groups. These activity categories included chewing, swallowing, tooth clicking, limbic movements, speech expression, bruxism or clenching, and resting or active patterns in the maximum mandibular opening position, maximum mandibular protrusion position, maximum mandibular lateral position, and resting jaw position.

[0033] Combining dynamic temporal pattern recognition and spectrum analysis algorithms, quantitative analysis of muscle activity frequency, duration, symmetry or coordination parameters is performed;

[0034] Build an abnormal pattern recognition and risk warning model to identify signals of frequent non-task muscle engagement, persistent muscle tension, or limited mouth opening, and form a typical abnormality library;

[0035] Use dynamic time warping algorithm to conduct comparative analysis of individual behaviors;

[0036] Spectral density analysis algorithms are used to identify the characteristics of high-frequency abnormal muscle discharges and low-frequency weakness;

[0037] Calculate the functional stability index, muscle imbalance risk index or abnormal behavior frequency and output the risk score.

[0038] In some embodiments, the ex vivo intelligent analysis module is configured to implement at least one of the following:

[0039] Establish individual patient baseline files and conduct personalized health analysis and trend tracking for each patient;

[0040] Build a telemedicine service platform to support remote diagnosis, rehabilitation guidance and / or personalized health management functions;

[0041] Establish a cloud server-based data access system to support the transmission, storage, classification, encryption or access of high-frequency sensor data streams;

[0042] Develop a doctor-side platform that supports viewing multi-day patient data, AI analysis reports, and trend charts; provides video consultation, remote annotation, and diagnosis and treatment recommendation entry modules, supporting remote intervention and follow-up; integrates voice / text quick entry or medical knowledge base docking and case archiving functions to improve doctor efficiency;

[0043] Build patient-oriented applications that support real-time data display, health score feedback, and interactive guidance; embed rehabilitation exercise guidance modules; and enable message exchange, report push, or intelligent consultation functions between patients and doctors to improve medical efficiency and compliance.

[0044] From the above, it can be seen that according to the above technical solution disclosed in the present invention, modern flexible electronics, AI intelligent analysis and telemedicine mode are integrated to provide a new solution for the diagnosis and rehabilitation of oral and maxillofacial muscle dysfunction. It is highly innovative, has clinical transformation prospects and social health benefits, and has broad application and promotion value.

[0045] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0047] Figure 1 A schematic diagram of the structure of an intelligent diagnosis and treatment system for oral and mandibular muscle dysfunction according to some embodiments of the present invention is shown;

[0048] Figure 2 A flowchart illustrating a method for implementing an intelligent diagnosis and treatment system for oral and mandibular muscle dysfunction according to some embodiments of the present invention is shown;

[0049] Figure 3 A schematic structural diagram of a micro control unit according to some embodiments of the present invention is shown. DETAILED DESCRIPTION

[0050] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0051] See also Figure 1 As shown, an embodiment of the present invention provides an intelligent diagnosis and treatment system for oromandibular muscle dysfunction, and the intelligent diagnosis and treatment system for oromandibular muscle dysfunction includes:

[0052] The flexible wearable sensing module includes an electromyographic signal acquisition unit, a strain sensing unit, and a pressure sensing unit. The electromyographic signal acquisition unit is used to collect electromyographic signals of the oral and maxillofacial muscles, the strain sensing unit is used to collect facial dynamic displacement signals, and the pressure sensing unit is used to collect maxillofacial pressure distribution signals. Specifically, the electromyographic signal (sEMG) acquisition unit is designed and the layout is optimized; flexible strain and pressure sensor design (such as graphene and silver nanowire materials) is designed; and integrated packaging, comfort, and stability testing are performed.

[0053] The data preprocessing and communication module can be integrated with the flexible wearable sensor module in the wearable device, including a data preprocessing submodule and a wireless communication submodule. The data preprocessing submodule is used to preprocess the data obtained from the flexible wearable sensor module, and this module can be developed in an embedded manner; the wireless communication submodule outputs the preprocessed data wirelessly; for example, the data preprocessing and communication module: micro MCU system development, embedded filtering and compression algorithms; Bluetooth 5.0 / WiFi wireless module debugging and transmission protocol optimization; power management design (rechargeable, low power consumption).

[0054] The off-body intelligent analysis module, wirelessly connected to the data preprocessing and communication module, includes at least one of the following: a mobile terminal application, a cloud-based platform, or a physician-side platform, at least one of which includes a machine learning algorithm module for displaying, analyzing, identifying, monitoring, evaluating, or assisting in clinical diagnosis. The off-body intelligent analysis module implements a software platform and intelligent analysis, including mobile application development (user recording, physiological visualization); cloud-based data management and a physician-side platform; and machine learning algorithms for myoelectric pattern recognition and functional assessment.

[0055] This embodiment can also achieve system integration and clinical verification, scenario-based deployment and user experience optimization, multi-center subject testing (children / adults), doctor-patient interaction process optimization and the establishment of a health intervention recommendation feedback mechanism.

[0056] The implementation and control method of the present invention adopts Figure 2 The method shown.

[0057] Therefore, according to an embodiment of the present invention, for patients undergoing postoperative rehabilitation of facial paralysis, chewing / swallowing disorders, temporomandibular joint disorders, and postoperative reconstruction of oral cancer, multimodal physiological signals such as electromyography (sEMG), strain and pressure of multiple key facial muscle groups (such as the masseter and temporalis muscles) can be collected in real time. The system is based on a machine learning algorithm model, such as a hybrid model constructed by splicing a convolutional neural network (CNN) and a long short-term memory network (LSTM), to recognize the user's chewing, swallowing, speaking and other actions, and evaluate the functional status of the muscles; through an adaptive calibration mechanism and signal fusion algorithm, the signal quality and analysis accuracy are improved; Bluetooth or Wi-Fi wireless data transmission is supported, and remote medical platforms can be connected to realize intelligent and continuous health monitoring and auxiliary diagnosis; through comprehensiveness and efficiency, the limitations of traditional diagnostic methods are solved, and it has clear commercial application value.

[0058] The detailed description of the specific modules is as follows:

[0059] (1) Flexible wearable sensing module

[0060] sEMG signal (surface electromyography signal): sampling frequency 1000-2000Hz, signal bandwidth 20-450Hz, using differential acquisition, reflecting the discharge of facial masticatory muscles, temporalis muscles and other muscle groups;

[0061] Flexible strain sensor signal: mainly reflects dynamic facial displacement, such as mouth opening and closing, jaw movement, with a sampling frequency of about 100-200Hz;

[0062] Flexible pressure sensor signal: used to monitor maxillofacial pressure distribution (such as bite force changes), with a sampling frequency of 100-200 Hz;

[0063] All sensors will be sampled through a multi-channel analog front-end (AFE) module + 12-bit ADC (analog-to-digital converter) and time-stamped uniformly through a micro MCU to build a unified time-aligned data packet.

[0064] The specific implementation is as follows:

[0065] 1. Electrode material selection: Flexible conductive materials (such as PEDOT:PSS and silver nanowire / elastomer composites) are used to construct skin-adhesive dry electrodes to improve signal stability and biocompatibility during long-term wear.

[0066] Layout optimization strategy:

[0067] Develop a standardized point distribution plan based on the maxillofacial muscle anatomy atlas and signal source modeling;

[0068] Introducing data-driven methods (such as PCA and signal quality assessment indicators) to optimize electrode layout and improve information redundancy and perception coverage;

[0069] Signal acquisition system design: Develop high-sensitivity, low-noise multi-channel amplification and filtering circuits to support multi-channel synchronous acquisition and real-time preprocessing functions.

[0070] 2. Flexible strain / pressure sensor design (such as graphene, silver nanowire materials)

[0071] Design goal: To sense facial epidermal deformation or mechanical stress changes caused by actions such as mouth opening and closing, chewing, and tongue movement, as a supplement to sEMG signals to achieve multimodal collaborative perception.

[0072] Sensing materials and structure design:

[0073] Use highly sensitive graphene, carbon nanotubes or silver nanowire / PDMS composites to construct stretchable strain / pressure sensing layers;

[0074] Design microstructures such as island-bridge structures and serpentine interconnected electrodes to improve their stretchability and dynamic response capabilities;

[0075] Packaging and integration strategy: Develop low-modulus medical-grade elastic materials for packaging to ensure the sensor conforms to the skin contours and withstands repeated stretching and deformation;

[0076] Performance parameter evaluation:

[0077] Response sensitivity (Gauge Factor), linearity, hysteresis, recovery time, etc.;

[0078] Conduct simulation tests on oral movements (such as opening and closing the mouth, and lateral movements) to assess the ability to accurately identify and distinguish them.

[0079] 3. Integrated packaging, comfort and stability testing

[0080] Modular integration:

[0081] Integrate multiple sensors such as electromyography, resistance, strain / pressure, etc. on a flexible printed circuit board (FPCB);

[0082] Develop low-power microcontroller units (MCUs) and wireless communication modules to enable local data preprocessing and real-time transmission (e.g., BLE, Wi-Fi);

[0083] System Package Design

[0084] Low-modulus medical-grade elastic material is used for packaging to ensure that the sensor fits the skin contour and withstands repeated stretching and deformation. Optionally, the flexible material body can be designed into a square, curved, U-shaped or other structure.

[0085] Provide two design options: "patch type" and "reusable type";

[0086] Apply processes such as hot pressing, spray encapsulation, and plasma surface treatment to enhance equipment durability and environmental adaptability (sweat, friction, etc.);

[0087] Comfort and user experience evaluation:

[0088] Conduct wearing comfort, skin-friendliness, and long-term wearing stability tests on volunteers;

[0089] Infrared thermal imaging and bioimpedance technology were used to evaluate the effects of the device on skin permeability and safety of electrical stimulation.

[0090] (2) Data preprocessing and standardization algorithm

[0091] Denoising:

[0092] The sEMG signal was subjected to a bandpass filter (Butterworth 4th order, 20–450 Hz) to remove motion artifacts and power frequency interference;

[0093] The flexible strain and pressure signals are processed by combining wavelet transform (Daubechies-4) and sliding average filtering to improve the signal-to-noise ratio;

[0094] sEMG RMS normalization: normalized based on the maximum voluntary muscle contraction (MVC) signal;

[0095] The strain and pressure signals are normalized to the range of [0,1] to reduce the differences in data distribution among channels;

[0096] Feature augmentation and dimension alignment:

[0097] Introducing signal envelope extraction (Hilbert transform) and frequency band energy decomposition;

[0098] Interpolate and resample the multi-channel signal (e.g., spline interpolation) to a unified time frame structure to construct a three-dimensional input tensor: [number of channels × time points × feature dimensions];

[0099] The adaptive calibration module design includes:

[0100] To address issues such as sensor wearing deviation and skin humidity changes, a signal drift calibration model based on self-supervised learning is constructed;

[0101] Apply feature distribution matching technology (such as CORAL and Deep CORAL) to achieve spatial alignment of source and target domain signals and improve model generalization capabilities;

[0102] Implement a dynamic threshold correction mechanism based on wearing history data to automatically adjust sensing sensitivity and signal mapping parameters.

[0103] Key software and implementation tools

[0104] Python 3.9+SciPy / Numpy / Scikit-learn: signal processing and data preprocessing;

[0105] MATLAB R2022b: Wavelet Analysis and Filtering Simulation;

[0106] LabVIEW (optional): for real-time hardware data flow visualization;

[0107] TensorFlow 2.0 (some functions are used for self-supervised alignment models);

[0108] The unified interface connects to the embedded acquisition system through MQTT / Serial or BLE communication protocol;

[0109] (3) Wireless Communication and Embedded Systems

[0110] This paper focuses on the development of embedded systems focusing on efficient acquisition, intelligent preprocessing, stable wireless transmission and low-power operation of sensor signals, thereby realizing a miniaturized, highly integrated, low-latency data acquisition and transmission platform suitable for facial wearable scenarios, laying a technical foundation for subsequent telemedicine and AI analysis.

[0111] 1. Micro MCU system development, embedded filtering and compression algorithms

[0112] Microcontroller selection and platform construction:

[0113] Choose a low-power, high-performance microcontroller (such as Nordic nRF52840 and STM32 series) with multi-channel ADC, BLE / WiFi communication capabilities, and rich GPIO resources;

[0114] Design flexible PCBs with multi-layer routing to fit within device space constraints and meet wearable requirements.

[0115] Signal preprocessing algorithm integration:

[0116] Integrated digital bandpass filter (e.g. 20-450Hz) for EMG signal noise suppression, filtering out power frequency interference (50 / 60Hz) and motion artifacts;

[0117] Implement the Fast Fourier Transform (FFT) module to perform preliminary analysis in the time and frequency domains;

[0118] Adding compression algorithms (such as adaptive Delta coding and RLE) effectively reduces the amount of transmitted data and improves communication efficiency;

[0119] Set up an adjustable mechanism for sampling rate and storage cache to support dynamic power consumption control under different analysis accuracy.

[0120] 2. Bluetooth 5.0 / WiFi wireless module debugging and transmission protocol optimization

[0121] Communication module debugging and integration:

[0122] Supports multiple communication protocols (BLE 5.0, WiFi IEEE 802.11b / g / n), taking into account both short-range fast synchronization and medium- and long-range stable communication;

[0123] The integrated antenna layout optimization algorithm improves signal penetration and stability, and adapts to complex wearing scenarios of the mouth and jaw (such as occlusion and angle changes when talking and chewing).

[0124] Data transmission protocol design:

[0125] Design a lightweight custom communication protocol stack that supports data packet packaging, timestamp marking, packet loss detection, and breakpoint resuming;

[0126] Optimize BLE multi-channel data exchange rate (through ATT MTU adjustment and connection parameter negotiation) to improve real-time performance and system energy efficiency;

[0127] Implement standard data docking interfaces (such as MQTT and WebSocket) with mobile APPs and cloud platforms, and support two-way interaction (data reporting and remote updates).

[0128] 3. Power management design (rechargeable, low power consumption)

[0129] Power supply system design:

[0130] Uses high-density flexible lithium polymer battery (LiPo) as the main power source to meet long-term working requirements (e.g. >8 hours);

[0131] Integrated Type-C rechargeable module, supporting intelligent fast charging, battery protection and overvoltage and overcurrent control;

[0132] Supports wireless charging expansion modules to achieve a more flexible user experience.

[0133] Low power strategy optimization:

[0134] Utilize the MCU's multi-level sleep modes (such as Deep Sleep and Stop Mode) to automatically switch during non-collection periods;

[0135] Set up dynamic sensor wake-up mechanisms (such as motion triggering and timed sampling) to extend standby battery life;

[0136] The power consumption of the entire system is controlled to a target of <15mW, ensuring that the energy burden and heat accumulation during daily wear are minimized.

[0137] This embedded system will serve as the data center of the entire diagnostic platform, realizing a closed loop from signal acquisition → preprocessing → wireless transmission → cloud interaction, providing strong support for the promotion and application of flexible wearable devices in clinical and home scenarios.

[0138] (IV) Function and implementation of the in vitro intelligent analysis module

[0139] In order to achieve accurate identification and dynamic assessment of oral and mandibular muscle dysfunction, highly robust processing procedures and intelligent analysis models are developed for different physiological signals (such as surface electromyography signals (sEMG), strain and pressure signals, etc.) to achieve symptom feature extraction, risk grading and personalized assessment.

[0140] First, a unified signal synchronization sampling mechanism is established to ensure that sEMG, strain and pressure signals have a unified time base; denoising, normalization and feature enhancement of each channel signal are implemented to improve the consistency and interpretability of the algorithm input data; and a self-calibration algorithm that can adapt to different wearing positions is constructed to reduce the risk of misjudgment caused by signal offset.

[0141] For the action and function recognition algorithm model, a hybrid model based on convolutional neural network (CNN) and long short-term memory network (LSTM) was developed to identify key facial muscle activities (such as chewing, swallowing, and speaking); combined with dynamic temporal pattern recognition (such as DTW) and spectral analysis technology, core parameters such as muscle activity frequency, duration, symmetry and coordination were quantitatively analyzed.

[0142] Implement an early warning model for abnormal patterns (such as high-frequency involuntary contractions, excessive tension in the masseter muscles, limited mouth opening, etc.) to preliminarily determine the type and severity of functional disorders.

[0143] Personalized health analysis and trend tracking, establishing individual baseline profiles, using the patient's own historical data as a comparison benchmark, and dynamically evaluating functional changes;

[0144] Introducing time series analysis and trend prediction models (such as ARIMA or Transformer) to monitor disease evolution and evaluate rehabilitation efficacy;

[0145] Using a lightweight "intelligent scoring system", multiple physiological indicators are integrated into a visual health index, making it easier for users and doctors to understand and track.

[0146] Based on the sequence characteristics of multi-channel signals, an intelligent model is constructed to automatically identify and grade key facial movements (such as chewing, swallowing, mouth opening, speaking, etc.) and their functional states (such as symmetry, frequency, muscle strength).

[0147] Specifically, the action label system and collection standards are formulated:

[0148] Category A: Chewing

[0149] Category B: Swallowing

[0150] Category C: Speaking

[0151] Category D: Open-Close

[0152] Category E: Resting state (Rest)

[0153] Use a video-assisted annotation system (such as ELAN or VIA tools) to synchronously collect motion data and build an accurate motion-signal pairing database;

[0154] The initial sample size requirement is: 30 subjects × 10 times per type of action × multiple angles, with a total of more than 3,000 groups of original samples.

[0155] In the constructed dual-channel convolution-recurrent hybrid model (CNN+LSTM), the CNN module is used to extract local temporal change features (such as sEMG burst patterns and strain change rates); the LSTM module is used to capture the dynamic dependencies of continuous action sequences; the output layer uses a softmax classifier or sigmoid for multi-label tasks; an attention mechanism (Attention Layer) is added to enhance the feature weighting of signals during critical periods; the model input structure is: [number of samples × time frame × feature channel], and the feature dimensions include RMS, frequency domain features (such as average power), symmetry indicators, etc.

[0156] When constructing an abnormal pattern recognition and risk warning model, a typical abnormality library is constructed (such as frequent non-task muscle group participation, continuous muscle tension state, and limited mouth opening signal templates); DTW (dynamic time warping) is introduced for comparative analysis of individual behaviors; and spectral density analysis (Welch method) is introduced to determine characteristics such as high-frequency abnormal muscle discharge and low-frequency weakness.

[0157] The indicators output by the risk score include: functional stability index FSI; muscle group imbalance risk index MBI; abnormal behavior frequency counter (per min).

[0158] Core software and computing platform:

[0159] Model training environment: Python + TensorFlow / PyTorch;

[0160] Data enhancement modules (Additive Noise, Time Warping, Window Slicing);

[0161] Transferable model frameworks (e.g., TS-TCC, SimCLR for time-series);

[0162] Model deployment platform:

[0163] Embedded deployment (TensorFlow Lite) + Android app or web server (ONNX Runtime);

[0164] Output results are uploaded to the remote cloud platform (supporting JSON format structured data upload)

[0165] (5) Telemedicine platform and doctor interaction system

[0166] In order to achieve remote closed-loop collaboration between wearable devices, users and doctors, a safe, intelligent and highly interactive telemedicine service platform has been built to support functions such as remote diagnosis, rehabilitation guidance, and personalized health management.

[0167] 1. Construction of cloud data reception and management platform

[0168] Establish a cloud server-based data access system to support real-time uploading and storage of high-frequency sensor data streams;

[0169] Implement data classification, encrypted storage and permission control mechanisms to comply with privacy policy requirements such as the "Personal Health Information Protection Specification";

[0170] Supports different permission data access for user terminals and doctor terminals to ensure medical information security and compliance.

[0171] 2. Functional design of doctor workstation

[0172] Develop a web / mobile work interface dedicated to doctors, supporting the viewing of multi-day patient data, AI analysis reports and trend charts;

[0173] Provides video consultation, remote annotation and diagnosis and treatment recommendation entry modules, supporting remote intervention and follow-up;

[0174] It integrates voice / text quick entry, medical knowledge base docking and case archiving functions to improve doctors' efficiency.

[0175] 3. User-side health management and interactive functions

[0176] A user-oriented mobile app interface that supports real-time data display, health score feedback, and interactive guidance;

[0177] Built-in rehabilitation exercise guidance module (such as intelligent prompts for masseter muscle relaxation and mouth opening training);

[0178] Realize user-doctor message exchange, report push and "smart consultation" functions to improve medical treatment efficiency and compliance.

[0179] The entire system ultimately realizes a closed-loop process from flexible perception → embedded collection → intelligent analysis → cloud interaction → clinical feedback, improving the ability to detect and intervene early in chronic diseases such as oral and mandibular muscle dysfunction.

[0180] (VI) System integration and clinical verification

[0181] This phase aims to complete the full-link integration of hardware systems, software platforms, and intelligent algorithms, and conduct functional testing and system evaluation in a real clinical environment to ensure that the system has good usability, stability, and clinical adaptability.

[0182] 1. System integration and functional joint debugging

[0183] (1) Hardware-software integration

[0184] Connect multi-channel flexible sensors (sEMG, strain, pressure) to the acquisition controller via BLE / USB and complete the programming of the embedded MCU driver module;

[0185] Open up the complete chain from raw data collection → local caching → real-time transmission → data display / analysis → feedback control;

[0186] Complete data format interface docking with cloud platforms (such as medical imaging platforms or electronic medical record systems) and support CSV / JSON format output.

[0187] (2) Functional module joint debugging test

[0188] The joint debugging projects include:

[0189] Multi-channel synchronous sampling accuracy error ≤ 5ms;

[0190] Real-time data visualization refresh rate ≥ 15 frames / second;

[0191] Intelligent recognition module delay ≤ 0.3s;

[0192] Using simulated signal sources and standard motion models, test various motion recognition accuracy, feature extraction stability, model classification accuracy, and response time;

[0193] System stability test cycle: 24 hours of continuous operation without packet loss or crash.

[0194] 2. Clinical Validation and Evaluation Indicators

[0195] (1) Clinical application scenario design

[0196] Indications: Postoperative rehabilitation of facial paralysis, chewing / swallowing disorders, temporomandibular joint disorders, and oral cancer postoperative reconstruction patients;

[0197] Cooperating unit: Rehabilitation Department / Maxillofacial Surgery Department of a Class-A Tertiary Stomatological Hospital;

[0198] Number of subjects: We plan to recruit 30-50 clinical patients and 20 healthy volunteers, with a 1:1 control design.

[0199] Verification period: The initial verification period is 3 months.

[0200] In summary, according to the embodiments of the present invention, the following beneficial effects can be achieved:

[0201] The present invention provides an intelligent diagnosis and treatment system for oromandibular muscle dysfunction. This system integrates modern flexible electronics, AI intelligent analysis, and telemedicine to provide a new solution for the diagnosis and rehabilitation of oromandibular muscle dysfunction. The system is highly innovative, has promising clinical translation prospects, and provides social health benefits. It has broad application and promotion value, including but not limited to:

[0202] Provides a non-invasive, comfortable and continuous physiological data collection method;

[0203] It can realize dynamic visualization of facial muscle function;

[0204] Improve the early detection rate of oral and mandibular muscle dysfunction;

[0205] It effectively improves patient compliance and medical efficiency, and is especially suitable for children, the elderly and people recovering after surgery.

[0206] It should be noted that:

[0207] The algorithm and display provided herein are not inherently related to any particular computer, virtual device or other equipment. Various general-purpose devices can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing this type of device. In addition, the present invention is not directed to any specific programming language. It should be understood that various programming languages ​​can be utilized to implement the content of the present invention described herein, and the above description of specific languages ​​is for the purpose of disclosing the best mode of the present invention.

[0208] In the description provided herein, numerous specific details are described. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques are not shown in detail so as not to obscure the understanding of this description.

[0209] Similarly, it should be understood that in order to streamline the invention and aid understanding of one or more of the various inventive aspects, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, this method of disclosure should not be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim.

[0210] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition may be divided into multiple submodules or subunits or subcomponents. All features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed herein may be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0211] Furthermore, those skilled in the art will appreciate that although some embodiments described herein include certain features and not other features included in other embodiments, the combination of features from different embodiments is intended to be within the scope of the invention and to form different embodiments.

[0212] The various component embodiments of the present invention can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components of the device for determining parameters of laser minimally invasive ablation of bone tissue according to an embodiment of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and a computer program product) for executing part or all of the methods described herein. Such a program for implementing the present invention can be stored on a computer-readable medium, or can be in the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0213] An embodiment of the present invention provides a non-volatile computer storage medium storing at least one executable instruction, which can execute the control method of the above-mentioned intelligent diagnosis and treatment system for oral and mandibular muscle dysfunction in any of the above-mentioned method embodiments.

[0214] Figure 3 The schematic diagram of the structure of an embodiment of a micro control unit in the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the micro control unit.

[0215] like Figure 3As shown, the micro control unit may include: a processor 302 , a communications interface 304 , a memory 306 , and a communication bus 308 .

[0216] Processor 302, communication interface 304, and memory 306 communicate with each other via communication bus 308. Communication interface 304 is used to communicate with other devices, such as client devices or other server network elements. Processor 302 is used to execute program 310, which specifically performs the steps described in the data preprocessing embodiment for the intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction.

[0217] Specifically, the program 310 may include program codes, which include computer operation instructions.

[0218] Processor 302 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in the onboard image processing board may be of the same type, such as one or more CPUs, or may be of different types, such as one or more CPUs and one or more ASICs.

[0219] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0220] The program 310 can be specifically used to enable the processor 302 to execute operations corresponding to the data preprocessing embodiment of the above-mentioned intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction.

[0221] It should be noted that the above embodiments illustrate rather than limit the invention, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between brackets should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present invention may be implemented by means of hardware comprising several different elements and by means of appropriately programmed computers. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

Claims

1. An intelligent diagnosis and treatment system for oromandibular muscle dysfunction, comprising: The flexible wearable sensing module includes an electromyographic signal acquisition unit, a strain sensing unit, and a pressure sensing unit. The electromyographic signal acquisition unit is used to acquire electromyographic signals of the oral and maxillofacial muscles, the strain sensing unit is used to acquire facial dynamic displacement signals, and the pressure sensing unit is used to acquire maxillofacial pressure distribution signals. The data preprocessing and communication module includes a data preprocessing submodule and a wireless communication submodule. The data preprocessing submodule is used to preprocess the signal data obtained from the flexible wearable sensor module, and the wireless communication submodule outputs the preprocessed data wirelessly. The off-body intelligent analysis module is wirelessly connected to the data preprocessing and communication module and includes at least one of the following: an application program provided on a mobile terminal, a cloud platform or a doctor-side platform, and a built-in machine learning algorithm module for displaying, analyzing, identifying, monitoring, evaluating or assisting clinical diagnosis of information.

2. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The electromyographic signal acquisition unit is used to monitor the surface electromyographic signals of the oral and maxillofacial muscles to reflect the resting or activation mode of the muscles during chewing, swallowing, teeth clicking, edge movement, language expression, nighttime teeth grinding or clenching, as well as the maximum mandibular opening position, maximum mandibular protrusion position, maximum mandibular lateral position, and resting jaw position; the electromyographic signal acquisition unit uses flexible conductive materials as electrodes that can be attached to the skin, and the electrodes are configured to: formulate a layout plan based on the anatomical atlas of maxillofacial muscles and signal source modeling, and introduce a data-driven algorithm to optimize the electrode arrangement; The strain sensing unit or pressure sensing unit is used to sense facial epidermal deformation or mechanical stress changes caused by opening and closing the mouth, chewing or tongue movement, and includes a sensing layer constructed of graphene, Mxene, carbon nanotubes, super-folded carbon materials, metal-based conductive materials, liquid metal and / or PDMS materials, and the sensing layer adopts an island bridge structure or a serpentine interconnection structure.

3. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The flexible wearable sensing module and the data preprocessing and wireless communication module are arranged on a flexible material body; the flexible material body is worn on the human head, craniofacial area and / or neck; or, the flexible material body is attached to the surface of the oral and maxillofacial muscles.

4. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The electromyographic signal acquisition unit, the strain sensing unit and the pressure sensing unit are integrated on a flexible circuit board.

5. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The data preprocessing submodule is arranged on a flexible circuit board, and includes a micro MCU and at least one of the following modules: a multi-channel digital-to-analog converter, a filter, a Fourier transform module, a data compression module or a data cache module.

6. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The wireless communication submodule is configured as at least one of the following: Support BLE or WiFi communication protocols, taking into account both short-range fast synchronization and medium- and long-range stable communication; Integrated antenna layout optimization algorithm improves signal penetration and stability, adapting to complex wearing scenarios around the mouth and jaw; Design a lightweight custom communication protocol stack that supports data packet packaging, timestamp marking, packet loss detection and / or breakpoint resuming; Optimize BLE multi-channel data exchange rate to improve real-time performance and system energy efficiency; Implement a standard data docking interface with mobile terminal applications, cloud or doctor-side platforms, and support two-way interaction.

7. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The data preprocessing and communication module further includes a battery module, and the battery module is configured as at least one of the following: Adopt high-density flexible lithium polymer battery to meet the needs of long-term work; Integrated Type-C rechargeable module, supporting intelligent fast charging, battery protection and overvoltage and overcurrent control; Supports wireless charging expansion to achieve a more flexible user experience; Utilize the multi-level sleep mode of the data pre-processing submodule to achieve automatic switching during non-collection periods; Set up a dynamic wake-up mechanism for the sensor unit to extend standby battery life; The system power consumption control target is maintained at less than 15mW to ensure that the energy burden and heat accumulation of daily wear are minimized.

8. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The data preprocessing and communication module is configured to perform at least one of the following: Establish a unified signal synchronization sampling mechanism to ensure that the myoelectric signals of facial oromandibular muscles, facial dynamic displacement signals, and maxillofacial pressure distribution signals have a unified time base; Achieve denoising, normalization or feature enhancement of each channel signal to improve the consistency and interpretability of algorithm input data; Build a self-calibration algorithm that can adapt to different wearing positions to reduce the risk of misjudgment caused by signal offset.

9. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The machine learning algorithm module is configured to perform at least one of the following: A hybrid model based on a convolutional neural network and a long short-term memory network was constructed. The convolutional neural network module was used to extract local temporal variation features, and the long short-term memory network module was used to capture the dynamic dependencies of continuous action sequences. The output layer used a softmax or sigmoid classifier to identify the activity categories of facial muscle groups. These activity categories included chewing, swallowing, tooth clicking, limbic movements, speech expression, bruxism or clenching, and resting or active patterns in the maximum mandibular opening position, maximum mandibular protrusion position, maximum mandibular lateral position, and resting jaw position. Combining dynamic temporal pattern recognition and spectrum analysis algorithms, quantitative analysis of muscle activity frequency, duration, symmetry or coordination parameters is performed; Build an abnormal pattern recognition and risk warning model to identify signals of frequent non-task muscle engagement, persistent muscle tension, or limited mouth opening, and form a typical abnormality library; Use dynamic time warping algorithm to conduct comparative analysis of individual behaviors; Spectral density analysis algorithms are used to identify the characteristics of high-frequency abnormal muscle discharges and low-frequency weakness; Calculate the functional stability index, muscle imbalance risk index or abnormal behavior frequency and output the risk score.

10. The intelligent diagnosis and treatment system for oral and maxillomandibular muscle dysfunction according to claim 1, characterized in that: The off-body intelligent analysis module is configured to implement at least one of the following: Establish individual patient baseline files and conduct personalized health analysis and trend tracking for each patient; Build a telemedicine service platform to support remote diagnosis, rehabilitation guidance and / or personalized health management functions; Establish a cloud server-based data access system to support the transmission, storage, classification, encryption or access of high-frequency sensor data streams; Develop a doctor-side platform that supports viewing multi-day patient data, AI analysis reports, and trend charts; provides video consultation, remote annotation, and diagnosis and treatment recommendation entry modules, supporting remote intervention and follow-up; integrates voice / text quick entry or medical knowledge base docking and case archiving functions to improve doctor efficiency; Build patient-oriented applications that support real-time data display, health score feedback, and interactive guidance; and embed rehabilitation exercise guidance modules; Realize message exchange, report push or intelligent consultation functions between patients and doctors to improve medical efficiency and compliance.

Images