A method and device for sensing and monitoring lip and tongue muscle pressure for orthodontic assessment

Abstract

This invention relates to a method and device for sensing and monitoring lip and tongue muscle pressure for orthodontic assessment. The method includes: acquiring a three-channel mixed signal comprising lip muscle pressure signal, pressure signal when the anterior part of the tongue contacts the palate, and pressure signal when the posterior part of the tongue contacts the palate; sequentially performing median filtering, exponential moving average smoothing, piecewise linear interpolation calculation, and static dead zone filtering to obtain a real-time mixed pressure signal; calculating the corresponding short-time energy and differential slope of the signal; performing feature analysis on the filtered three-channel mixed signal to capture the corresponding temporal relationship, morphological pattern, and duration; identifying the corresponding oral cavity movements based on the temporal relationship, morphological pattern, and duration and obtaining the identification and classification results; and evaluating the effect of orthodontic treatment based on the filtered three-channel mixed signal and the identification and classification results. Compared with the prior art, this invention has the advantage of good noise filtering effect.

Description

Technical Field

[0001] This invention relates to the field of orthodontic technology, and in particular to a method and device for sensing and monitoring lip and tongue muscle pressure for orthodontic assessment. Background Technology

[0002] Oral muscle function plays a crucial clinical role in orthodontic treatment, rehabilitation of swallowing disorders, and assessment of oral motor abilities. The lip and tongue muscles, as important functional muscle groups in the oral cavity, directly influence the normal development of the dental arch, the stability of orthodontic treatment outcomes, and the coordination of oral functions through the balance of their mechanical properties. However, traditional oral muscle function assessment has long relied on manual palpation and empirical judgment by physicians, lacking objective and quantitative assessment methods, making it difficult to meet the personalized treatment needs of the precision medicine era.

[0003] Existing publicly available technologies mainly follow several development paths: First, a small airbag is placed at the tongue-palatine or lip-tooth interface and connected to an external pressure gauge via a catheter. This approach is intuitive and its sensitivity is easily adjustable, but the catheter alters the oral posture and vocalization conditions, making it unsuitable for long-term wear in natural mouth closure and everyday scenarios. Second, point sensors such as thin-film piezoresistive (FSR), strain gauges, piezoelectric, or capacitive sensors are used, attaching a single-point probe to a designated location on the palate or lip side for localized point measurement. These devices are thin and easy to place, but single-point information is limited, easily affected by installation posture and temperature drift, and multi-point expansion leads to complex wiring and increased size. Third, sensing elements are embedded in the denture or mouthguard structure to improve wearing comfort. This approach is mostly used for occlusal / molar or tooth-mouthguard interface pressure monitoring; however, there are few publicly available methods for "synchronous measurement" of soft tissue (lip / tongue) pressure, and the integration levels vary at the overall device level, including wireless connectivity, power consumption, and maintenance. Some research-grade systems simultaneously record multimodal signals such as pressure and electromyography, but these mostly rely on wired or backpack-type mainframes, suitable for short-term experiments but difficult to transfer to orthodontic follow-up and home training. Commercially available devices such as tongue depressors and lip closure force meters are often used for single-item screening or training effect evaluation, typically operating under single-path, short-term, desktop measurement conditions, making it difficult to simultaneously reflect the dynamics of the two force paths of the lips and tongue under the same wearing posture. Specifically: (1) In terms of measurement dimensions and comprehensiveness, most devices have a single function, or only focus on lip pressure monitoring, or only design for tongue pressure, and lack an integrated system that can simultaneously collect pressure from multiple points such as the lips, tongue and cheek muscles. Such single-site monitoring cannot reflect the coordinated action and force balance of the oral muscle group, while the effect evaluation of orthodontic treatment depends precisely on the understanding of the overall biomechanical environment of the oral cavity; (2) In terms of power consumption management, the fixed sampling rate design lacks an intelligent control mechanism and cannot dynamically adjust the sampling strategy according to the connection status and clinical needs, resulting in short device battery life (usually less than 24 hours), and frequent charging seriously affects user experience and monitoring continuity. (3) In terms of comfort, the universal shell design is difficult to adapt to individual differences in dental arch shape, resulting in poor wearing stability. Furthermore, the rigid material is prone to causing mucosal discomfort or pressure ulcers, leading to low patient compliance. (4) In terms of data processing, the original signal transmission mode occupies a lot of bandwidth and storage space, and lacks effective noise filtering and feature extraction algorithms. Clinicians need to spend a lot of time on data processing and it is difficult to quickly obtain valuable diagnostic information.

[0004] Therefore, the field of orthodontics and rehabilitation medicine urgently needs an innovative solution that can achieve long-term, comfortable, and accurate monitoring. This invention addresses the aforementioned technical bottlenecks by providing a lip and tongue muscle pressure monitoring device based on flexible electronics technology, featuring adaptive sampling capabilities and personalized customization. Through multimodal sensor layout, intelligent power management algorithms, and a biomimetic shell design, it effectively solves the systemic shortcomings of existing technologies in terms of measurement dimensions, battery life, wearing comfort, and data processing, providing a reliable technical platform for the quantitative assessment of oral function.

[0005] The invention disclosed in CN119548141A is a system for acquiring and analyzing the biomechanical characteristics of the lip and tongue muscles. The system includes: a fixation component, which is combined with an oral appliance to be fixed in the user's mouth; a acquisition component for acquiring lip biomechanical signals generated by the lips and / or palatal biomechanical signals generated by the tongue muscles; and an analysis component for identifying the user's tongue muscle strength information and / or lip muscle strength information based on the lip biomechanical signals and / or palatal biomechanical signals, and generating the user's patient assessment information based on the tongue muscle strength information and / or lip muscle strength information. This application embodiment, combined with a corresponding oral appliance and fixed in the oral cavity, enables real-time monitoring of the strength of the patient's tongue and lip muscles, allowing monitoring of the strength of the tongue and lip muscles during training and non-training periods. This allows patients and doctors to view the patient's lip and tongue muscle condition in real time, and through information processing and analysis by the terminal system, to assess the patient's condition and provide reminders and guidance. However, the system provided by this solution lacks effective noise filtering and feature extraction algorithms, requiring clinicians to spend a significant amount of time on data processing, making it difficult to quickly obtain valuable diagnostic information. Summary of the Invention

[0006] This invention addresses the shortcomings of existing oral pressure monitoring devices by providing a method and device for sensing and monitoring lip and tongue muscle pressure for orthodontic assessment, thus solving problems such as poor comfort, short battery life, and poor signal quality in existing technologies.

[0007] The objective of this invention can be achieved through the following technical solutions: A method for sensing and monitoring lip and tongue muscle pressure for orthodontic assessment includes: Step 1: Acquire a three-channel mixed signal including lip muscle pressure signal, pressure signal when the front of the tongue contacts the palate, and pressure signal when the back of the tongue contacts the palate; Step 2: Perform median filtering on the three-channel mixed signal, then perform exponential moving average smoothing on the median-filtered mixed signal to obtain the baseband voltage characteristic value; based on the baseband voltage characteristic value, perform piecewise linear interpolation calculation to obtain the equivalent absolute physical force value; perform static dead-zone filtering on the equivalent absolute physical force value to obtain the real-time mixed pressure signal; Step 3: Calculate the short-time energy and differential slope of the real-time mixed pressure signal; when the short-time energy of the signal is detected to be greater than a preset personalized resting threshold and the differential slope is greater than a preset slope threshold, increase the signal sampling rate and return to Step 1; Step 4: Perform feature analysis on the filtered three-channel mixed signal to capture the corresponding temporal relationship, morphological pattern, and duration; based on the temporal relationship, morphological pattern, and duration, identify the corresponding oral movement and obtain the identification and classification results; based on the filtered three-channel mixed signal and the identification and classification results, evaluate the effect of orthodontics.

[0008] Furthermore, the median-filtered mixed signal is smoothed using an exponential moving average to obtain the baseband voltage characteristic value. The corresponding processing expression is: in, This is the smoothed output value at the current moment. This is the median filter result input at the current time. α is the smoothed output value of the previous time step, and α is the smoothing coefficient.

[0009] Furthermore, based on the baseband voltage characteristic value, piecewise linear interpolation is performed to obtain the equivalent absolute physical force value, and the corresponding calculation formula is as follows: Where Grams is the mapped equivalent absolute physical force value, and ADC is the input baseband voltage characteristic value. and To find two adjacent calibration anchor points in the table that enclose the current ADC value, X is the voltage sample value and Y is the corresponding standard physical weight.

[0010] Furthermore, based on the temporal relationship, morphological pattern, and duration, the corresponding oral cavity movements are identified and classification results are obtained. The corresponding identification process includes: When the pressure signal at the back of the tongue first peaks, followed by the pressure signal at the front of the tongue, and the pressure signal of the lip muscles remains below the resting threshold, it is classified as a normal swallowing pattern. When the peak pressure signal of the anterior part of the tongue is missing or below the normal threshold, the pressure signal of the lip muscles increases and exceeds the preset compensatory force threshold, and the timing of the pressure signal of the posterior part of the tongue is disordered, it is classified as an abnormal tongue protrusion and swallowing pattern.

[0011] Furthermore, after evaluating the effect of the orthodontic treatment, the corresponding evaluation results are obtained, and the personalized resting threshold and slope threshold are updated based on the evaluation results.

[0012] Furthermore, while increasing the signal sampling rate, a pre-set circular buffer is latched; the circular buffer is used to cyclically store the latest data at the default sampling rate within the last 2-3 seconds.

[0013] Furthermore, oral motor activities specifically include swallowing, biting, and resting.

[0014] Furthermore, the process of setting the initial values ​​for the personalized resting threshold and slope threshold includes: Three-channel mixed pressure signals were collected under resting closure, and the median of each channel signal was calculated for zero-point calibration. After zero-point calibration, three-channel mixed pressure signals were collected again under multiple empty swallows and multiple biting actions, and the corresponding peak pressure, duration, and rise slope were calculated. Based on the peak pressure, duration, and rise slope, the corresponding personalized resting threshold and slope threshold were calculated.

[0015] Furthermore, the methods also include: While performing pressure monitoring, the device's power and temperature are monitored in real time; when the power is below the first preset threshold, an alert is issued; when the power is below the second preset threshold, the signal sampling rate is reduced to 5Hz and unnecessary peripherals are turned off; when the power is below the third preset threshold, the device enters sleep mode. When the temperature exceeds the first preset temperature threshold, frequency reduction or alarm is triggered; when the temperature exceeds the second preset temperature threshold, the machine is forced to shut down.

[0016] The present invention also provides an apparatus for a method of monitoring the pressure of the lip and tongue muscles for orthodontic assessment as described above, comprising: The braces-like shell, two lingual pressure sensing units, a labial pressure sensing unit, circuit components, a battery, and a wireless charging receiver coil; The braces-like outer shell serves as the overall support for the device, comprising an outer layer, a middle layer, and an inner layer; the outer layer is a wear-resistant protective layer, the middle layer is a circuit encapsulation layer, and the inner layer is a skin-friendly and comfortable layer. The two lingual pressure sensing units, the labial pressure sensing unit, the circuit assembly, the battery, and the wireless charging receiving coil are all located in the middle layer of the brace-like shell; the two lingual pressure sensing units are located in the anterior maxillary region and the posterior maxillary region of the middle layer, respectively; the labial pressure sensing unit is located in the labial region of the maxillary incisors of the middle layer.

[0017] Compared with the prior art, the present invention has the following advantages: (1) This invention acquires three pressure signals simultaneously: lip muscle pressure signal, tongue anterior contact pressure signal with palate, and tongue posterior contact pressure signal with palate, and uses independent signal conditioning channels to ensure the temporal consistency of the three channels, thus realizing dual-path synchronous monitoring of lip and tongue pressure. This overcomes the fundamental defect of traditional single-sensor or dual-sensor time-sharing acquisition that cannot accurately capture the temporal relationship of lip and tongue coordinated movement. Moreover, traditional single-tongue sensors cannot determine the propagation direction of tongue movement; they can only know that the tongue has moved, but cannot distinguish between normal wave-like swallowing and abnormal forward-pushing swallowing. This invention, by placing sensors on the anterior and posterior parts of the tongue respectively, can accurately identify the swallowing pattern type by analyzing the corresponding temporal relationship, providing a direct basis for clinical judgment on whether myofunctional training is needed. For orthodontic patients, abnormal swallowing patterns are an important cause of open bite, anterior interdental space, and narrow dental arch, while the dual-tongue sensor of this invention can objectively quantify the existence and severity of this abnormal pattern.

[0018] (2) This invention constructs a four-level digital signal processing architecture that includes disturbance rejection, smoothing, interpolation, and dead-zone control. It effectively suppresses the 50Hz sinusoidal noise floor using exponential moving average (EMA) filtering; it adopts a piecewise linear interpolation engine based on personalized multi-point calibration to replace complex floating-point power function calculations, thereby achieving linear mapping of nonlinear curves; and it combines an absolute dead-zone truncation algorithm to eliminate zero-point baseline drift caused by oral temperature and saliva. This architecture significantly reduces the computational power consumption of the microprocessor while improving the signal-to-noise ratio of the final output pressure signal by more than 40dB, greatly improving the initial measurement accuracy and long-term stability of the flexible sensor.

[0019] (3) This invention can accurately distinguish different types of oral functional activities by analyzing the temporal correlation of the three-channel signals: swallowing events are characterized by a three-segment temporal pattern, with the peak of the posterior tongue sensor appearing earlier, the anterior tongue sensor following, and the lip sensor showing slight changes; occlusal events are characterized by a sharp increase in lip pressure and slight changes on the lingual side; and the resting state is characterized by all three channels being in a low-amplitude, stable state. This multi-channel joint judgment mechanism significantly improves the accuracy of event recognition and avoids misjudgments caused by occasional interference from a single sensor. Attached Figure Description

[0020] Figure 1This is a flowchart of a method for monitoring the pressure of the lip and tongue muscles for orthodontic assessment provided in an embodiment of the present invention; Figure 2 This is a structural diagram of a device for monitoring and sensing lip and tongue muscle pressure for orthodontic assessment provided in an embodiment of the present invention. Figure 3 This is a schematic diagram of the sensor location of a sensor for a pressure monitoring device for oral orthodontic assessment provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the circuit components of a pressure monitoring device for sensing the lip and tongue muscles for orthodontic assessment provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of an adaptive sampling algorithm for a sensory lip and tongue muscle pressure monitoring method for orthodontic assessment provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0022] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0023] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0024] Example 1 like Figure 1 As shown in the figure, this embodiment provides a method for monitoring the pressure of the lip and tongue muscles for orthodontic assessment. The method includes the following steps: S1: Acquire a three-channel mixed signal including lip muscle pressure signal, pressure signal when the front of the tongue contacts the palate, and pressure signal when the back of the tongue contacts the palate; S2: Perform median filtering on the three-channel mixed signal, then perform exponential moving average smoothing on the median-filtered mixed signal to obtain the baseband voltage characteristic value; based on the baseband voltage characteristic value, perform piecewise linear interpolation to obtain the equivalent absolute physical force value; perform static dead zone filtering on the equivalent absolute physical force value to obtain the real-time mixed pressure signal; S201: The first stage of processing is performed using pulse suppression median filtering: a sliding window of length N is maintained, and the current and historical continuous sampling points are sorted and the median is extracted to remove sudden spike noise caused by instantaneous friction of the oral mucosa or spatial electromagnetic pulse.

[0025] S202: The second stage of processing is performed using power frequency anti-interference exponential smoothing filter to eliminate the environmental power frequency sinusoidal noise introduced by the human antenna effect and obtain pure baseband voltage characteristic values.

[0026] Preferred, The expression for power frequency immunity exponential smoothing filter is: in, This is the smoothed output value at the current moment. This is the median filter result input at the current time. This is the smoothed output value from the previous time step, where α is the smoothing coefficient, ranging from 0.1 to 0.3.

[0027] S203: The third-level processing is performed using nonlinear piecewise linear interpolation mapping: a pressure-voltage lookup table containing multiple calibration anchor points is pre-constructed, the baseband voltage characteristic value is sent into the lookup table for piecewise linear interpolation calculation, and the nonlinear electrical response of the flexible thin-film pressure sensor is mapped to the equivalent absolute physical force value; the inverse proportional nonlinear electrical response of the flexible thin-film pressure sensor is straightened.

[0028] Preferred, The expression for the nonlinear piecewise interpolation mapping is: Where Grams is the mapped equivalent absolute physical force value, and ADC is the input baseband voltage characteristic value. and To find two adjacent calibration anchors in the table that enclose the current ADC value, where X is the voltage sample value and Y is the corresponding standard physical weight.

[0029] S204: The fourth stage of processing employs anti-creep static dead zone filtering to eliminate minute baseline fluctuations caused by temperature drift and material physical creep in flexible sensors. The corresponding dead zone filtering expression is: Among them, Outputt The final output is a real-time hybrid pressure signal, where Grams is the equivalent absolute physical force value after the third-level mapping, and Th... head The set dynamic dead zone threshold.

[0030] Specifically, A dynamic dead zone threshold is set. When the calculated equivalent absolute physical force value is lower than the dead zone threshold, the output result is forced to be zero to eliminate the small baseline fluctuations caused by temperature drift and material physical creep of the flexible sensor.

[0031] S3: Calculate the short-time energy and differential slope of the real-time mixed pressure signal; when the short-time energy of the signal is detected to be greater than the preset personalized resting threshold and the differential slope is greater than the preset slope threshold, increase the signal sampling rate and return to step S1; Specifically, An adaptive sampling scheduling mechanism based on a low-power event tracking system using a dual-condition AND gate and hysteresis fallback is employed: The short-time energy (absolute amplitude) and differential slope corresponding to the real-time mixed pressure signal are calculated. An event candidate is triggered only when the absolute amplitude exceeds a preset personalized resting threshold and the differential slope exceeds a preset slope threshold. This instantly increases the signal sampling rate from the default low-power standby state (10Hz) to a high-frequency capture state (100Hz) and returns to step one. After the event ends, a hysteresis comparison strategy is used. When the signal falls back to below a preset proportion of the resting threshold and remains there for a certain period, low-frequency sampling resumes.

[0032] Preferred, While increasing the signal sampling rate, a pre-set circular buffer is latched; the circular buffer is used to cyclically store the latest data at the default sampling rate within the last 2-3 seconds.

[0033] Preferred, The personalized resting threshold and slope threshold are dynamically generated by a parameter self-convergence algorithm based on patient interaction. The specific steps include: (1) Entering the zero-point calibration state, the system guides the patient to remain at rest with their mouth closed for a preset time, and extracts the steady-state median of the mixed signals of each channel during this period as the absolute zero-point baseline for this wear; (2) Entering the dynamic threshold learning state, the system guides the patient to perform standard empty swallowing and light biting actions, and captures the peak pressure of each channel under the standard actions; (3) Based on the peak pressure, multiply by the preset personalized weight coefficient, dynamically calculate and update the personalized resting threshold and slope threshold of the system to adapt to the differences in oral structure and muscle strength of different patients.

[0034] Preferred, The process of setting the initial values ​​for personalized resting threshold and slope threshold includes: Three-channel mixed pressure signals were collected under resting closure, and the median of each channel signal was calculated for zero-point calibration. After zero-point calibration, three-channel mixed pressure signals were collected again under three empty swallows and one biting action, and the corresponding peak pressure, duration and rise slope were calculated. Based on the peak pressure, duration and rise slope, the corresponding personalized resting threshold and slope threshold were calculated.

[0035] S4: Perform feature analysis on the filtered three-channel mixed signal to capture the corresponding temporal relationship, morphological pattern, and duration; based on the temporal relationship, morphological pattern, and duration, identify the corresponding oral movement and obtain the identification and classification results; based on the filtered three-channel mixed signal and the identification and classification results, evaluate the effect of orthodontics.

[0036] Preferred, Identifying corresponding oral cavity movements specifically includes targeted identification of abnormal swallowing patterns based on a three-channel spatiotemporal feature matrix: Extract the peak occurrence timing and amplitude distribution of the three-channel mixed signal in the time domain; When the pressure signal at the back of the tongue first peaks, followed by the pressure signal at the front of the tongue, and the pressure signal of the lip muscles remains below the resting threshold, it is classified as a normal swallowing pattern. When the peak pressure signal of the anterior part of the tongue is missing or below the normal threshold, while the pressure signal of the lip muscles increases sharply and exceeds the compensatory force threshold, and the timing of the pressure signal of the posterior part of the tongue is disordered, it is classified as an abnormal tongue thrust swallowing pattern (TongueThrust), and the frequency and intensity of this abnormal pattern are included in the orthodontic effect evaluation results.

[0037] Specifically, After evaluating the effectiveness of orthodontics, corresponding evaluation results are obtained, and personalized resting thresholds and slope thresholds are updated based on the evaluation results.

[0038] Preferred, The method also includes: While performing pressure monitoring, the device's power and temperature are monitored in real time; when the power is below the first preset threshold, an alert is issued; when the power is below the second preset threshold, the signal sampling rate is reduced to 5Hz and unnecessary peripherals are turned off; when the power is below the third preset threshold, the device enters sleep mode. When the temperature exceeds the first preset temperature threshold, frequency reduction or alarm is triggered; when the temperature exceeds the second preset temperature threshold, the machine is forced to shut down.

[0039] Preferred, The method also includes a data scheduling mechanism that is tightly coupled with the Bluetooth Low Energy (BLE) link state: Real-time monitoring of wireless communication link status; when the link is detected to be in a connected state, real-time mixed pressure signals and event characteristics are packaged into low-latency packets for streaming pass-through. When a link disconnection is detected, the system triggers a blind zone buffer mode to maintain the sampling of underlying signals and event recognition, and stores the captured oral movement features and corresponding timestamps locally in time blocks in a non-volatile manner. When the link is re-established, the handshake synchronization protocol is executed first to resume the local storage of block data.

[0040] Example 2 This embodiment provides an apparatus for monitoring lip and tongue muscle pressure in orthodontic assessment, as described in Embodiment 1, comprising: Overall structural composition and layout: like Figure 2 As shown, the oral pressure monitoring device of the present invention mainly consists of the following parts: a personalized braces-style silicone shell, three flexible thin-film pressure sensors, a signal conditioning circuit, an NRF52832 main control chip, a customized wing-shaped lithium battery, a power management module, and a wireless charging module.

[0041] The personalized brace-style silicone shell is made of medical-grade silicone rubber and features an asymmetrical three-layer structure: an outer 0.3mm thick wear-resistant protective layer, a middle 0.7mm thick circuit encapsulation layer, and an inner 0.2mm thick skin-friendly comfort layer. The shell is manufactured through the following steps: first, a dental scanner is used to acquire the user's three-dimensional dental arch data; then, a personalized brace mold is created using 3D printing technology; and finally, medical-grade silicone is injected for molding. The shell thickness is controlled at 1.2mm to ensure strength without affecting normal oral function. The inner side of the shell has a reference surface for mounting the sensor units, facilitating repeated assembly. The effective sensing areas of the three pressure sensor units located on the front, back, and labial sides of the palate are all aligned with the inner skin-friendly pressure layer to reduce stress attenuation and maximize the physical sensitivity to the subtle muscle forces of the lips and tongue.

[0042] Preferred, The brace-like outer shell is U-shaped or horseshoe-shaped to fit the dental arch.

[0043] like Figure 3As shown, three flexible thin-film pressure sensors are positioned as follows: Sensor B: located on the labial side of the maxillary incisors, specifically for monitoring pressure changes in the labial muscles (orbicularis oris) during functions such as mouth closure and phonation. Sensor C: located in the anterior maxillary region, to the left of the mid-palatal suture, monitoring the pressure distribution when the anterior part of the tongue contacts the maxilla. Sensor D: located in the posterior maxillary region, to the right of the mid-palatal suture, monitoring the pressure characteristics when the posterior part of the tongue contacts the maxilla. Temporal correlation analysis of the three-channel data can be used to distinguish different types of oral functional activities.

[0044] Circuit system architecture and connections: Circuit system such as Figure 4 As shown, it includes: an NRF52832 main control chip, responsible for signal processing and data transmission; and a three-channel instrumentation amplifier composed of TLV2254AIPWR operational amplifiers, with a gain of [missing information]. It features a single-pole low-pass RC filter with a cutoff frequency of 10Hz. Signals from the three sensors are connected to the P0.28, P0.29, and P0.30 pins of the NRF52832 for ADC sampling. A PCA9420 power management chip is used for intelligent power distribution. The wireless charging module employs the MWPR1516 receiver chip, utilizing electromagnetic induction for energy transfer. During charging, the power management module monitors battery voltage, current, and temperature parameters in real time. When abnormal conditions are detected (such as over-temperature or over-current), it automatically adjusts charging parameters or pauses charging to ensure system safety.

[0045] This customized wing-shaped lithium battery utilizes lithium polymer battery technology and has a capacity of 120mAh. Its shape is designed according to the buccal contour of the posterior segment of the dental arch, featuring a double-wing-shaped curved arrangement that perfectly conforms to oral anatomy. The battery incorporates a temperature sensor and overcharge / overdischarge protection circuitry to ensure safe use. It is packaged in a food-grade soft-pack, offering excellent flexibility and sealing.

[0046] System working principle and data processing: The working principle of this invention is as follows: When pressure is applied to the flexible sensors in three areas by the lip or cheek muscles, the electrical signals generated by the sensors are conditioned by the TLV2254 and then sent to the ADC channel of the NRF52832 to be converted into digital signals. The main chip uses a hybrid filtering algorithm to process the signals, eliminating noise and interference. The processed data is transmitted to a mobile APP via Bluetooth BLE for real-time display and recording. Power management adopts a multi-level power reduction strategy: when the battery level is below 30%, the user is alerted; when it is below 20%, the sampling rate is automatically reduced to 5Hz and unnecessary peripherals are turned off; when it is below 5%, it enters sleep mode. This smooth degradation strategy avoids data loss due to sudden shutdown. In terms of thermal management, when the temperature exceeds 40°C, frequency reduction or alarm is triggered, and when it exceeds 50°C, a forced shutdown is performed, providing active safety protection. The wireless charging process is also monitored throughout, and parameters are immediately adjusted or paused in case of abnormality to protect battery life.

[0047] To overcome the inherent severe nonlinearity, physical creep (slow resistance drift after pressure) and power frequency interference introduced by the oral cavity human antenna effect in flexible thin-film pressure sensors (FSRs), the NRF52832 main control chip abandons the traditional fixed bandpass filter and embeds a highly efficient four-level digital signal processing (DSP) architecture with physical characteristic compensation.

[0048] First, the sensor's weak voltage undergoes a first-stage median filter (window N=3) to precisely eliminate instantaneous electromagnetic pulses and extreme spikes caused by oral cavity movements, without disrupting the sharp rising edge characteristics of swallowing events like mean filtering. Its expression is: Among them, X t x is the median filter output for the t-th sampling point. t The input signal is the original sampled value at the current moment, and N is the set sliding window size (preferably 3).

[0049] The signal then enters the second-stage power frequency immunity exponential smoothing filter (EMA). Due to the high impedance characteristics of the FSR, the system easily couples 50Hz AC radiation into the ADC, forming a sinusoidal noise floor. The EMA filter, through specific weighting (α=0.2 in this example), completely smooths it into a pure DC baseband voltage. Its expression is: Among them, S t S is the smoothed output value at the current time. t-1 α is the smoothed output value of the previous time step, and α is the filtering smoothing coefficient.

[0050] To address the most challenging aspect of FSR—the nonlinear resistance characteristic—the system did not employ computationally expensive floating-point power function fitting. Instead, it utilized a third-level nonlinear piecewise linear interpolation engine. Based on multiple anchor points obtained during patient calibration (e.g., 0g, 10g, 50g, 100g), the system used lookup tables and linear interpolation formulas to mathematically "straighten" the response curve, accurately converting the voltage into an "equivalent absolute physical force value (gram weight)."

[0051] Finally, after the fourth stage of anti-creep static dead zone filtering, false values ​​below 2.0g caused by temperature drift or slight physical creep of the material are forcibly truncated to 0. This four-stage processing improves the output signal-to-noise ratio by more than 40dB, enabling the flexible sensor to be worn and accurately positioned immediately.

[0052] like Figure 5As shown, the adaptive sampling algorithm of this invention is as follows: The device defaults to a low-power monitoring state, operating at a sampling rate of 10Hz. In this state, the algorithm performs two tasks in parallel: first, it performs real-time filtering on the three-channel signals and calculates short-time energy and differential slope; second, it continuously monitors the Bluetooth (BLE) connection status. When the system detects a disconnection from the mobile app, it automatically enters a batch caching mode, still maintaining monitoring of basic physiological signals, storing data locally in 1-minute blocks, and resuming transmission after the connection is restored, greatly reducing communication power consumption and ensuring data integrity. The event triggering mechanism uses AND logic, requiring simultaneous satisfaction of the amplitude condition (signal > personalized resting threshold Th_rest) and the slope condition (signal change rate > threshold Th_slope). This dual-condition judgment mechanism greatly reduces the false alarm rate caused by slow head movements or sensor drift. Once an event candidate is identified, the system quickly switches to the high-frequency event capture state within 5 milliseconds, increasing the sampling rate to 100Hz. Simultaneously, the pre-triggered circular buffer is immediately latched. This buffer continuously stores the latest 2-3 seconds (10Hz) of data in the monitoring state, thus perfectly solving the trigger delay problem and achieving the capture of complete event context information. During the event duration, the algorithm continuously tracks signal characteristics and adaptively adjusts the monitoring duration according to the event intensity. Under high-fidelity sampling at 100Hz, the algorithm performs multi-evidence event confirmation and classification. It does not simply rely on a single-point threshold, but analyzes the temporal relationship, morphological pattern, and duration of multi-channel signals. For example, a true swallowing event is characterized by the early appearance of the peak value of the posterior tongue sensor, followed by the anterior tongue sensor, while the lip sensor shows slight changes; while an occlusal event is characterized by a sharp increase in lateral labial pressure. This pattern recognition-based classification method has an accuracy far exceeding that of simple threshold judgment. After the event ends, a hysteresis comparison strategy is used to determine whether the signal has returned to the baseline. After confirmation, the sampling rate is automatically reduced to 10Hz, and the event's feature values ​​(such as peak value, duration, energy integral, etc.) are extracted for local storage or subsequent transmission. This algorithm, through an "on-demand sampling" mechanism, achieves up to 62% power consumption optimization in typical usage scenarios, extending battery life to over 72 hours while ensuring high-fidelity recording of important physiological events. In signal feature extraction, the system first performs baseline calibration at 10Hz to ensure signal stability; then, it performs real-time feature extraction, identifying various oral activity patterns including swallowing, saliva production, biting, and resting. A multi-evidence cross-validation mechanism confirms events, improving classification accuracy. During the event capture phase, 100Hz high-frequency sampling ensures detailed integrity, and data is temporarily stored in a cache queue awaiting submission. The system also incorporates a hysteresis fallback strategy to avoid false triggers caused by signal fluctuations. Furthermore, the system possesses anomaly handling capabilities, capable of real-time monitoring of abnormal states such as device disconnection, low battery, and overheating, and executing corresponding protection procedures to ensure system reliability and user experience.

[0053] Upon first use, the system guides the user through a brief guided calibration process: First, the user rests with their mouth closed for 10-20 seconds, and the system automatically calculates the median signal value for each channel to complete zero-point calibration. Then, the user is guided to perform three empty swallows and one gentle bite. During this process, the system collects high-quality data and automatically extracts feature values ​​such as peak pressure, duration, and rise slope, thereby calculating the initial event detection threshold Th_event = 0.6 * average peak swallowing pressure.

[0054] Explanation of reference numerals in the attached figures 1—Braces-like outer shell; 2—Lingual pressure sensing unit; 3—Lip pressure sensing unit; 4—Circuit assembly; 5—Battery; 6—Wireless charging receiver coil.

[0055] This invention addresses the long-term follow-up needs of orthodontic assessment by proposing a brace-like device that can be worn in a natural closed-mouth posture. It incorporates two types of compliant sensing surfaces—lingual and labial—within the same outer shell to simultaneously record soft tissue forces. At the electronic and algorithmic levels, it introduces oral function pattern recognition and adaptive sampling / transmission scheduling, combined with a programmable power supply and battery health monitoring to form an energy-constrained closed loop. Furthermore, it establishes an anti-interference signal processing system and a personalized parameter self-learning calibration algorithm to achieve accurate and personalized use of the signal data. The following will provide a detailed description of the overall structural composition and layout, circuit system architecture and connections, system working principle, and data processing to enable those skilled in the art to implement the device.

[0056] The problems with existing technologies include: 1. Main problems (1) Data collection is limited to a single dimension and lacks collaborative analysis. Commercially available devices have limited functions, measuring only lip pressure or only tongue pressure, and cannot simultaneously collect pressure data from multiple points such as the lips and tongue at the same time reference. This single-path measurement cannot reflect the coordinated action and mechanical balance of the oral muscle groups, while the evaluation of orthodontic treatment effectiveness depends precisely on the understanding of this overall mechanical environment.

[0057] (2) Inability to balance high-precision capture with long battery life. Existing devices use a fixed sampling rate. If set to a high sampling rate (≥100Hz) to capture details of transient events such as swallowing and biting, power consumption is enormous and battery life is severely insufficient (usually <24 hours), failing to meet the needs of continuous monitoring over multiple days in orthodontics. If set to a low sampling rate (e.g., 10Hz) to ensure battery life, waveform details of key events will be lost, making accurate time-domain and frequency-domain analysis impossible. Existing devices' sampling and transmission strategies fail to intelligently coordinate with Bluetooth connection status. Continuous high-speed transmission during connection and the lack of effective data caching and management mechanisms after disconnection result in high average power consumption, a high risk of data packet loss, and impacts data integrity and user experience.

[0058] (3) Lack of system-level safety and energy management. The equipment lacks real-time monitoring and closed-loop control of its own power, current and temperature. It continues to operate at high load under abnormal conditions such as low power or overheating, which may not only lead to data loss, but also pose safety hazards and poor equipment reliability.

[0059] 2. Secondary issues (1) The oral environment is complex and the signal is easily affected by saliva fluctuations, mucosal friction, temperature drift, etc., resulting in low signal-to-noise ratio, poor data stability, and large measurement error of traditional monitoring equipment.

[0060] (2) Different users have huge differences in oral structure and muscle strength habits, and the general parameter settings lead to large measurement deviations; the equipment requires complex calibration by professionals before use, resulting in poor user experience and difficulty in promotion.

[0061] The beneficial effects of this invention include: (1) This invention proposes a hardware-software co-operated architecture for nonlinearity and creep compensation in flexible sensors. Addressing the technical shortcomings of existing flexible thin-film sensors, such as severe nonlinearity, susceptibility to interference from environmental power frequency antenna effects, and physical creep, this invention constructs a four-level digital signal processing architecture encompassing anti-interference, smoothing, interpolation, and dead-zone control. Specifically, exponential moving average filtering effectively suppresses the 50Hz sinusoidal noise floor; a piecewise linear interpolation engine based on personalized multi-point calibration replaces complex floating-point power function calculations, achieving linear mapping of nonlinear curves; and an absolute dead-zone truncation algorithm eliminates zero-point baseline drift caused by oral temperature and saliva. This architecture significantly reduces microprocessor computational power consumption while improving the signal-to-noise ratio of the final output pressure signal by more than 40dB, greatly enhancing the initial measurement accuracy and long-term stability of the flexible sensor.

[0062] (2) This invention provides a targeted identification technology for abnormal swallowing based on a three-channel spatiotemporal sequence matrix. Traditional single-point pressure measurement devices cannot effectively identify the core cause of orthodontic open bite relapse and dental arch stenosis—abnormal swallowing. This invention simultaneously collects three-channel pressure spatiotemporal feature matrices from the labial side, the anterior part of the tongue, and the posterior part of the tongue, and compares and analyzes the normal swallowing sequence with the abnormal tongue protrusion sequence. This technology not only achieves accurate measurement of multi-point muscle pressure, but also directly outputs quantitative diagnostic conclusions about swallowing dysfunction, filling the gap in the field of objective assessment of orthodontic muscle function.

[0063] (3) This invention establishes a low-power event tracking mechanism based on dual-condition judgment and hysteresis fallback. Addressing the technical challenges of long-duration dynamic monitoring of orthodontic devices, this invention abandons the traditional single absolute amplitude triggering method and innovatively adopts a dual-condition AND gate triggering mechanism based on pressure amplitude and differential slope. High-frequency sampling is triggered for event capture only when the absolute pressure value exceeds a set threshold and its upward slope conforms to the physical limits of human muscle exertion; and a hysteresis fallback strategy is introduced after the event ends. This mechanism effectively eliminates frequent false event awakenings caused by the patient's slow head turning and changes in gravity posture, significantly improving the overall battery life of the device while ensuring high-fidelity capture of key physiological events.

[0064] (4) This invention designs a parameter self-convergence calibration algorithm based on guided action interaction, realizing adaptive and accurate monitoring under individual differences. Due to the huge differences in oral anatomy and muscle strength habits among different patients, fixed threshold devices are prone to false alarms or missed alarms. The device of this invention forces the device into calibration state each time it is turned on and worn, extracts the resting median as the absolute zero baseline to offset the physical pretension force; and guides the patient to perform a standard empty swallowing action. The system automatically extracts the peak pressure and dynamically generates personalized event trigger thresholds. This algorithm effectively adapts to the individual differences of different patients, reduces the calibration cost for professional physicians, and improves the clinical universality of the device.

[0065] (5) This invention constructs a state-aware breakpoint resume and blind zone caching mechanism to ensure the integrity of medical-grade monitoring data. To adapt to complex offline wearing scenarios such as orthodontic patients at home and during exercise, the underlying sampling scheduling and low-power Bluetooth link state of this invention are deeply decoupled. When the system detects a wireless link disconnection, it does not interrupt the underlying physiological signal monitoring, but automatically and seamlessly switches to the batch blind zone caching mode, packaging the extracted event features and timestamps into local non-volatile memory; after the link is restored, handshake synchronization and breakpoint resume are performed first. This mechanism completely eliminates the risk of data interruption caused by unstable connection. (6) This invention adopts an asymmetric three-layer biomimetic flexible shell structure, which balances wearing comfort and sensing sensitivity. The braces-like shell innovatively adopts an asymmetric three-layer design with gradually varying thickness (0.3mm outer wear-resistant layer, 0.7mm circuit encapsulation layer, and 0.2mm inner skin-friendly stress-bearing layer). While ensuring that the whole device achieves medical-grade IP68 waterproof and high biocompatibility, the inner sensing area adopts a stress-free physical isolation bonding process, avoiding the pre-compression of the sensor caused by the deformation tension of the silicone shell itself, and maximizing the physical sensing sensitivity of the flexible thin film device to slight changes in muscle force.

[0066] The innovative aspects of this invention include: 1. Main Innovations (1) Integrated dual-path flexible sensing structure and layout optimization: A medical silicone shell that fits the shape of the dental arch is designed to integrate three flexible thin film pressure sensor arrays distributed on the labial side (orbicularis oris muscle), the anterior tongue (left side of the palatal midline), and the posterior tongue (right side of the palatal midline), so as to realize the synchronous acquisition and collaborative analysis of lip and tongue pressure information under natural closed mouth posture. (2) Event-driven multi-level adaptive sampling and dynamic power management algorithm: Based on oral function pattern recognition (swallowing, biting, resting, etc.), a state machine model with "low power monitoring (10Hz) → event candidate trigger (dual condition judgment) → high frequency capture (100Hz) → feature extraction and caching → hysteresis fallback" as the core is constructed, and intelligent linkage with BLE connection status (optimize throughput when connected, and batch cache when disconnected) is realized to achieve on-demand sampling and transmission; (3) Integrate a programmable power management chip, through I 2 The C interface monitors battery power, current, and temperature in real time, and implements multi-level power reduction strategies (downsampling when battery power is <20%, sleep mode when battery power is <5%) and active thermal protection (frequency reduction when battery power is >40°C, shutdown when battery power is >50°C) to ensure long-term reliable operation of the system.

[0067] 2. Secondary innovation points (1) Establish an anti-interference signal processing system for the oral environment, adopt a hybrid scheme combining moving average filtering (MAF) and IIR low-pass filtering, and integrate an adaptive baseline correction algorithm to estimate and compensate for slow baseline fluctuations caused by temperature changes and sensor drift in real time.

[0068] (2) Design a parameter self-convergence algorithm based on initial usage behavior to achieve "wear and use immediately".

[0069] The principle behind innovation in problem-solving is as follows: (1) The personalized custom-made brace shell is made of medical-grade silicone rubber, which perfectly fits the user's dental arch shape, ensuring comfortable and stable wear and providing a consistent geometric reference for synchronous measurement.

[0070] (2) A dynamic switching mechanism between 10Hz (baseline) and 100Hz (event) sampling rates is adopted. In the low-power monitoring state (10Hz), the algorithm monitors signal characteristics (amplitude, slope) and system status (BLE connection) in parallel. The "amplitude-slope dual-condition AND gate trigger" logic is introduced as an event candidate criterion, which greatly reduces the false alarm rate. After the event is confirmed, the algorithm responds quickly within 5ms, switches to 100Hz high-frequency sampling, and latches 2-3 seconds of context data in the pre-triggered ring buffer to ensure that the event waveform (especially the rising edge) is captured completely and with high fidelity. After the event ends, a hysteresis comparison strategy is used to confirm that the signal has recovered the baseline, and then switches back to 10Hz sampling, and extracts event feature values ​​(peak value, duration, energy, etc.) for local storage or transmission. A connection status-aware intelligent wireless transmission and caching strategy is constructed: the sampling strategy is deeply coupled with the BLE connection status to achieve decoupled management of transmission power consumption and acquisition power consumption. When connected, the MTU and connection interval are optimized to ensure real-time performance; when disconnected, the system automatically enters batch caching mode, packs data into Flash in minutes, and resumes transmission after the connection is restored, greatly reducing average power consumption and ensuring data integrity.

[0071] (3) Set multi-level safety thresholds and control strategies: When the battery level is below 20%, the sampling frequency is automatically reduced to 5Hz; when the temperature exceeds 40°C, an alarm is triggered and the frequency is reduced; when the temperature exceeds 50°C, the device is forced to shut down. The wireless charging process is also monitored, and the charging parameters are intelligently adjusted when abnormalities occur, forming a comprehensive power supply and thermal safety protection.

[0072] (4) Moving average filter (MAF) effectively suppresses high-frequency random noise.

[0073] (5) IIR low-pass filter: adopts first-order Butterworth design, cutoff frequency is set to 15Hz, filter coefficient α is set to 0.2, smooths the signal and retains the physiological frequency band of 0.5-20Hz.

[0074] (6) Automatic zero-point calibration is adopted to guide the user to rest and close their mouth for 10-20 seconds, automatically calculate and subtract the sensor baseline value; and prompt the user to complete several swallows and one light bite. The system automatically extracts the feature values ​​(peak, slope, duration) of these standard actions, and uses them as the basis for adaptive convergence algorithm threshold parameters to establish a user personal feature model.

[0075] The technological effects brought about by the innovation include: (1) With a single wear, the lip / tongue dual-path synchronous time course can be obtained in a natural posture, which significantly reduces the systematic bias caused by posture and fitting differences; without sacrificing battery life, it captures key events such as swallowing, saliva swallowing, biting / clenching with high fidelity, and forms indicators such as event count, incidence, peak, rise edge, duration / recovery time, lip-tongue coordination coefficient and event load, which are used for treatment assessment and behavioral training; the connection / energy sensing strategy reduces average power consumption and temperature rise, and reduces disconnection and packet loss; the closed loop of power supply and thermal safety improves data integrity and user experience, and improves clinical usability and follow-up efficiency overall.

[0076] (2) The anti-interference signal processing system for the oral cavity environment improves the signal-to-noise ratio of the output signal by more than 25dB and reduces the root mean square error by about 60%, effectively overcoming interference from saliva, friction and temperature drift, and ensuring the accuracy and stability of data acquisition.

[0077] (3) The personalized parameter self-learning calibration algorithm can automatically complete zero-point calibration and individual characteristic parameter convergence within 10 minutes by guiding users to complete simple standard actions, thus eliminating measurement deviations caused by differences in users' oral cavity structure and muscle strength habits.

[0078] This invention designs a wireless sensing lip and tongue muscle pressure monitoring device for orthodontic assessment. It achieves simultaneous dual-path acquisition of lip and tongue pressure without altering the subject's natural closed-mouth posture. It balances event-aware adaptive sampling with energy-constrained linkage to ensure temporal resolution of key transients (swallowing / occlusion, etc.) and multi-day battery life. Programmable voltage regulation and battery health monitoring enhance power supply stability and safety. Furthermore, it establishes an anti-interference signal processing system tailored to the oral environment and constructs a personalized parameter self-learning calibration algorithm to eliminate measurement biases caused by differences in user oral structure and muscle strength habits. This device is suitable for orthodontic assessment, myofunctional training follow-up, and related clinical research scenarios.

[0079] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A method for monitoring and sensing lip and tongue muscle pressure for orthodontic assessment, characterized in that, include: Step 1: Acquire a three-channel mixed signal including lip muscle pressure signal, pressure signal when the front of the tongue contacts the palate, and pressure signal when the back of the tongue contacts the palate; Step 2: Perform median filtering on the three-channel mixed signal, and then perform exponential moving average smoothing on the median-filtered mixed signal to obtain the baseband voltage characteristic value; based on the baseband voltage characteristic value, perform piecewise linear interpolation calculation to obtain the equivalent absolute physical force value; Static dead zone filtering is applied to the equivalent absolute physical force value to obtain a real-time hybrid pressure signal; Step 3: Calculate the short-time energy and differential slope of the real-time mixed pressure signal; when the short-time energy of the signal is detected to be greater than a preset personalized resting threshold and the differential slope is greater than a preset slope threshold, increase the signal sampling rate and return to Step 1; Step 4: Perform feature analysis on the filtered three-channel mixed signal to capture the corresponding temporal relationship, morphological pattern, and duration; Based on the temporal relationship, morphological pattern, and duration, the corresponding oral activity is identified and the identification and classification results are obtained. The effect of orthodontic treatment is evaluated based on the filtered three-channel mixed signal and the identification and classification results.

2. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, The median-filtered mixed signal is smoothed by exponential moving average to obtain the baseband voltage characteristic value. The corresponding processing expression is: in, This is the smoothed output value at the current moment. This is the median filter result input at the current time. α is the smoothed output value of the previous time step, and α is the smoothing coefficient.

3. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 2, characterized in that, Based on the baseband voltage characteristic value, piecewise linear interpolation is performed to obtain the equivalent absolute physical force value. The corresponding calculation formula is as follows: Where Grams is the mapped equivalent absolute physical force value, and ADC is the input baseband voltage characteristic value. and To find two adjacent calibration anchor points in the table that enclose the current ADC value, X is the voltage sample value and Y is the corresponding standard physical weight.

4. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, Based on the temporal relationship, morphological pattern, and duration, the corresponding oral cavity movements are identified and classification results are obtained. The corresponding identification process includes: When the pressure signal at the back of the tongue first peaks, followed by the pressure signal at the front of the tongue, and the pressure signal of the lip muscles remains below the resting threshold, it is classified as a normal swallowing pattern. When the peak pressure signal of the anterior part of the tongue is missing or below the normal threshold, the pressure signal of the lip muscles increases and exceeds the preset compensatory force threshold, and the timing of the pressure signal of the posterior part of the tongue is disordered, it is classified as an abnormal tongue protrusion and swallowing pattern.

5. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, After evaluating the effectiveness of the orthodontic treatment, the corresponding evaluation results are obtained, and the personalized resting threshold and slope threshold are updated based on the evaluation results.

6. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, While increasing the signal sampling rate, a pre-set circular buffer is latched; the circular buffer is used to cyclically store the latest data at the default sampling rate within the last 2-3 seconds.

7. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, The oral movements specifically include swallowing, biting, and resting.

8. The method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, The process of setting the initial values ​​of the personalized resting threshold and slope threshold includes: Three-channel mixed pressure signals were collected under resting closure, and the median of each channel signal was calculated for zero-point calibration. After zero-point calibration, three-channel mixed pressure signals were collected again under multiple empty swallows and multiple biting actions, and the corresponding peak pressure, duration, and rise slope were calculated. Based on the peak pressure, duration, and rise slope, the corresponding personalized resting threshold and slope threshold were calculated.

9. A method for monitoring lip and tongue muscle pressure for orthodontic assessment according to claim 1, characterized in that, The method further includes: While performing pressure monitoring, the device's power and temperature are monitored in real time; when the power is below the first preset threshold, an alert is issued; when the power is below the second preset threshold, the signal sampling rate is reduced to 5Hz and unnecessary peripherals are turned off; when the power is below the third preset threshold, the device enters sleep mode. When the temperature exceeds the first preset temperature threshold, frequency reduction or alarm is triggered; when the temperature exceeds the second preset temperature threshold, the machine is forced to shut down.

10. An apparatus for monitoring lip and tongue muscle pressure in orthodontic assessment as described in any one of claims 1-9, characterized in that, include: The braces-like shell, two lingual pressure sensing units, a labial pressure sensing unit, circuit components, a battery, and a wireless charging receiver coil; The braces-like outer shell serves as the overall support for the device, comprising an outer layer, a middle layer, and an inner layer; the outer layer is a wear-resistant protective layer, the middle layer is a circuit encapsulation layer, and the inner layer is a skin-friendly and comfortable layer. The two lingual pressure sensing units, the labial pressure sensing unit, the circuit assembly, the battery, and the wireless charging receiving coil are all located in the middle layer of the brace-like shell; the two lingual pressure sensing units are located in the anterior maxillary region and the posterior maxillary region of the middle layer, respectively; the labial pressure sensing unit is located in the labial region of the maxillary incisors of the middle layer.

Images